What I have now heard from a couple of recent clients, is the importance of actually tracking progress in AI search. As with any parts of marketing you want to creating great messaging material, content and so on; but you also need to know whether that effort is actually working or not.
I use a simple spreadsheet for doing this, attached below:
As part of my work on being seen by various AI tools I am using a simple spreadsheet to track my progress on different platforms. Just posting it here for now if anybody finds it useful:
One of the biggest blockers I have had is trying to get my head around how you can run some sort of large language model which is good enough for useful tasks, on a regular laptop. Before I looked into the details of how the magic worked, this just didn’t seem possible!
But there are some very smart people out there, and your starting point is the understanding of Ollama, downloaded from https://huggingface.co/ .. Once you have this installed and running on your laptop, you can start implementing some of the models and seeing how far you can take it based on your machine spec..
Measuring the performance of marketing departments has always been a primary objective for marketing leaders. There have been many models and proposals for how to do this, most of which I’ve used over the years.
But now there is an added dimension – how well is your AEO and GEO optimisation going? What do you need to add to your quarterly deck? The attached is my suggestion, based on work I have done before, but now with a small update for AI.
Why AI Visibility Lets Us Measure Brand Impact. Finally.
For as long as I’ve worked in marketing, there’s been one problem I could never quite solve: how to measure the impact of brand work.
We’ve always had solid tools for performance — clicks, conversions, funnel stages. But brand strength? Hand-wavey and unmeasurable. I’ve tried surveys, awareness studies, proxies like search volume. They gave some data, but they were slow, expensive, and rarely told us what people actually thought at the point of decision.
That’s why I find AI search and generative engines so interesting. I believe for the first time, we can open that black box and see the impact of your brand in action.
The Measurement Problem
When I’ve been a CMO, I always felt the tension between what the numbers said and what the market was actually feeling. Performance dashboards looked precise, but they ignored the harder question: what position does our brand hold in people’s minds?
Surveys and brand trackers tried to answer this, but they were blunt instruments. They lagged behind reality and rarely influenced day-to-day decisions.
So “brand impact” remained something we talked about, but couldn’t measure in any meaningful way.
AI as a Brand Mirror
Generative engines don’t just list links. They generate answers — which means they have to synthesize a point of view.
When someone asks an AI tool about your category, the answer is shaped by:
What the model has learned over years of training.
What it’s picking up now from your site, your thought leadership, your competitors.
That makes AI outputs a kind of brand mirror. They show what the system thinks your brand (or your category) stands for in real time. As mentioned here, “Brand” can be considered as the priors in a Bayesian model where they represent previous knowledge that we bring when making a product selection. Simply put your prior knowledge that you love Apple products has to be considered when you are searching for a product solution – you are unlikely to buy the new expensive iPhone because of a feature list 🙂
The big difficulty as we all know as marketers is that non-marketers don’t get this. This is partly our fault of course, we are notoriously bad at “Marketing marketing”. And I don’t think words are enough – we need some maths! we need to understand why these things are related at a deeper level. Then, the next time when your boss says “Why are we spending all this money on brand advertising can’t we just get some leads?”, you can link these two things together (brand spend and company performance) much more directly.
So GEO is a new and exciting frontier. But how can we measure its impact?
Five Ways AI Visibility Changes Brand Measurement
It’s real-time – no more waiting six months for a brand tracker; AI outputs shift as your content does.
It’s observable – you can literally see how your brand is described in answers, not just infer it.
It’s comparative – you can check not only how you appear, but how competitors are framed.
It’s testable – publish new messaging or thought leadership and see if it shows up in generative answers.
It’s scalable – instead of surveying a few hundred people, you’re checking the same engines millions already use.
Why This is a Breakthrough
For the first time, we can ask: what does AI say about us? And we can track how that changes.
Publish new messaging → does it appear in AI answers?
Run a campaign → does it shift how the category is described?
Invest in thought leadership → does it influence which sources are cited?
That’s measurable brand impact. Not perfect, but visible in a way it’s never been before.
Why I’m Testing This
This post is part of a wider project I’m running here on bjrees.com. Over the next few weeks I’ll:
Publish structured content designed to test how AI engines update their outputs.
Track before-and-after snapshots in Perplexity, Bing, and ChatGPT.
Share the results, whether they work in my favour or not.
If brand has always been the hardest thing to measure, maybe AI finally gives us a way in.
FAQ
Why has brand impact been so hard to measure? Because the tools we had — surveys, awareness studies — were indirect, slow, and disconnected from real buying moments.
How does AI change that? AI systems generate answers by combining historical knowledge with fresh signals. That gives us a live view of how a brand is represented.
Is this just another form of SEO? Not really. SEO is about rankings. GEO shapes long-form generative answers, and AEO structures content for short, direct responses. Together, they make brand visibility something we can observe and track.
For more than two decades, digital marketing has revolved around search engines. Brands competed for rankings on Google, invested in SEO playbooks, and built inbound marketing machines. But the terrain has shifted. Increasingly, customers don’t just “search” – they ask AI. Tools like ChatGPT, Perplexity, Bing Copilot, and Google’s AI Overviews are rewriting the journey from curiosity to consideration.
But this raises a critical question: if this is true, then what does AI say about your brand? What impact does it have on your marketing?
Unlike search engines, which display lists of links, generative engines produce answers. If your brand is absent from those answers, you are invisible – even if your site is technically well-optimized for search. Visibility in the AI era requires new strategies, grounded in both marketing practice and cognitive science.
That’s where Generative Engine Optimization (GEO) comes in: the discipline of ensuring your brand is embedded in the priors that AI systems draw on when producing content. Alongside GEO, Answer Engine Optimization (AEO) provides near-term tactics for appearing in AI-generated answers today.
This article draws on my background in mathematics, psychology, and early machine learning, together with two decades leading marketing in B2B SaaS, to explore the theory and practice of GEO and AEO – and why they matter for the next decade of marketing.
From Psychology to AI: How Decisions Are Made
Marketers often forget that human decision-making is not purely rational. People don’t start from a blank slate when choosing between brands – they begin with priors: mental shortcuts, schemas, and brand associations.
Psychologists call this process schema-based decision-making. Everyone arrives at a buying decision with a pre-existing set of beliefs. For example:
New evidence – reviews, product pages, recommendations – gets filtered through these priors (or at least “combined with”). The process is mathematically described by Bayesian updating: starting from a prior belief, and updating it with new evidence to form a posterior belief (the decision).
AI Works the Same Way Generative AI mirrors this process. Large language models (LLMs) have priors (training data, embeddings, parameters) and evidence (retrieved documents, prompts). When asked a question, the model integrates both to generate an answer – much like a human updating beliefs with new information.
That means marketers must now think in Bayesian terms not only about human buyers, but about AI engines as decision-makers.
SEO in Decline: A Saturated Market
For years, SEO was the growth engine of B2B marketing. Publish content, build backlinks, climb the rankings. But in 2025, SEO faces three challenges:
1. Market Saturation – Most categories are dominated by entrenched incumbents with high domain authority. – New entrants struggle to gain traction without massive resources.
2. Standardization – Everyone uses the same tools (Ahrefs, SEMrush, Clearscope). – Agencies follow the same playbooks. – What once gave an edge is now a commodity.
3. AI Intermediation – Google AI Overviews, ChatGPT, and Perplexity strip away organic clicks. – Even well-optimized content is paraphrased by AI, with fewer users clicking through.
SEO is no longer an offensive play – it’s defensive. If you already rank, you defend that ground. But building a new presence from scratch is brutally hard.
GEO vs. AEO: Two Paths to AI Visibility
To navigate this new landscape, I distinguish two complementary disciplines:
Generative Engine Optimization (GEO) – Analogy: Brand advertising. – Nature: Long-term, cumulative, hard to measure. – Goal: Shape the priors inside AI systems so your brand is included in their answers. – How: Structured content, metadata, thought leadership, embedding strategies, visibility in trusted sources. – Reference: Pranjal Aggarwal (2023): https://arxiv.org/abs/2311.09735
SEO: The “Middle Child” – Once fast and high-impact. – Now slow, saturated, and heavily intermediated. – Still necessary, but no longer sufficient.
Why GEO Is the Differentiator
For most marketers, AEO will be the entry point – testable, tactical, accessible. But GEO is where differentiation happens.
– It’s hard. Understanding embeddings, training data, and AI salience requires a mix of technical and strategic skill. – It’s rare. Few agencies truly know how to influence LLMs. – It’s durable. Once your brand is part of an AI’s priors, it becomes “sticky.”
In other words: GEO is the smart play for thought-leaders, challengers, and consultancies who want to own the frontier.
You need to understand marketing, Bayesian models, schema theory and machine learning to understand how cognitive and probabilistic models underlying how people make decisions about which product to go for.
Practical Strategies for Marketers
So what does this mean in practice?
1. Think Like a Bayesian – Ask: what priors do people and AI systems already hold about your category? – Provide new evidence that can shift those priors. – Balance brand-building (priors) with tactical content (evidence).
2. Balance GEO and AEO – Use AEO for immediate visibility in AI answers. – Invest in GEO for long-term brand salience inside AI systems. – Accept that both are necessary.
3. Reframe SEO – Treat SEO as defensive, not a growth engine. – Protect your rankings, but don’t bet your future on them. – Redirect resources into GEO and AEO experiments.
4. Build for Multi-Channel Demand – Inbound is no longer enough. – Combine GEO/AEO with outbound, partnerships, and category design. – Build resilience into your marketing mix.
The Human Side: Teams and Capabilities
Technology alone won’t save you. Winning in the AI era requires marketing teams that are: – Strategic – able to connect brand, content, and AI salience. – Technical – comfortable with schema markup, embeddings, and AI monitoring tools. – Adaptive – willing to test new channels before they’re fully standardised.
The challenge is not just tools but culture. Embedding AI into workflows, hiring for curiosity and rigour, and coaching teams to thrive in uncertainty are as important as any technical tactic.
Conclusion: Shaping What AI Says About You
The shift from search to AI answers is not a fad. It is a structural change in how information is mediated, how buyers form beliefs, and how brands achieve visibility.
– SEO is defensive. – AEO is tactical. – GEO is strategic.
The companies that succeed will be those that learn to shape both human beliefs and machine priors.
If you’ve ever read Zen and the Art of Motorcycle Maintenance by Robert Pirsig, you’ll know his reflections on craftsmanship go far beyond fixing bikes. As he wrote:
“[In a craftsman], the material and his thoughts are changing together in a progression of changes until his mind’s at rest at the same time as the material is right… The state of mind which enables a man to do work of this kind is akin to that of a religious worshipper or love. The daily effort comes from no deliberate intention or program, but straight from the heart.”
So what does this have to do with marketing B2B software? Everything. Quality content doesn’t just appear—it’s the product of care, thought, and a refusal to cut corners. And in a world where ChatGPT and Gemini can churn out endless filler, the question is: how do we balance speed with true understanding?
1. Quality comes from craftsmanship, not shortcuts
Pirsig’s idea of quality begins with the mindset of a craftsman: someone who approaches their work as an act of care. Marketing is not so different. It’s easy to produce “good enough” content that ticks the boxes—an SEO-friendly headline, a keyword-laden body, and a quick conclusion. But “good enough” rarely makes an impact.
Craftsmanship in marketing means paying attention to the details that others overlook. It’s about making sure the story fits the audience, the examples resonate, and the argument stands up to scrutiny. That effort creates trust with your audience.
Shortcuts, by contrast, show through quickly. Readers can tell when a piece was rushed or when it lacks real understanding. In B2B, where buying cycles are long and decisions are complex, that kind of shortcut can erode credibility.
2. It’s about how you conduct yourself to get the right results
Content is at the core of almost every marketing role. But while output matters, how you go about producing it matters even more. The process you follow—whether you take time to think, reflect, and refine—directly affects the quality of what you publish.
This isn’t about perfectionism. Few teams have the luxury of unlimited time, and deadlines are real. Instead, it’s about finding the balance between speed and depth. Are you creating something your company can stand behind, or are you cutting corners for the sake of throughput?
The companies that consistently build trust with their audiences are the ones that find this balance. They know that content is not just a task to be completed, but a reflection of their brand, their people, and their standards.
3. AI tools can’t replace real understanding
Many marketing teams are now experimenting with ChatGPT, Gemini, and other AI writing assistants. At first, it seems like a time-saver. You can generate an outline, expand bullet points into paragraphs, even draft an entire blog in minutes.
But there’s a problem. If you rely entirely on AI, you end up with content that looks right but lacks depth. It’s the equivalent of a student handing in an essay written by someone else: the words are there, but the understanding is missing. Readers pick up on that lack of authenticity quickly.
I know this from experience. I once tried using ChatGPT to write content as an experiment. On the surface, it looked fine. But the work felt hollow, and I learned nothing from the process. I decided never again—because without understanding, content has no foundation.
4. Real learning builds pride in your work
A better approach is to use AI sparingly—as a support, not a substitute—and to invest in real learning. Recently, I needed to use Docker to isolate two jobs running in Linux. I could have relied on ChatGPT prompts like: “Use Docker to create 2 new environments for me.”
The AI gave me fragments of an answer, but nothing that built my understanding. I didn’t know why containers worked, what the benefits were, or how environments could communicate. If I had stopped there, I would have had a shallow solution—and no ability to troubleshoot later.
Instead, I worked it out for myself, guided by Pirsig’s principle of focusing on quality. I learned enough to understand how Docker really worked, and I felt a sense of pride in the result. That pride matters—it creates confidence in your own expertise, and it shows through in the content you create.
5. Human insight will always beat AI slop
Ultimately, content written with care—by fallible, thoughtful humans—will always outperform what I call “AI slop.” Audiences crave originality, perspective, and voice. These are things that AI, for all its power, cannot truly replicate.
There’s also a practical point. Every AI query consumes processing power and energy. By doing the thinking yourself, you save not just your credibility but also resources. It’s a small act of responsibility in a world increasingly flooded with machine-generated text.
Quality content comes from insight, not automation. It comes from asking the right questions, reflecting on your experiences, and presenting ideas in a way that matters to your audience. That’s why, even as AI becomes more common, human-centered quality will continue to stand out.
Conclusion
In marketing, as in Pirsig’s philosophy, quality is not just about the output but about the mindset of the person creating it. AI can be a useful tool for sparking ideas, but it can’t replace the depth of insight you gain from doing the hard work yourself.
Writing content that matters takes time, thought, and pride. It’s about more than hitting publish—it’s about understanding your craft well enough to stand behind it. That’s the kind of content that will always stand out from the noise.
“If your job isn’t what you love, then something isn’t right”1
If you are not passionate or at least interested in what your company does for customers then working in marketing is quite a slog. Of course parts of the marketing role which are less interesting than other parts (I’m no fan of doing expenses…) but if you aren’t interested in the world of marketing how it works and how you actually get customers – then you will struggle to give it your all. And part of that should be having a passion project in your job ❤️
Mine is AI. I studied it at college and I’ve always been interested in how the AI world is evolving. At times I have struggled to use it in work, but that has all changed with AI taking over the world.
I’ve been working on Project Skynet for a while now, but I am particularly excited about this next stage – running everything on my laptop.
Initial Setup Stage
I’ve only put one or two notes here because it will really depend on your machine and there are plenty of other places that give much better insights about how to set up Linux on a laptop. Below is what I did on my machine.
This was the most interesting and exciting part for me. I had a presumption that it would not be possible to run this system on my laptop. Surely to be running a mini brain on my Dell just doesn’t make sense!? How wrong I was.
It definitely is the case that I can’t create ChatGPT 5 model! I might be able to run one of the ChatGPT 4 models, and I will experiment with that another time.
But I only want to run quite simple tasks so I’m actually just going to use Phi-3. as I say, I’m not planning to plot out a new moon orbit with this, just run some simple processes on my laptop.
I will pause there mostly because I was very pleasantly surprised at how easy this all was to set up. I had originally thought that it would be multiple stages to get this up and running, but Ollma has made this so simple. So for now I will leave you with my first query on my laptop, not in the cloud:
But why am I bothering to do this? Why not just run this in the cloud? It certainly would be the simpler option. The problem with doing this is you aren’t really learning anything or developing your expertise and something new and exciting. This is what I personally enjoy, and so now that I have this up and running I will move on to the next stages – building a marketing assistant that is just running on my laptop.
There are other benefits too – I feel happier about privacy and security when it is running on my machine, I also feel more in control of the spend. but there is also a strange feeling that I don’t want to “rent” the infrastructure I would rather have it on my machine and own it. This was always my ambition but, as I say, I didn’t think it could be done on a laptop. I was wrong.
If you’re growing a SaaS company, there often comes a moment when success with SMBs isn’t enough. You’re ready to move upmarket. But selling to larger organisations requires more than just a bigger sales target – it means changing how you think about customers, teams, tools, and messaging.
Here’s a straightforward checklist, based on what I’ve seen work, to help you scale from SMBs to enterprise.
1. Hire a Sales Team That Can Sell to Committees
You can’t grow into the enterprise if your sales team isn’t ready for it. Selling to one or two decision-makers in a small company is very different from selling into a team of senior stakeholders with complex buying processes. Even if your organisation is small, make sure you have at least one or two people who can manage enterprise deals properly – structured sales approach, strong qualification, and an ability to drive deals forward across a buying group.
2. Redefine Your Ideal Customer Profile (ICP)
In the early days, your ICP might’ve been simple: one end-user, one pain point. That changes when you move into larger organisations. Now you need to understand who your senior decision-makers are – IT, Finance, Operations—and what matters to them. And you still need to win over the end-users. Most enterprise deals depend on both.
3. Understand Why Your Product Solves Their Problems (What is the “Job to be Done”)
Enterprise buyers won’t connect the dots for you. It’s not enough that your product worked well for a team lead at an SMB. You need to be clear about which problems your target stakeholders are trying to solve, and why your product is the right solution. This takes more than just positioning work – it takes real insight from customer conversations and a solid understanding of the world they operate in.
4. Build a Target Account List
You need a focused list of the right organisations to go after. Use data — firmographics, technographics, and any buying signals you can get hold of. If you’ve already got larger organisations using your product, even in a limited way, start there. Prioritise warm accounts.
5. Set Up Account-Based Marketing (ABM)
You’ll need marketers who can work alongside sales and run targeted campaigns for specific accounts. Whether it’s one-to-one or one-to-few, the key is relevance and coordination. This isn’t cheap or quick to build – but it’s essential.
6. Align Sales and Marketing at the Frontline
If sales and marketing aren’t working from the same account list, or aren’t aligned on messaging and strategy, things break down. Bring in BDRs and SDRs who can work directly with field marketers. Shared goals, shared plans.
7. Build the Right Tech Stack
You’ll need the basics: a good CRM (usually Salesforce) and a solid marketing automation platform (like HubSpot). On top of that, I recommend:
These help you run coordinated outreach and target the right accounts at the right time.
8. Revisit Pricing and Packaging
Your pricing needs to match how enterprises buy. That’s not just about the number—it’s about the model. Whether you go per-user, per-transaction, annual or monthly—make sure your pricing is clear and credible. If pricing becomes a point of confusion, you’ll lose momentum. Tools like ProfitWell can help here.
9. Build Out Case Studies and Thought Leadership
If you’re not already the market leader, your content needs to show that you understand the market better than anyone else. That means writing clearly about customer problems and how to solve them. Get case studies from the best-known companies you’ve already worked with – even if the deals were small.
10. Don’t Forget PR and Analyst Relations
Enterprise buyers do their research. That includes the press, analyst reports, and third-party content. Make sure your message shows up in the places they look.
If you’re moving from SMB to enterprise, I’ve created a detailed checklist in the Resources section.
I’ve updated my marketing pyramid, adding in a new layer for LLMOs – I feel it has got to a point where a marketing strategy that doesn’t reference this new world will start to look a little dated.
I’m working on a new version of this pyramid which property understands the impact of LLM developments, though the core principles here remain unchanged. Put simply you can’t do absolutely everything in a normal sized marketing department. you have to make strategic decisions Specifically choosing to place your budgets and people to address the problems that you face today. Sometimes that will be long term LLMO work, sometimes that’ll just about getting a blog live. Sometimes you just need to go out and meet some more customers!
But you can’t do all of that today and this is where this pyramid really helps me out.
Over the past year, many marketing teams have opened Google Search Console, seen a drop in clicks, and asked: “Is our content failing?”
That’s a reasonable question—until you look a little closer. Because what’s actually happening isn’t failure. It’s a structural change in how people search.
Meet the “Crocodile Effect”
Since mid-2023, a consistent pattern has emerged in Search Console data for many sites:
Impressions are going up
Clicks are going down
This diverging trend has been nicknamed the Crocodile Effect. It’s being driven by changes in how Google surfaces information—specifically, the rollout of AI Overviews (formerly “Search Generative Experience”).
Why It’s Happening
Google’s new AI-generated summaries often answer user queries directly on the results page. These responses pull from multiple sources, cite content, and increasingly eliminate the need to click through to a website.
As a result:
You might still rank highly or be cited in an AI summary.
But the user gets their answer immediately, without clicking.
This is classic zero-click behaviour, accelerated by generative AI.
What This Doesn’t Mean
It doesn’t mean your content isn’t valuable.
It doesn’t mean you’re being outranked.
It doesn’t mean your SEO strategy is broken.
In fact, if your impressions are rising, it likely means your content is still being seen—it’s just being surfaced in a different format.
This is why first-party data matters more than ever. Many third-party SEO tools can underreport traffic by a factor of 5–10x compared to Search Console. Always trust the primary source.
What to Do About It
1. Shift the success metric
Clicks alone are no longer the best proxy for value. Visibility and influence on the buying journey—even without a click—are now key.
2. Optimize for “fan-out”
One large topic may now need to be split into multiple, specific pieces. AI Overviews tend to pull from narrowly focused content that aligns tightly to individual user intents.
Example: Instead of “Microsoft 365 Security Best Practices,” consider also writing posts on:
“Conditional Access Policy Setup”
“How to Audit Weak Passwords in Microsoft Entra”
“Microsoft Defender for Office 365 Configuration Tips”
3. Track LLM visibility
It’s not just about Google anymore. Users are also searching with tools like ChatGPT, Perplexity, and Copilot. Some marketers are starting to track presence across these surfaces, too.
What Might Come Next
While traditional web traffic may drop, purchase intent might actually rise. Users who’ve researched via AI and LLMs could arrive on your site more informed and ready to convert.
In one case I came across recently, traffic originating from ChatGPT converted at 7x the rate of regular organic. That makes sense—if the AI has already explained your value, you’re meeting the visitor mid-funnel, not top.
Final Thought
SEO isn’t dying, but it is evolving.
It’s no longer just: “How do I rank?”
But: “Where am I surfaced, and how?”
Understanding this shift—and adjusting accordingly—will separate the frustrated from the forward-thinking in the next wave of digital strategy.
I’ve been experimenting with large language models (LLMs) and vector databases like Pinecone — not just as a research interest, but as a working prototype. My goal was to build a system that could retrieve, structure, and surface my own content in a way that’s useful to both people and machines.
What started as a technical exercise quickly turned into a content strategy rethink. The more I worked with embeddings, retrieval, and prompting, the more obvious it became that most B2B SaaS content — mine included — isn’t really designed to be useful in an LLM-shaped world.
This post is a set of observations from that process. It’s not a how-to, and it’s definitely not marketing advice. It’s just a few things I’ve noticed while trying to make my content more legible — to machines, yes, but also to myself.
1. LLMs don’t skim, they distill
One of the first things I noticed was how differently LLMs process content. They’re not scanning a web page for formatting cues or crawling a hierarchy of headings. They’re vectorising meaning — pulling intent and structure from the text itself.
This rewards clarity over cleverness. Vague intros, overused analogies, and “setting the stage” paragraphs get flattened. What works best is directness: “This is what the user needs to know, and here’s what we know about it.”
2. Most content is badly stored
I had to dig through slide decks, half-written blog drafts, and internal notes to feed the system anything useful. And even when I did, it wasn’t in a format the LLM could make much sense of.
A lot of our content isn’t unfindable because it’s private — it’s unfindable because it’s scattered, fragmented, and inconsistently written. Structuring information (even just basic metadata and formatting) turned out to be more useful than adding “AI” to anything.
3. Answerability is the new readability
When I tested my system by asking questions for Syskit — “What are common governance risks in Microsoft 365?”, for example — it only worked if the source material actually contained answers. Not positioning. Not messaging. Actual sentences that respond to an implied question.
I started to think of this as “answerability”: could this content, in its current form, directly answer a user or AI prompt? If not, it’s probably not useful — not to the system, and not to anyone else either.
4. Consistency matters more than tone
LLMs are surprisingly good at detecting contradiction. If one post says we support something and another implies we don’t, the system flags ambiguity. That’s useful — but also a bit exposing.
I used to think consistency was about branding. Now I think it’s about information integrity. If the machine can’t reconcile what you’re saying across multiple assets, it won’t confidently say anything at all.
5. Structure beats style
There’s nothing wrong with good writing. But good structure — clear subheadings, defined sections, and consistent terminology — outperforms style every time when you’re working with LLMs.
Most of what I had to rewrite wasn’t because the sentences were bad. It’s because the paragraphs had no job. There was no signal about what a block of text was meant to do: define, explain, compare, warn, resolve.
Once I started thinking about content structurally — almost like documentation or an API — everything started working better.
6. You can’t fake this with ChatGPT
There’s a temptation to take short-cuts: paste your post into ChatGPT, ask for SEO suggestions, then call it LLM-optimised. But when you’re building your own retrieval stack, you realise pretty quickly that what matters isn’t how AI generates content — it’s how it understands it.
Most B2B content isn’t referenceable because it’s too shallow, too scattered, or too brand-filtered. You can’t prompt your way around that. You have to fix the source.
Final thought
Building with LLMs — even in a small way — forced me to re-evaluate how I write, store, and structure information. The tools didn’t just change the output. They changed how I think about the inputs.
In most companies — especially SaaS, tech, or B2B — the product is always evolving. New features are added, roadmaps shift, priorities change. But while your product can and should evolve, your messaging to customers, investors, and even your internal team needs to remain consistent and clear.
The risk of constantly shifting messaging is simple: if you’re always talking about the latest feature, your audience struggles to understand what your product actually does, who it’s for, and why it matters.
So how do you create messaging that stays relevant even as your product changes? The answer is to anchor your message in value — not features.
Start with a Clear Positioning Statement
The foundation of any strong product message is a Positioning Statement. This is not a tagline or a piece of marketing copy. It’s a simple, internal declaration that keeps your story focused as the product evolves.
A good Positioning Statement answers three basic questions:
Who is your target customer?
What unique value do you deliver?
How do you prove that value?
For example:
“We help mid-market SaaS companies prevent Microsoft 365 governance risks by giving IT teams full visibility, automation, and control — without adding admin overhead.”
That type of clarity gives you a reference point when you’re briefing the sales team, writing web copy, or speaking to investors. Features may change; the Positioning Statement keeps your message stable.
And of course, positioning isn’t necessarily permanent. As your market, product, or customer base evolves, you can refine it. But changes should reflect major shifts — not minor feature releases.
Define Your “One Simple Thing”
After positioning, refine your core message into your One Simple Thing (OST) — a short, memorable expression of the emotional power your product delivers.
Steve Jobs famously introduced the iPod as:
“1,000 songs in your pocket.”
He didn’t mention storage capacity, file formats, or syncing. The OST made the iPod instantly understandable and appealing.
Your One Simple Thing should aim for the same effect: immediate clarity, emotionally resonant, easy to share. This helps customers, prospects, and even your own team explain what you do.
Test, Refine, and Iterate
Early on — especially during customer discovery — your messaging will need active testing. Present your Positioning Statement and OST to real prospects. Watch what resonates. Pay attention to where they pause, what they repeat, and what confuses them.
Iterative messaging work often makes the difference between early traction and stalled growth. Strong product teams treat messaging as part of product development — not just something for marketing to handle after the product is built.
Use “Social Intents” for Early Conversations
In customer conversations, it can be helpful to use simplified “social intents” — very short, informal versions of your core message that you can test in early calls and pitches.
These are not formal pitch decks or marketing headlines. Instead, think of them as working hypotheses you test in real time:
“We help fast-growing companies stop Microsoft 365 permission sprawl before it becomes unmanageable.”
Social intents are particularly useful in discovery and early-stage sales, where you’re learning as much as you’re selling. Just be cautious where customer data or privacy regulations apply if you’re collecting feedback at scale.
Messaging Stability Is an Advantage
In fast-moving product organizations, stable messaging is a competitive asset. When your team knows how to describe the product simply and consistently, everything else gets easier: customer conversations, onboarding, sales enablement, investor pitches, even hiring.
The product will keep changing. The roadmap will keep evolving. But a clear message anchored in customer value can stay stable for years.
“And whether or not AI might already be, as some scientists believe, sentient, and there’s this little piece on the front of the Daily Telegraph this morning about an AI model that was created by the owner of ChatGPT that apparently disobeyed human instructions and refused to switch itself. Researchers say that this particular model is 03 model, described as the smartest and most capable to date was observed tampering with the code that was meant to ensure its automatic shutdown, and it did it despite an explicit instruction from searches that it should allow itself to be shut down, which is fascinating, isn’t?” – Anna Foster, Today program 26.5.2025
I’ve noticed recently how the advent of AI tools has significantly changed my work routine during the course of the week. Historically, Monday morning has been “Well that was a nice weekend, what the hell was I doing at work?”. This then takes me to my scribbled notes about what I was working on and what was left undone.
The problem with this approach is that it can keep you in the task-oriented world rather than stepping back to looking at your higher level goals. Rather than thinking “How are we going to win this market? What were those great ideas I had a few weeks ago? How can I make progress on the bigger picture?”, we jump straight into the pile of TODOs.
My way round this is to use my own AI bot “Skynet”, as an additional copilot*. All of my knowledge from the last 13 years including a recency bias adjustment accessible through a simple interface. This means that every Monday morning I can do research from my knowledge about how to address the issues at hand.
A simple example from this morning, I was looking for some research I had done a couple of years ago about marketing performance measurements – leads, MQLs, opp values, reporting approaches. Rather than trawling through my old notes, a lot of which is in my own childlike handwriting (so yes, good OCR is key to this process), I just ask Skynet and it gives me a well structured starting point which includes all of those little details that I would otherwise forget. This last point is the most important of why building your own chatbot is important. ChatGPT will give you generic answers but you have your own experience that you need to include, otherwise you might as well just do a Google search.
* Microsoft has the best brand name here, just not quite the best product yet…
So far in this project, I’ve been building a system that can scan all my blog posts, documents, and notes, extract the useful stuff, and make it searchable via natural language. The aim is to get something that works like a real-time assistant — answering questions using my own content as the source.
We’re now at Stage 4. Here’s what’s happened so far, and what comes next.
What’s Working
There are two sides to this system:
One piece of code handles data ingestion. It scans my files, pulls out the text, and stores it in Pinecone (a vector database).
The other piece lets me query that data using natural language.
The ingestion script (PopulateChatSystemDataRepository.py) currently runs manually — mostly because I’m trying to avoid hitting API rate limits. Eventually, it’ll move to Google Cloud Run so it runs continuously without needing my laptop open.
On the querying side, I started with basic keyword search. It was fine, but not great — too brittle. Now I’m using embedding-based retrieval with Pinecone, which is far better at handling fuzzier, more conversational queries.
The current setup includes a FastAPI service deployed on Cloud Run. It accepts queries via a simple URL. For example:
Type that into a browser, and it returns a relevant result from my content. It’s rough around the edges, but it works.
Why Speech-to-Text Is in the Mix
You might notice I’ve already wired in Google’s Speech-to-Text API — even though I’m still in the text-only phase. That’s for later. Eventually, I want this system to handle real-time conversations — voice in, answers out. But for now, I’m keeping things simple.
What’s Next: A Text Interface
This is the next step. I want to build a simple text interface — something that lets me talk to the system like an old-school text adventure game. No need for a fancy UI yet. Just a clean loop where I type a question, the system replies, and I can keep the conversation going.
Why this? Because before I worry about polish, I want to know the core experience works — the retrieval is accurate, the flow makes sense, and I can actually use it.
The checklist:
Add all my content to the system (done)
Build a basic interface for interaction (next)
That’s Stage 4. Getting the interface up and running is the next focus — and from there, it gets a lot more interesting.
When building search tools, intelligent assistants, or AI-driven Q&A systems, one of the most foundational decisions you’ll make is how to retrieve relevant content. Most systems historically use keyword-based search—great for basic use cases, but easily confused by natural language or synonyms.
That’s where embedding-based retrieval comes in.
In this guide, I’ll break down:
The difference between keyword and embedding-based retrieval
Real-world pros and cons
A step-by-step implementation using OpenAI and Pinecone
An alternative local setup using Chroma
Keyword Search vs. Embedding Search
Keyword-Based Retrieval
How it works:
Searches for exact matches between your query and stored content. Works best when both use the same words.
Example:
Query: "What is vector search?"
Returns docs with the exact phrase "vector search".
Pros:
Very fast and low-resource
Easy to explain why a match was returned
Great for structured and exact-match data
Cons:
Doesn’t understand synonyms or phrasing differences
Fails if the words aren’t an exact match
Embedding-Based Retrieval (Semantic Search)
How it works:
Both queries and documents are converted into dense vectors using machine learning models (like OpenAI’s text-embedding-ada-002). The system compares their semantic similarity, not just their words.
Example:
Query: "How does semantic search work?"
Returns docs about “meaning-based search” even if the words are different.
Pros:
Understands intent, not just keywords
Great for unstructured content and natural queries
Can surface more relevant results even if phrasing is varied
Cons:
More computationally intensive
Results are harder to explain (based on vector math)
Requires pre-trained models and a vector database
Feature Comparison Table
Feature
Keyword-Based Retrieval
Embedding-Based Retrieval
Search Logic
Matches words exactly
Matches by meaning
Flexibility
Low
High
Speed
Fast
Slower
Resource Use
Low
Higher
Explainability
High
Low
Best For
Structured search
Chatbots, recommendation, unstructured data
Common Tools
Elasticsearch, Solr
Pinecone, Chroma, FAISS
Setting Up Embedding-Based Retrieval
Let’s build a basic semantic search system using:
OpenAI (text-embedding-ada-002)
Pinecone (hosted vector DB)
Chroma (optional local alternative)
1. Choose Your Tools
Embedding model:
OpenAI’s text-embedding-ada-002 or a local Hugging Face model.
This is the part where all the content sources came together into a centralized system I could actually interact with.
This post is a cleaned-up record of what I built, what worked, what didn’t, and what I planned next. If you’ve ever tried to unify fragmented notes, decks, blogs, and structured documents into a searchable system, this might resonate 🙂
What I Built
There were two main components at the heart of the system:
Batch Processing Script PopulateChatSystemDataRepository.py — this was run manually to gather and format all source data into a single repository. My plan was to automate it later.
Continuous Scanner A lightweight background service monitored for new blog posts and updates.
At that point, the batch script did the heavy lifting, though I intended to shift it onto Google Cloud Run to handle scale.
Where the Data Lived
The sources I processed included:
PowerPoint files These were manually selected and hardcoded into the script — a reasonable tradeoff given how few I needed to track.
Notes from a Cambridge Judge Business School programme
Third-party and personal research logs
iCloud Backups These contained archived slide decks and supporting materials.
All of this data was funneled into a staging area for eventual vector embedding and retrieval.
Microsoft Graph API + OneNote
To pull content from OneNote, I used the Microsoft Graph API. First, I installed the required libraries:
pip install msal requests
msal handled authentication via Azure Active Directory
requests allowed me to interact with the Graph API endpoints
Once I authenticated, I could enumerate and query notebooks like this:
python ExtractNotes.py
After logging in via a Microsoft-generated URL, I could successfully extract content from all the notebooks I needed.
Licensing Curveballs
At the time, I hit a snag: my Microsoft 365 Family plan didn’t include SharePoint Online, which was required to query OneNote via the Graph API.
I weighed my options:
Pay for a Business Standard plan (~£9.40/month)
Try and use my home license in some way, even thought it didn’t seem to have what I needed for OneNote
I went with option 2, supported by a one-month free trial of Microsoft Business Basic to help validate the approach.
Google Sheets as the Backbone
The ingestion script used a JSON keyfile to interact with Google Sheets. It opened the sheet like this:
client.open_by_key(sheet_id).sheet1
Sheets acted as a live database — but I ran into 429 rate-limit errors, especially when repeatedly reading the same files. To solve this, I built a basic checkpointing system so the script would:
Cache previously processed records
Avoid re-downloading the same content every time
Track progress and only fetch new entries on each run
The GitHub Reset
After a short break from the project, I realized the codebase had grown too complex. I had introduced a lot of logic to deal with throttling and retries, but it made everything harder to understand.
So I rolled back to a much earlier commit and started again from a simpler foundation.
It was the right move.
What Came Next
Here’s what I tackled after that cleanup:
Migrated the whole project to an old home laptop
Simplified the ingestion pipeline
Ensured each run processed only new data, not the full archive
Finalized access and querying via Microsoft Graph API for OneNote and SharePoint content
Reflections
Skynet began as a chatbot experiment, but evolved into something bigger — a contextual knowledge system that drew from years of notes, presentations, and personal writing.
Stage 2 was about turning chaos into structure. The next phase was even more exciting: embeddings, retrieval, and building a system that could answer real questions, grounded in my own work.
Over the past few months, I’ve been building something a bit different: a real-time AI-powered assistant designed to help me work better with my own content. The goal is to create a system that can scan and catalog documents, blog posts, audio recordings, and notes, then surface that information back to me as I need it—almost like a second brain. I wanted it to pull from tools I already use daily, like Google Sheets, OneNote, and GitHub, and use technologies like Pinecone, OpenAI, and Google Cloud to power the intelligence behind it.
This blog series is a step-by-step breakdown of how I built it—from a messy OneNote notebook into a working system. Each post will focus on one key stage, including the code, architecture, and lessons learned along the way. This first post is solely about the tech stack that I chose, actually one of the most fun stages.
Setting Up the Environment for a Python-Based AI Chatbot
This project runs on Linux, primarily because I want to use Python, which I have some basic experience with. Here’s how I set up the development environment and supporting tools.
Core Tools and Services
Google Cloud
Speech-to-Text for audio transcription
Cloud Run to execute background processing tasks
Google Sheets for structured data storage (e.g., cataloging blog posts)
OneDrive (Personal) for general document storage
iCloud Drive for mobile voice recordings
GitHub to manage Python code and version control
Trello as a lightweight project tracker
ChatGPT to assist with development and planning
Getting Linux + WSL Working Smoothly
There were a few initial stumbling blocks in getting everything up and running, especially around Windows Subsystem for Linux (WSL). Here’s the distilled process:
Launch CMD as Admin, then enter WSL with: wsl
Activate the virtual Python environment: source myenv/bin/activate
Navigate to the correct project folder: cd ~/skynet/
Code and Repository Setup
All code is version-controlled in GitHub
To update code:
git add PopulateChatSystemDataRepository.py
git commit -m "Message"
git push origin main
git pull origin main
Libraries required when switching machines:
pip install gspread
pip install oauth2client
Purpose of This Setup
One of the key tasks here is to catalog blog posts into a Google Sheet using Python:
python3 SendDataToGoogleSheets.py
This setup forms the backbone of a knowledge base that can be queried by an AI chatbot.
Why Use Google Cloud Run?
I plan to regularly publish new blog posts and documents. These need to be automatically picked up by a cloud-based system, not just left on a drive. To do that, I’m using Google Cloud Run to host the background process that parses and ingests this content.
The service is named skynet, though it hasn’t been deployed live yet—waiting until the code is fully tested.
Setting Up the Google Sheet Database
Create a Google Sheet and give it a meaningful name (e.g., “Chat System Data Repository”).
Enable the Google Sheets API in the Google Cloud Console.
Set up the API key and credentials for access.
Code integration is done using Python with the gspread library—no Zapier or low-code tools.
Data Format for Each Entry
Each blog post entry should include:
Message ID
User ID
Timestamp
Message Text
Source (e.g., Slack, WhatsApp)
Response Status (e.g., Processed, Pending)
Parsed via Python’s feedparser, which extracts standard RSS fields such as title, link, description, content, and publication time.
Next Steps
There are two major next steps:
Add additional content sources into the pipeline (see Trello board).
For any new source, take the data through the same ingestion process.
Currently, everything runs through the PopulateChatSystemDataRepository.py script, which has been updated to handle edge cases like escape characters.
Now that the core data is in place inside a Google Sheet, the next stage is testing the pipeline end-to-end. Once that’s working, I’ll expand to include additional data sources.
Such a well known quote, it barely needs a reference (but I will do anyway – it’s Peter Drucker).
One of my side projects is an attempt at taking all of my own knowledge from the last few years of marketing and putting them into an AI model. One of the issues with a platform like ChatGPT is that it will give you answers based on “What the world thinks”. For 95% of cases this is great of course – if you want to know the capital of France, you want the consensus, not an opinion. But these models don’t reflect the more difficult parts of marketing, the knowledge gained at the coalface of trying out real campaigns and seeing them fail or win – the things that make you stand out from your competitors.
As an example I have recorded short posted a short video below of what my own personal AI tells me about company culture. This is based on years of notes on marketing strategy and so genuinely reflects my view rather than perhaps a more generic outlook.
I also believe this is an important part of marketing. If you are saying the same as everybody else in your marketing then no one will read it. Why would they? You need to present an opinion.
I work at https://www.syskit.com/, and I have spent the last few weeks trying to figure out why things are going so well for performance of the company. You can look at charts, you can create models but somehow they don’t quite give you the answer. The one thing I keep coming back to is the culture. A culture of openness, mutual support, great people and, the secret sauce – far fewer meetings – has created a culture where great work can be done and enjoyed. I listened to a great program this week on workplace culture which I could more or less summarise as “Fewer meetings” 🙂
I have abandoned attempts to control the whole thing by voice as it just seems an enormous amount of work the minimal return. It’s cool, but not practically helpful.
Instead and I think more interesting, is going into the world fine-tuning more. If I just wanted general information about marketing, I can do a Google search. The value comes from producing something based on my specific views. As a simple example, I attach much more credence to expertise to search engine optimization. This isn’t the view held by everybody in marketing :).
The other interesting point is about security. I have thought about putting this chatbot on the web for anybody to use. But I would need to do an audit of every single piece of information in the underlying database. So for now this is something that I use in general conversation as an aide.
Anyway, I will post more videos as it starts giving me more and more insight.
I’ve had a lot of fun over the last few weeks building a marketing chatbot. I started here and thought “How difficult can it be?!”
Two months later I have found out both how difficult some of it is, but also how easy some of it is. The easy bits were putting together sources of information and using ChatGPT to help with the structure of the app. The most difficult bits were anything to do with permissions (even on my own data) and anything where I needed to use Microsoft Graph API. Still a long way to go, but I thought I’d post an update on the tech stack I’ve used so far.
Screenshot of the app so far! I will publish the real thing when I’ve had a look at the security implications…
Tech Stack
The very simple architecture/workflow for the app can be split in half:
Server-side: Run various processes to collect together everything I know about marketing and put it into a database.
Client-side: A way of running queries against that database
Having these guidelines helped enormously with choosing the technology and also helped to make sure everything worked together.
I wanted to use the following guidelines:
As free as possible. Only pay for something if the cost was minimal and it saved me an enormous amount of time/reduced complexity
Linux based. Again for cost purposes, but also because it’s just much easier to use 3rd party libraries
I ended up with the following, first for server-side then for client-side.
Server-side
Google Cloud. I have used Google Cloud in three places. If you are in the world of Linux then it is so easy to call the Google Cloud API, the online help is superb and it is relatively easy to use. But the biggest winner was the integration between GoogleSheets and the API. I thought this would be tricky, but actually you were guided through it very simply. Specifically I used it for:
Speech-to-text encoding
Cloud Run to run the independent scans of various types of content
Google Sheets as document storage, including a sheet which lists all of my blog posts
Microsoft Personal OneDrive for documents. This is a significant source of information as I have used Microsoft personal OneDrive for a decade or more to keep notes on various ideas. With hindsight, I would have kept this much cleaner and better governed. There has been a lot of work cleaning up what was in here in all sorts of formats
GitHub for storing all Python code. A private repo for now
ChatGPT to help me with everything. It would be a lie to say that I wrote everything from scratch. I went to a very impressive lecture recently at the Cambridge Marketing Meetup from Siok Siok Tan who talked about how, going forward, people would work side-by-side with the machines splitting out tasks as appropriate. The way forward for this sort of work is being smart about what you should do and what the machine should do. What I found particularly interesting was the symbiotic relationship needed for writing code. For example, chat gpt wrote quite a lot for me as a starting point but I usually found I had to do a reasonable amount of work on top of that, because the ChatGPT sources were out of date, and APIs seem to be getting updated all the time.
Microsoft Graph API. Sigh. I use this extensively to programmatically access my OneNote notes. I had just finished working with the Google technology and thought “Surely the Microsoft APIs are just as easy?”. No they are not. A lot of this is to do with security of course, and that is good as I don’t want everybody reading my personal files. However the structure of OneNote files is nontrivial and getting API access to work correctly there’s a lot of hard work.
Client-side
Pinecone was a big breakthrough for me. I want a chatbot which isn’t just a simple text search. I wanted to try and do something which allowed me to use some sort of fuzzy search mechanism. As an example, my records might say something like “Account based marketing is a great strategy”, But I want that to come up if I searched for something like “What is ABM?”. I needed something more than just simple keyword based retrieval, so moved over to using embedding based retrieval with OpenAI.
Docker to containerize everything. Is that a real word!?
Google’s speech-to-text API. I’m going to save this for the 3rd blog post, partly because it is just so impressive it deserves a post of its own.
That will do for now. The next stage is all the testing, particularly to make sure the time using the right security for my documents. One of the key issues with using AI technologies on your own notes (essentially what I’m doing here), is the problem of oversharing information when you shouldn’t. It is something that we help customers with at Syskit, albeit in a different context (Microsoft 365 governance). But the principles are the same – are you sharing things outside your organisation that you shouldn’t be?! Be careful out there… I will eventually put a link in here that links to my app, but not until I have gone through all of the information and documents that I am using to make sure I am not oversharing. I will need to do this by hand so I may be some time…
I’ve started a new project to try and write some Python code so that I can talk to the laptop in real time as I’m working. The ultimate goal is to create a real time assistant who is making suggestions for me in conversations.
Today was day one. Despite many many false starts I finally managed to get the laptop to recognise my voice and start transcribing it. Let Judgement Day begin.
I started doing some manual research into the security risks with different AI tools. But then I thought, why not get the AI to do it for me? So that’s what I did. Once again I am very impressed…
1. Data Privacy and Confidentiality
ChatGPT: Users may inadvertently share sensitive information, potentially leading to data leaks. ChatGPT conversations may not always have the same enterprise-level data security controls, depending on deployment.
Gemini: Google advises against sharing confidential information with Gemini, as conversations may be reviewed to improve quality, raising privacy concerns for sensitive data. (searchenginejournal.com)
Apple Intelligence: Apple prioritizes on-device processing and privacy, with Private Cloud Compute to protect user data. However, the effectiveness depends on consistent use of these privacy features by users. (security.apple.com)
Microsoft Copilot: Integrated within Microsoft 365, Copilot has enterprise-grade security and compliance features built-in, including support for GDPR, HIPAA, and other regulatory standards. However, the risk of users accidentally sharing confidential data through Copilot remains, especially if the model is trained on organization-specific data without strict data policies in place.
2. Phishing and Social Engineering
ChatGPT: Can be exploited to craft convincing phishing emails or social engineering scripts by generating content that mimics corporate communication styles.
Gemini: Google Gemini has been found to have vulnerabilities that could be exploited for phishing, potentially enabling attackers to take over chatbots or impersonate users. (securityweek.com)
Apple Intelligence: While Apple Intelligence AI hasn’t been specifically linked to phishing exploits, any AI with language generation capabilities could potentially be leveraged for social engineering if misused.
Microsoft Copilot: As it interacts with Microsoft 365 tools, Copilot has the potential to automate and personalize phishing messages within Microsoft’s suite, particularly within Outlook or Teams. Enhanced by organizational knowledge, phishing attacks crafted by Copilot could mimic familiar internal communication patterns, making them harder to detect.
3. Malicious Prompt Injections
ChatGPT: Susceptible to prompt injection attacks that could manipulate the AI’s behavior to provide unintended or sensitive information.
Gemini: Vulnerable to indirect prompt injection, which could enable phishing or chatbot takeovers. (securityweek.com)
Apple Intelligence: Apple has measures to guard against vulnerabilities and offers rewards for identifying AI security flaws. However, the risk of prompt injection remains if used in complex workflows without safeguards.
Microsoft Copilot: As Copilot becomes embedded across Microsoft 365 applications, prompt injections could potentially allow malicious users to exploit workflows or access sensitive data by manipulating Copilot’s responses. This is particularly concerning in applications where sensitive data is routinely processed, like Excel or SharePoint.
4. Data Exfiltration and Unauthorized Access
ChatGPT: Without proper security configurations, there is a risk of data exfiltration if ChatGPT is misused or linked to sensitive applications.
Gemini: Accused of unauthorized data scanning on Google Drive, Gemini has raised concerns over potential data exfiltration or access without user consent. (techradar.com)
Apple Intelligence: Apple’s approach emphasizes on-device data handling to mitigate unauthorized data access, though secure implementation and user adherence are necessary to minimize risk.
Microsoft Copilot: As an AI system integrated with Microsoft 365, Copilot has extensive data access, which could be exploited if permissions are misconfigured or if attackers find ways to bypass controls. Because Copilot can access files, emails, and other stored information, this could expose sensitive data if not closely monitored.
5. Compliance and Regulatory Risks
ChatGPT: Compliance risks may arise if sensitive or regulated data is inputted, especially if stored outside organizational control, potentially violating GDPR, CCPA, or other regulations.
Gemini: Google’s Gemini AI practices could pose regulatory challenges, particularly around user consent, data retention, and control over how user data is handled.
Apple Intelligence: Apple’s privacy focus and on-device data processing align well with regulatory standards, but enterprises must ensure their specific use cases remain compliant with industry standards.
Microsoft Copilot: Copilot aligns closely with Microsoft’s regulatory and compliance frameworks, making it a safer choice for organizations bound by strict regulations. However, organizations still need to ensure data governance policies are enforced to avoid regulatory risks related to AI-driven data processing.
Conclusion
While ChatGPT, Gemini, Apple Intelligence, and Microsoft Copilot each bring distinct features and security controls, core security risks are common across all platforms. These include data privacy, phishing, prompt injection vulnerabilities, data exfiltration risks, and compliance challenges. Microsoft Copilot offers the advantage of built-in enterprise security and compliance support, but also presents risks, especially around data handling, unauthorized access, and phishing automation.
For all platforms, implementing clear data-sharing policies, monitoring AI interactions, user training, and regular security audits will help organizations mitigate these risks effectively.
Most of your time as a marketing leader is spent trying to make decisions with inadequate data. In an ideal world, we would have run an A/B test on everything we wanted to do, looked at the numbers and then made a decision. Which image should we use for our new advert? What message? What tone? Which type of customer are we trying to reach? And 1,000 other things.
A/B testing is one way of approaching this problem. The difficulty is that most marketers will – and should! – already have a view. If I was given the choice between two headlines:
Find out how our products can help you
Click here to give us some money
I know which one I would click on, I definitely don’t need to do a test!
But here is a more realistic example. You are trying to sell into a company and you are not sure who makes the decision. Is it the end user? The manager? The person holding the strings?
How on Earth do you do an A/B test for something like this?
You will soon find with a question like this, that you quickly hit the “Everyone has an opinion“ problem. You ask various members of the team and outside your team and everyone gives a different answer. There are a couple of ways out of this situation as I’ve mentioned, but doing an A/B test is generally completely impractical.
So what can you do? The approach that I take now is to use some of the concepts from Bayesian logic to help me make the decision. The key concept is is the idea that every decision you make is a combination of your prior knowledge plus the data that you see. And the real issue with prior knowledge is that everyone comes to the table with their own history.
As an example, if you have been running a marketing team for years, and all you have been doing is numbers-driven digital marketing for B2C businesses – and crucially, you have had success with that, then you are going to start your analysis with that approach in mind – the answer to the question “what should we do next for our marketing?“ will very likely be something around digital marketing strategy. In contrast, if you come to the table from a brand marketing background, then it is likely that your initial opinions will favour this sort of approach. Why? Because this is what you know and there’s a good chance you’ve had success with it at some point in the past.
Crucially, just asking the question “which is right?“ will not get you anywhere! You each have prior knowledge that you are bringing into the process. So what do you do when you start running a campaign and you start to get results, albeit with very low numbers? How do you combine your prior knowledge of what should happen with what is actually happening?
This is where the Bayesian approach can be very useful. I don’t think you need to understand any maths to use this approach, it is about the principles behind it.
I first read about this principle in the book below:
I have recommended this book about six times on this blog, so I am definitely a fan! The key part that is relevant to this blog post I have copied below. I tried to paraphrase it, but then I realised that Sean Carroll’s short explanation is better than anything I could come up with:
“Prior beliefs matter. When we’re trying to understand what is true about the world, everyone enters the game with some initial feeling about what propositions are plausible, and what ones seem relatively unlikely. This isn’t an annoying mistake that we should work to correct; it’s an absolutely necessary part of reasoning in conditions of incomplete information. And when it comes to understanding the fundamental architecture of reality, none of us has complete information.
Prior credences are a starting point for further analysis, and it’s hard to say that any particular priors are “correct” or “incorrect.” There are, needless to say, some useful rules of thumb. Perhaps the most obvious is that simple theories should be given larger priors than complicated ones. That doesn’t mean that simpler theories are always correct; but if a simple theory is wrong, we will learn that by collecting data. As Albert Einstein put it: “The supreme goal of all theory is to make the irreducible basic elements as simple and as few as possible without having to surrender the adequate representation of a single datum of experience.”
Everyone’s entitled to their own priors, but not to their own likelihoods.“
This might feel like a slightly obscure deviation from the subject matter of this blog (marketing!) but I don’t think it is.
Unlike many other areas, It is very difficult to come up with definitive evidence for why one approach is better than another. This can lead to the back and forth, debate about messaging and other areas.. or worse than that, you can even end up with the HiPPO principle for making decisions (“Highest Paid Person’s Opinion”).
But – If you understand that this is where a lot of people are coming from, that they are making decisions based on their prior experience and not necessarily the facts in front of them – then this makes it much easier to have a rational discussion. There will be a very reasonable logic behind why somebody is arguing for something. Listen to that person, interpret and apply intelligently.
This is a personal view. All I can write about is what I find to be a great culture, and I know others will have a very different point of view.
And I definitely have it in my employer, Syskit. I have been lucky to work at some great companies (and some not so great companies…). But for some very specific reasons I believe Syskit to be at the top of that group. I have now been there for a full year so I wanted to do a quick review of why it’s so good, in no particular order.
Fewer meetings
As a company grows it gets more and more difficult to keep to a common path and collaborate effectively across so many different people.
The answer to this problem is most definitely not having more meetings though. At first look, this seems like a good response. We all want to keep in sync with each other we’re struggling working remotely (if this is the case). Why not put in some regular weekly meetings or “catch-ups” to achieve that?
From my experience, a great company culture needs as little time as possible in meetings. Why? First of all the obvious – nobody likes meetings! But there is something more subtle. There are some very important meetings that are needed for a company to run and those should be kept. What I believe should be lost, and what I see in great companies is the removal of “catch-ups” and other weekly meetings. My starting point with all of my colleagues is that I trust them all and know that they know their jobs. Why then do I need a meeting to check in? What information is actually being communicated, that couldn’t be sent in an email?
At Syskit, I have now managed to get my weekly meetings down to about 4 or 5 per week. Most days, I have days with absolutely no meetings at all. Crucial to a well run company because it frees up your time for more impactful deep work. Looking back at the times when I have made big strategic changes or fixed really knotty problems they are always on days which are completely free. This is why I protect them so closely.
Crucially, this is the culture at Syskit and it fits me very well 😊
Informality
It is hard to explain why this is so important to me and it’s not just a “work thing”. It is important to have structures and models in the workplace, but that doesn’t mean to say we need a culture of formality. I can genuinely say that I enjoy the company of my colleagues, that we chat about things other than work, and that we “get on”.
My test for this is always:
“If you were told that you were going to be stuck on a three hour train journey with a colleague with no books or phones, how would you feel about that?”.
I believe this tells you a lot about whether you are on the same wavelength as your colleagues and for me informality is very important. We are not doing God’s work here (as an old boss put it), and I believe informality is a crucial part of any culture of smart people.
A great product
Why should this matter when we are talking about culture? Surely a great product is just about your company’s commercial viability?
I think it is a lot more than that. You want to believe that you are genuinely creating something great. You might not be saving the world with your product, but you should be proud of it and for me, that is a really important part of company culture.
Again, I have a test for this:
“Would I recommend this product to a friend? I mean, a “non-work” friend?”
I have worked at places where I wouldn’t recommend the company’s product to my friends, and that is fine. But I would recommend Syskit’s product to friends (who happen to be Microsoft 365 administrators! – that’s not all of them 😊), and that means that there is an honesty about the work which I think is important.
This is strongly related to another point which is a focus on the numbers. Is the company hitting its targets?
Again, why does this matter when we are talking about culture? Personally, I much prefer conversations in the workplace which are about customers, the product and whether the company is doing well. I find this more interesting than conversations about HR policy, organisation, and holiday entitlement.
So what I like in a workplace is spending time talking about the commercials and driving the success of the company. It is only the numbers that show whether you are succeeding in your role and again, that is part of the culture.
Expertise
Again, a strange one perhaps. But I would argue in marketing in particular, that there is a very wide range in expertise in the market. Marketing well is very difficult. It is subtle, complex, often unintuitive and I would argue that B2B are marketing is just a completely different field to B2C. Budget allocation is difficult; explaining what you are doing to non marketers is difficult; getting alignment between marketing and other areas is difficult; focusing on three things instead of 30 is difficult. And getting into the minds of customers is the most difficult thing of all.
Working with other experts and experienced people is crucial for me and I believe crucial for the success of any company. If you are working with people who think that marketing is just about optimising some Google ads then I believe you might struggle to grow in that environment.
All of this is why I think expertise in a company is crucial to the culture. You need to be having intelligent conversations with people about tricky problems and personally, that is what I enjoy most about it.
If any of this sounds even remotely interesting and you are looking for a new role in any field – marketing, development, product, sales – then reach out to me through LinkedIn and let’s talk.
We’ve been doing some digital marketing work recently and the more and more time I spend on digital work the more beasts I feel need to be slain.
NB: I’m talking specifically about B2B marketing here – which is important. It’s important because many of the problems that B2B marketers face come from taking a “copy and paste” approach from B2C in to B2B. But I think these two jobs are completely different.
Myth 1 – A/B testing is valid when writing copy
i’m not a fan of A/B testing generally, mostly for statistical reasons – these tests are almost never done on a large enough volume to be valid. But even if you did have an enormous data set, would it still be useful?
I don’t think so. When potential customers are looking for a product that fulfils their needs, the language that you use, particularly on a digital advert, has to be as good as perfect as it can be. Not just the words, the insight, the phrasing, the context and so on. We’ve all seen ads where the copy is just “not quite right”. Are you presenting your product In the most appropriate way? Should you be describing a feature or the advantages and benefits? Should you be targeting someone more senior or more junior? Should the wording be laser focused on a specific use case, or more generic?
The answers to these questions won’t come from an A/B test. They will come from sitting in front of a customer talking to them about their business and drivers, then finding a way to formulate that into something appealing and simple. And that’s why marketing is such an interesting discipline to work in!
Myth 2 – More is better
Surely if you have 10 different messages going out to customers about 10 different value propositions, that’s better than one or two? Surely?
I don’t think so. I love the phrase “You will get bored of your marketing before your customers do”. If we are lucky, very lucky then our customers will be able to remember one thing about us, about what we do. It might be something like “Do you do security software or something?”. Or “Are you a Microsoft add-on?”. To try and get this message inside the heads of potential customers, it has to be repeated over and over and over. Then, if you are lucky, when they have a problem that you can solve they will have an aha moment when they remember “oh yes Syskit, they do something for that don’t they?”. They will then Google search your company name, find you read your website and make a decision about whether to go further.
This is a massive win – It’s your brand advertising dollars at work. That you undermine this advertising if you keep chopping and changing what you are saying. If one day you’re selling on price, the next day on functionality, the next day on something else then they won’t know what you do and they won’t think of you when they have a problem you could solve.
So choose the single killer feature, figure out why customers should care and then repeat, repeat, repeat.
Myth 3 – If I can’t show the ROI of a campaign, I shouldn’t run it
Perhaps one of the most dangerous in marketing. There are ways of showing the ROI of certain sorts of activities, for example I think it is possible to show the return on exhibiting at a conference (add up the spend, add up the opportunity value from the people who attended over the subsequent few months, and so on).
But for 95% of what you do, this isn’t possible. And this is where the big difference between B2B and B2C marketing is apparent. I believe it is impossible to measure “the experience” of the customer interacting with your advertising or not. For most messages that a customer sees, that isn’t measured anywhere, particularly not by Google. Of course they say they do, but if you spend time with the numbers you realise how much is missed.
Given this, I feel there’s an enormous amount of value that comes from certain sorts of marketing work, weather content, advertising or whatever. But it would be very dangerous to switch that off just because you couldn’t “quote “prove” its value. You wouldn’t understand the mistake that you had made until it was too late, when you have cut the advert and moved on to something else. So have faith that it is working and keep your eye on the messaging, to make sure you’ve got it laser focused.
Myth 4 – Exhibiting at events is a waste of time
Events are expensive to attend. The exhibition fees, travel, hotels, meals and much more. So the question is, are they worth it?
I think they are but not necessarily in the most obvious direct way. For me, meeting potential customers in any way is the most important activity you can do. It is very difficult to just bump into customers so you need to find somewhere where they congregate (NB: going to visit them one to one is also a great use of time).
The reason I think it’s so important is because you can have proper in depth honest conversations with attendees. What do they really value? What do they really think of your company? Who else do they like and why? I have often spent 20 minutes with a customer on a stand going through the details of their problems, and a lot of that content went straight into adverts or blog posts the next week. It makes the copy very easy – Just parrot back what’s your visitors said, with a little anonymity and hey Presto! It almost feels like cheating.
There is of course a question of ROI which often comes up. And that should definitely be considered – You shouldn’t be flying around the world for an expensive event while there will only be 15 attendees. But assuming you’re making smart decisions about budget and the types of people who attend then, it’s very possible you will generate some interest which will cover your costs and you will get the incredible insights about the market for free.
There are many more myths to be slain, and I’ll add them in as I remember them! At Syskit, we are clearly a 100% B2B company. All the marketing we do is in that model. This makes working here much easier as you know which advice to take on and which to ignore. It also focuses your time more on understanding customers, and a bit less on the latest tricks and trips from Google.
I’ve written a few times before about AI and it’s impact on marketing. I’ve written here about the impact on Google search, and here about openAI and ChatGPT.
Finally though Microsoft is making a bit more of a song and dance about their offering, Copilot. I was sceptical at first because a previous version I had tried wasn’t great. But now I am nothing but impressed. What I’ve been most impressed by is the option to make your own chat bots based on your own data using Microsoft Copilot Studio. So what I’ve done here is take all of my old blog posts about marketing from the last 10+ years and created a Copilot! A sort of “B2B marketing Copilot”.
It’s been a great first few weeks at Syskit. I’ve really enjoyed it, and I’ve loved meeting the team. But more than this, I can directly see why the company is going to keep growing at an even faster pace than it has so far.
Why? Is it the product? Is it the strategy? The market? Yes to all of these, but more than that it’s the culture of the company. It’s an open, communicative organisation with the right mix of autonomy and alignment.
Even more than that though, the biggest strength I found is simply the lack of meetings! I’ve seen both sides over the years – places with multiple half-hour meetings every day and now Syskit where my calendar is pretty empty. The reason I think this is so strong strategically is that by freeing up people, up it gives them the chance to actually do the work and make a real impact. It’s a strategic decision because you pay a price of losing constant comms and interaction. But the payoff is worth it – multiple days in “flow“, feeling that you’re making significant progress on important tasks. I see more clearly now how time slicing is a killer for productivity and I would struggle to go back to that approach!
I’ve worked on projects where there has been a tenfold difference in the productivity of the teams. Specifically, given a piece of work like “Let’s build a new campaign” or “Let’s redesign the homepage”, I’ve seen the same team size do that in one week or in 10 weeks. What was that difference? Specific processes? Using a standard framework? Project management approach? How do you create a culture of speed?
I don’t think so. From what I’ve seen over the years it’s something quite specific – a culture of trust and honesty – which then indirectly leads to much faster output.
I believe a high trust/honest culture has the following traits:
(almost) no meetings. The only meetings you should be having are celebratory – celebrating a new release or a deal closing. Why? Firstly, the obvious, almost everybody hates them! Most meetings are run on the HIPPO principle, which almost everyone dislikes. But why does it slow the business down? Because it delays decision making. Too many times I hear the phrase “That sounds like a problem, let’s put it on the agenda for our weekly meeting”. When you do this you’ve immediately slowed the whole project down by a number of days if not weeks. If you’re not at the meeting where thing X is being decided, then maybe you don’t need to be there. Trust your colleagues that the right decisions are being made. Trust that they understand your needs and you don’t need to spend your valuable time “Keeping an eye on things”.
Move fast and break things. Coincidentally, this week I’ve seen two errors from very well known organisations I suspect because they’re moving fast:
But crucially, I don’t think they’re moving too fast. I think it’s good that mistakes are being made because it shows a certain approach, that you are trying to get things to your customers as soon as possible. Note this is a strategic decision because there is a very plausible alternative (spend time and effort making everything perfect – the Apple approach). What’s this got to do with trust and honesty? You trust that when others make mistakes it was with the best of intentions. Trying to get things out trying to help customers is not being careless.
Proactively telling people about your mistakes. It takes some courage at first, but your colleagues will respect you more if you get in front of your mistakes first. Very difficult to do when you first start at a company. but worth the perseverance.
Sharing your work and results. if you can stand up in front of your colleagues, tell them how well you and your team have done, tell them what you did wrong and share all of the data (the good, the bad and the ugly), and be open to where you still need to do better – then you’re showing maturity as a leader, exactly the sort of thing that a good company will be looking for. Again this speeds everything up because you don’t need the endless one to ones. Everyone sees the same info, everybody knows what’s going on so you obviate the need for lots of half hour “catch-up” meetings killing your diary.
These are all great things, but how do they save you time? They save you and everybody else time because you preempt the endless back and forth debating ideas and suggestions. “Perhaps this person knows what she’s doing, perhaps we should give her idea a shot before we shoot it down!”. That way you can go from idea to trial to implementation in a week rather than going round and round trying to decide something when frankly nobody has the answers.
But really the true advantage in terms of speed is the possibility of getting into a flow therefore getting through a higher volume of work. The book “Deep Work” by Cal Newport has been very influential on me. You need 4 to 8 hour blocks without interruption so that you can really get your head into a problem, to do significant and valuable work. I’d go even further and say the most valuable work I’ve ever done is when I’ve had multiple days free to remove myself from the day to day and create something significant. And that’s something we all want to do.
There are always too many things to do running a marketing department. The essence of your job as a marketing leader is making strategic choices about where to double down and where to pause. This is what makes the job both interesting and difficult.
Below I’ve provided a scorecard that I’ve used many times in the past. It comes from the great Gabe Larsen, someone that I’ve followed for years on Linkedin, and I’d urge you to do the same.
This is just a example of his scorecard with some semi random numbers filled in. These are the scores that you give yourself and your department to show where there needs to be improvement and where everything is “just fine”.
An obvious first point – be honest. This is something for internal use only (as I say, the example above is semi-random). If you publish something like this to a wider group then there’s a risk you’re creating a problem for yourself (“I thought you said we were great on social media. It says here we’re hopeless!”). It takes a growth mindset to use this sort of thing really effectively. As always, fix the culture first.
So once you get over the cultural and communication issues, the very next question is “So what? What do I do with this?”. To quote a friend:
Eight out of 10 ideas are bad ideas. So the worst thing you can do is try to do everything. you’ll just burn everybody out. The “OK” thing you can do is to do nothing. At least you’re not burning everybody out. But of course the skill is choosing the right two things to take forward.
What I like about this scorecard approach, is that it gives you the superset of all the things you could do which forces you to consider them. In the example above, the marketing leader has proactively decided to not do any business development work at all. That’s not because she just doesn’t have time or doesn’t know anything about it. It’s a proactive decision that this isn’t something that will move the needle. Crucially, this frees up time to double down on something else. And that’s how you move from being a tactical marketer to a strategic marketer.
Measuring marketing performance is tough. At times I’ve argued that it’s impossible and therefore pointless.
But I don’t think that’s true. As a starting point pretty much any company you work for in a senior marketing role, will expect you to understand the impact of your work (and that of your colleagues). Part of this is looking at flow metrics like “Leads generated this quarter”. But that only tells part of the story, specifically about the flow of business. The other side which is discussed more by the Board than less by marketing departments is the need for balance metrics. As a simple example – you might have one business that’s generating 100 leads per month, but has, already, historically built up a mailing list of 500,000 people (all of whom have given you permission to market). Another company might be generating 100 leads per month but all from scratch and every lead costing money. The former company is more valuable because they have built up a balance of people over time who known them. And that is a very valuable thing.
This makes a significant difference to how you measure marketing performance. Ultimately the goal of the company is likely to be valuation. Part of that valuation is the value of the customer base and/or the awareness of the company. I believe the awareness of a company and its offerings is directly linked to valuation and so, as a marketing department you should be tracking it closely.
Below is a brief PowerPoint covering these different metrics. Any questions, let me know.
Is ChatGPT a threat Google? Obviously there’s nothing I can write about ChatGPT that hasn’t already been written 400 times, so this post is a sort of “naive plagiarism”. Still, with that in mind…
Where do you go first when you want an answer to a problem? This week I wanted to rewrite some old vb code (of which I am thoroughly ashamed) in C#. I didn’t want to do it from scratch so I looked for a tool to do the first draft.
In the past, my process for figuring out how to do this would be as follows:
1️⃣ Do a Google search on something like “How do I convert vb code into C#?”
2️⃣ Look at the top few search results and pick one or two. Often based on brand awareness or the bit of ad copy that they’ve written. In this example, https://converter.telerik.com/ appears at the top and I clicked through to have a look at their offering. Sometimes this will be an organic result sometimes it will be a paid for #ppc ad – and of course this is how Google make a lot of their money
3️⃣ So like most people my first port of call for a question is just to type it in the browser which then uses Google
4️⃣ But I noticed this morning that I’ve started doing something different. I’ve started asking questions in ChatGPT *before* I type it into the browser. Instead of a Google search I typed “Convert the following code into C#” into ChatGPT.
5️⃣ The results were excellent (see below)
6️⃣ The point here is not about how to update code to C#. The point is how we get answers to questions. Google has completely dominated this scene for a long time. But will it continue to do so? For the first time ever, for certain types of questions, I’m starting with ChatGPT then occasionally falling back on Google search.
So the good thing about this as a user is that we’re getting back to a world where great content wins rather than just the winner being the company with the best marketing team. But what if you’re a marketer??
And as soon as someone has a chrome plugin to change the default search to ChatGPT – then this starts to get interesting. In fact I’d be amazed if there wasn’t already something out there. I’m going to start hunting now, and you know where I’m going to start…
Here’s a short presentation on how to calculate return on marketing investment. It’s heavily oversimplified but the principle is simple – do a back of the envelope calculation before you start spending money, particularly if looking at a new system that only promises an uplift (rather than a significant change in performance).
Strategy is about making choices. If you’re not stopping some activities or choosing to not do something then you’re not being strategic.
Easy to say, less easy to implement, particularly when the organisation wants “More leads, more leads, more leads!”. How do you say “I want to do less”?
Almost everything in the world of strategy can be shown in a 2×2! I use the following simple model to help me decide what I need to stop doing and what I need to start doing more. It categorises activities into one of four groups:
Opportunities (top left)
“The Spanish prisoner”
The place where 90% of the discussion takes place. What else can we do? What are we missing? What new things can we do to fill the gap?
There’s a good reason why most time is spent here. Most earlier stage companies are in an aggressive growth phase – they’re not yet trying to cut costs and optimise the organisation. They just need to grow by any means. And this generally gets translated into “What new things can we do?”.
But you only have finite resources. Given that most new activities don’t work, the skill is in choosing the right things and I’d argue that’s the difference between success and failure; between good leadership and poor. It’s better to do nothing at all than everything! At least you’ll still have a team at the end of the endeavour.
Secret Sauce (top right)
“Marcella Hazan’s Tomato Sauce”
These are the things that are working for you right now and making you money. Be very careful before you start reducing effort here, particularly if you’re not sure. For example, it may be that there are a number of review sites that are writing rave reviews about your offering. You may be completely oblivious to this but you’ll know soon enough if the product quality goes down and the reviews start getting worse. And by then it might be too late to stop the decline.
The difficulty is being very honest about why customers buy or don’t. You may wish for example, that it’s because you have the greatest user experience on the market. But that might not be true! It might be that 80% of your sales have come through simple reliability, perhaps for quite a dull product. Not much fun, but if you move resource away from the wrong team It might be hard to get that good reputation back.
The Past (bottom right)
“Knight, death and the devil, Albrecht Dürer“
The market changes. How you got here might not be the best way of moving forward. This can be a very hard pill to swallow, and can lead to some difficult decisions. But as I say, the world changes. In marketing there are activities that were very strong 15 to 20 years ago but do you now rarely see. An obvious example is print advertising. Yes there’s still some going on but it’s definitely not the business it was. I would also put an activity like cold calling into this category – It worked in the days of Glengarry Glen Ross but that was a long time ago!
To stop some of these activities takes teamwork, trust and bravery. It’s far easier to keep spending the money, than tell someone you don’t want to do their favourite activity anymore. But that’s the job I’m afraid.
Sirens (bottom left)
“Tim Buckley”
Sirens are activities that seem oh so tempting, but you need to resist. For example, most marketing departments want to do publicity stunts. They’re fun and the employees will love it! but they can be very expensive indeed. What seemed like a quick brainstorm decision can turn into months and months of graft with little return.
Sadly as a marketing leader, your job is to say no, however tempting the idea. Again, I believe decisions here demarcate great leadership from poor leadership. If a member of your team comes up to you and says “I want to spend the next three years engaging with Gartner, when should I start?”, and you don’t think it’s the right thing to do, then you need to ignore that particular siren.
Finally, it’s worth mentioning again, my examples above are specific to a particular company at at a particular moment in time. These will obviously be different for you! But the process – listing the hundred things that you could do next and pushing them into one of the four quadrants – Is crucial if you want to hit your goals without completely burning out the team.
Scaling up marketing is tough. You’ve missed the fun of the early days where you can try a new thing each week and you’ve not yet reached the safety of a well established brand. The board will be on your back, expecting enormous growth yesterday when you know in your heart that building a brand takes time.
Assuming you don’t have infinite budget, infinite resource and a super-experienced team sitting there waiting for their next job, you need to prioritise. And to prioritise you need a strategy. Part of choosing a strategy is looking at your starting point. What are your unfair advantages? Where are you weak? Trying to start from scratch where you don’t have the expertise or traction in the market can be very tough, particularly if the board want results ASAP.
I use the diagram below as a way of focusing efforts where it’s needed:
There are five strategies listed from top to bottom. The further down the list you go, the less targeted the strategy becomes, though of course you are casting a wider net.
From left to right, the diagram provides a summary of what the strategy actually is, a grid showing where the strategy is most effective (is it better for large or small companies? And is it better for companies who already know what they want?), then most importantly, the cost of the strategy both in terms of budget but also in terms of the number of people you need to implement it effectively.
Crucially, when looking at costs you have to look both at the financial cost but also the number of people you need. To take a simple example, Google ads don’t need a large team to implement and maintain. However they cost a lot, whether you believe they are effective or not. In contrast, high quality content is cheap if not free to create. But you need a team of very talented people for this to be effective.
Finally I’ve put in a column of “how to scale”. Not everything can be scaled by pouring more money in. More money won’t give you higher quality content. More money won’t fix your hiring process. More money won’t fix your culture (in fact it does the opposite). When looking at how to scale it’s crucial to do the strategic diagnosis first of what you are capable of doing as an organisation, where the real problems are and so on. Without this you’ll burn everybody out, but also it will take far too long to hit the targets that you have.
This is where the real skill comes in, in senior marketing roles. How do I make an impact soon? What numbers can I show that actually show progress for a long term objective? Where are the quick wins to keep the board onside? What can I do if I’ve got pots of money? What can I do if I haven’t?
One thing I learned very early on in my career – you can’t do a marketing job without almost constant interaction with the rest of the business. Whether that’s the sales team, the product team, HR or anyone else, you need constant input and feedback from both inside and outside the building.
One very specific case of this – how do you get help with those difficult but crucial decisions that could make a long term impact on the business? one of the issues with being in a role for a long time is that you “Don’t know what you don’t know”. The only way that I’ve ever seen this problem successfully addressed is with external help. And by this I mean “Find somebody who knows more about this problem than I do”. A few days of insight from a guru in your field can save you months off time looking into something where you don’t know where to start. I’ve done this many times in my career, whether it’s “How does B2B marketing work in France!?” through to “What does world class product marketing look like?”
As I say, I’ve worked with lots of these gurus over the years, but i just wanted to pick out three in particular who have made a significant impact on my development and understanding. If you want to make a significant step change in one of these areas I would strongly recommend following these people and subscribing to their feeds – I’ve been very lucky to have worked with all of these people directly.
Gareth runs eqsystems.io and is the most knowledgeable person I know on senior leadership. Most of the problems I’ve ever seen with the impact of work done by teams, comes down to leadership. Why aren’t we getting more leads? Poor leadership. Why isn’t the product fit for purpose? Poor leadership. Why does it take us six months to do something that our competitors do in six weeks? Poor leadership. These problems are almost never caused by employees not knowing what they’re doing. It’s far more likely to be lack of direction and a poor culture.
So if you have a problem with, say, lead generation then before you start looking at your Marketo implementation, look at how you’re leading the team. And if you need help with this then Gareth Marlow is the first person I would talk to. You can’t do everything yourself, you need your team to work In sync with you, aligned to the strategy. This is difficult, very difficult and I’ve found Gareth one of the few people who can help you navigate through the actions that you need to take.
As April puts it herself “Positioning has a positioning problem”. I would agree sadly. It’s a term that is thrown around a lot in marketing departments with very little understanding of what it means. The best way to get started with understanding this problem is to just buy this book!
Additionally, if you can go and see April talk at an event or similar then I would very strongly recommend it.
Another term that is heavily misunderstood – “Strategy”. If you’ve ever sat in a room and heard the phrase “Our strategy is to be the best in the market and to win” – then you need this book. True strategy (i.e. making choices based on diagnosis) is rare and again, an afternoon or two with this book will give you a step change in your understanding. What’s also great about the proper understanding of strategy is that of course it’s not just restricted to marketing. This book has helped me understand how decisions get made, how to make a proper argument for something and why some projects work and others don’t. Again, highly recommended.
When I moved into the CMO role I realised a number of ways in which the position was different to what I had done previously. A lot of these differences are to do with being in a C-level role, but I’ll focus here on the CMO position and the new things you need to learn for that job.
When I first started in this role years ago, it was overwhelming. Suddenly I was responsible for the marketing done by an already successful organisation, and there were 1,000 things I needed to get my head into. Yesterday. I realised then that what I needed was a structure for how to think about this set of responsibilities. After years of trying different models I managed to boil it down to the post-it note above, which is still stuck to my monitor years later. In my view, if you are in a CMO role, you should have a plan for each of these seven things. It might be that your plan is “This isn’t important to us, we’re a different type of business so I’m just going to ignore it”. But that’s still a proactive plan. If somebody asks you “Why aren’t you doing Google ads?” then you should have a well thought through answer, not just “Oh I haven’t really looked at that yet!”.
1. Market research
The core work on which everything below is founded. Market research – who actually wants the thing you sell? Why? What are the problems they are trying to solve? How many potential customers are there? How can you reach them? Who are the competitors? How do they position themselves? How does the market view their positioning vs. yours? And so on and so on. Often a founder knows these things (else you wouldn’t have got to a point where you have a viable company!). But as you mature you need product managers, product marketers and others around the business to understand the market more than anybody else. That last point is important – if there is another company in the market who understands how the market works better than you, then there’s a solid chance that they will beat you. Octopus Deploy is a good example of this. There is literally no company in the world that understands the deployment market better than they do. I believe this is a key reason why they do so well – if you are a customer why would you buy from the second most knowledgeable organisation?
How do you do this in your organisation? Simple, not difficult. You (personally) need to go out there and talk to customers. Yes, surveys are good and there are other ways of getting quantitative data. But you need qualitative data to and that comes from talking to people. If you’re going to write copy for an ad, how can you do that if you haven’t personally spoken to somebody that you’re trying to sell to? It can be difficult finding these people but there are options. Put up a stand at an event. E-mail existing customers offering them £50 for a chat. Run online surveys. Find a friendly industry expert who can give you the lowdown. But as a senior marketer, you or one of your colleagues needs to be doing this task. Oh and buy the book Obviously Awesome by April Dunford!
2. Content
No one likes adverts. Digital ads do work as part of the marketing flywheel, you need great content. What is great content? Simple – it’s content that people actually want to read. 98% of the content out there is bland, copycat, written by people who don’t understand the market and can be generated by ChatGPT. Your content needs to be remarkable in both senses of the word.
This is difficult. Finding people in your industry who can produce great content should be one of your highest priorities, but it will take up a lot of your time. You only need one or two but it’s much better to have one great writer than three average ones. And in my experience, it’s better to have somebody who’s an expert in your field and not the greatest writer in the world rather than vice versa. Most customers don’t mind about your grammatical errors.
But the thing that is so great about having good content is that it’s investment that will last you for a long time. A well written article about your domain can work for you for years to come. And then you’ll start to get people sharing links, commenting on posts and reposting articles. “Marketing” from someone that doesn’t work for you is always far more powerful because it is genuine and authentic.
But again, what can I do other than just writing great content? Well you can place that content on 3rd party sites. You can sponsor articles. You can encourage others to write about you. You can reuse content – if you’ve written a great white paper, cut it up in 20 different ways and use all of your snippets – a short advert, a podcast, a written article, a presentation. if you can get both things right (the creation and the distribution of information), then you’re maximising your use of content marketing as an activity.
3. Sales enablement
There are lots of different parts to sales enablement, particularly if you have a complicated sale that needs support. I’ll just focus here on the part that marketing can play.
But whose job is it to do sales enablement? Sales or marketing? I’ll dodge that question by saying that “You need some group in your company that is worried about sales enablement”. That may sit in sales or it may sit in marketing, or even somewhere else. But you can’t leave this crucial role to salespeople themselves. Without a sales enablement function supporting sales team you’re sending troops into battle unarmed.
There is an exception – where you don’t really have Sales at all! If you’re an online retailer and customers never really talked you during the sales process then sales enablement is less relevant for you. But otherwise, this is a key part of the senior marketing role and it is your responsibility to make sure it happens.
So your job in marketing is to set up the opportunities for the sales team to smash home (and generally they are Opportunities in Salesforce that you want to get to Closed Won). Sometimes of course a salesperson will generate an opportunity themselves, nurture it through and close it. But you’re not doing your job if you’re not enabling them with support, collateral, training, insights, and thirty other things which make their life easier. You may not get the glory when the points are scored, but without that teamwork between sales and marketing, with sales enablement as the bridge, the opportunities won’t close regardless of how good the original lead was.
4. Awareness
As per the earlier image, if you don’t have a large number of people who already know about you then you need to start earlier in the marketing life cycle. Nobody ever went from “Never heard of you” to “Here’s £25,000” in a week. They went through a whole series the complex stages ending with the sale. And this starts with basic awareness. Have they even heard your name? Is there any association between your name and the value you offer? A large part of your job as a marketing person is building the brand awareness, so that when the customer is looking at solutions there’s already a name there that they can trust (at least in part).
Building awareness is expensive, and in almost all cases, impossible to track. Yes you can do brand surveys and ask customers “Where did you first hear about us?”. And this is something you should do, perhaps a little later on. But either way, if you’re starting from scratch then it’s a long road ahead. And yes, there’s no way to go from zero to hero without a very large pot of money/investment.
How do you do this? Depends enormously on the industry. Paid-for advertising is an obvious answer but there are others. If your budgets are low then some ingenuity and creativity is needed. You need a crystal clear value proposition (why should the customer care?), good designers, a “Keep it simple” approach, but hopefully this should be one of the easier things to get going. These are some of the things I’d be thinking about:
5. Demand Gen
This is a tough stage to summarise. Theoretically everything you’re doing above is generating demand. I.e. the demand comes from the accumulation of all the other activities you’re doing.
But, in a senior role, you must know how to measure this even if you don’t know exactly what causes the generation of demand. To measure the performance of demand Gen, I would strongly suggest using “Value of opportunities created” as your KPI. i.e. Take the dollar value of opportunities created in a given month from Salesforce and use this to measure demand Gen performance. There are a few alternatives, but here’s why I don’t think they work:
Number of leads. Anyone can generate leads. You can fool Google it’s giving you lots of leads – but they won’t convert. The goal here is money and without knowing whether the leads are any good this is a meaningless metric.
Total revenue. Of course, this is what we’re all ultimately interested in. But there are so many contributions to this final figure (every department contributes to revenue), that it’s really difficult to pull apart where marketing has had an influence.
Awareness or some other early stage metric. You should know what these numbers are, but they’re just too far away from revenue. And it can undermine your credibility with the leadership team if you’re reporting on “Page visits” in your marketing performance meetings.
So you must show the top level opportunity value generation numbers. But can you show further detail? I’d suggest the following as a starting point:
Split the figures out by region
Split the figures out by sales team (if different to above)
Show the numbers for revenue from existing customers vs. new customers
Show the numbers by different industries or segments
And a final tip – start simple then build up. Start with the super-reliable opportunity value metric, then when you and the rest of the leadership team believe that number, move on to the next one. If you don’t do this then you’ll spend management meetings explaining 20 charts but without actually saying anything useful for the rest of the team.
6. Partnerships
Crucial for reaching customers you simply can’t reach on your own. This could be for a number of reasons, but often it’s a matter of size. If you’re a company of, say, 100 people and you’re trying to sell to Lloyds Bank then it’s very unlikely they’ll even look at you. So your job is to try and find a bigger organisation willing to partner with you. NB: there are also obviously times when you can partner with smaller companies, but that’s not the big win here. Of course the mountain to climb here is convincing then to work with you at all. The prize is worth it, but it can be a long slog (often years). And generally, “Go big or go home” – a single partnership with Apple will provide more value 10 partnerships with smaller firms.
The two main types of partnership where I’ve seen most success are:
Technical partnerships. Someone else provides a bit of technology that you don’t have, something that might take you years to develop and you need to get to market quickly.
Commercial partnerships. Access to customers that you can’t reach on your own. A simple example of this is a new country – if you’re trying to open up Spain as a market, for example, and you know nothing about Spain then you need a partner or you’re going to spend a lot of time and money going in the wrong direction.
Because building partnerships can take a long time and be quite arduous, it is often better to hire a separate “Partner manager” of some sort, not only to get their expertise but because otherwise it can be an enormous distraction from the core business.
7. Customer Retention
I thought I’d end on something slightly controversial. The business that you work for should have a deep understanding and strategy for how customers are retained. You should know why people stick with you.
What’s controversial is what you can actually do about it in a marketing department. I believe most customers stick with the vendor because the product is great. Once you’ve bought from someone do you ever see their marketing again? You were busy using the product and getting on with your own projects and activities. A lot of money is spent on customer success and customer retention teams, I’m just not sure that’s money well spent. The product retains people. What that means, is that if you as a company have a problem with retention I would invest budget in the product rather than marketing. That’s a difficult argument to make if you work in the marketing department (“What do you mean you can’t help!?”). But that’s the essence of good strategy – making difficult calls where resource is constrained.
So I won’t write more about marketing customer retention as I’m not sure it’s something you should spend a lot of time on. But as always, very happy to be proven wrong!
These are the seven categories that I’ve used to try and understand all of the work that is going on in the marketing departments that I’ve been fortunate enough to run. as I say, in a senior role you should know why you are or are not doing something in each of these categories, and what value this work adds.
Finding a website with a chatbot on it (for example, https://chat.openai.com/), and having a go yourself, if only to see what everybody is talking about
Adding a generic chatbot to your own website. I’m not quite sure why you would do this, but it’s part of the process understanding how to integrate ChatGPT into your website
(now it starts to get more interesting…) automatically creating an FAQ for your website based on your content
Creating a ChatGPT bot that can go on your site for your customers to use to find out more about you and your company
I’ll talk about the first four points here and then, in the next article, the last point. This is a considerably bigger task, so needs a post of its own. The end goal is to allow customers and potential customers to come to your site and ask questions about your offering. There are two advantages to this approach:
If you’re resource constrained, you don’t have the people to be on the phones answering questions all the time.
Consistency. You can manage and see what’s being said to your customers on the website.
But where should I start?
Reading about it on the Internet
Not a whole lot more I can add here. If anything it’s hard to escape articles about the topic. The BBC has some good articles.
Playing with a chatbot yourself
The first question most people have is “But are these things any good? They’ll never fool me!”. Don’t listen to others, try it out yourself. I’d suggest that the openAI website itself is a great starting point. You may need to create credentials first, but spending some time here will really show you the power of what everybody is talking about. Here’s a pretty random example. I asked “What is account based marketing?”:
That’s very good. Yes, it’s a little generic, but I’ve done that with no effort, no research. If you wanted to find out about a new topic at work, 30 minutes with chatGPT would get you on your way.
Adding a generic chatbot to your site
Really this is a preparatory step before going on to the next more interesting stage. But it does introduce some of the useful resources.
I use WordPress for my site, set my example here is for WP. But the principle is the same – the difficult bit is creating the training data and then training up a model. If you can do that then getting it on WordPress is easy.
In particular, I want to highlight the Jordy Meow plugin. This is an incredible bit of kit, I was repeatedly pleasantly surprised by what was available and how easy it was to install and get working. This is no mean feet given that we’re moving into the territory have training AI models.
Like any WordPress plugin you install it from your dashboard. Then, on your WordPress site you’ll have something that looks a bit like the following:
Creating an FAQ for your website
So far we’ve looked at generic chatbots which are all over the web. But you want something for your website, based on your industry.
Again, I’m not going to go through the details of doing this because there are some fantastic notes on the AI engine help pages, and it will be different for different sites. But the most important point, the place where you need to spend most time and the place where you can really differentiate is on the training content. This will sound familiar to anybody who’s worked in marketing, but if you’re creating something to help you generate interesting content, then you have to have some interesting content to start with. I’ve used the process on this site to create 100+ questions and answers, without having to write a single question myself. The engine is so powerful that you can just give it a block of well written marketing text and it will automatically create some questions and answers from that text. To create the FAQs on this site I simply fed the engine the 94 blog posts I’ve written over the last 10 years and asked it to give me some questions based on this input.
You can see some of the results on this page. Remember all of these were auto generated, including the actual questions:
I’ve been enormously impressed by AI engine and the work done by Meow apps.
All great, but isn’t this a marketing blog? This just feels like a lot of technical detail! Well yes, that’s true. But one of the ways you can differentiate yourself from the crowd as a marketer is by moving on from just talking about technology to actually showcasing it. You can significantly boost your career by properly understanding how AI technologies can impact marketing. This needs to be more than just “Add AI to your marketing efforts!”. Your claims need substance and this is where the hard work comes in. I had the advantage that I’ve been writing blog posts for 10 years or more, so I had the source material. But you have to start somewhere, and this is one way to take the content that you’re writing and getting out to more people in a more palatable form.
Any further questions please feel free to get in touch to discuss how I can help.
Of course, really, we all want to build Skynet. However, until Judgement Day comes, we’ll have to make do with ChatGPT. ChatGPT is obviously a Big Deal right now for marketers, so I wanted to find out for myself.
Firstly, as a general point I do think it’s important to try technology out yourself before moving forward with a project. There are plenty of ways of trying out bits of tech if only for your own understanding. ChatGPT is no exception – you can try it online with almost zero effort:
This should start you off on your journey into the world of ChatGPT. For example, here are the results I got when I asked “What is the marketing flywheel model?”:
Definitely not wrong. And if I had to write a short piece on this topic I could do worse than copy and paste this into a blog post.
If your role is something like “content creator” then you can definitely get away with getting a machine to do your work instead. So, what’s the problem?
The issue can be seen in the response above. Though this answer is “not wrong” that is a long way from it being an insightful and useful piece of content. Customers coming to your site want insight, new ideas, new perspectives. They want to hear from industry experts, else why read your articles at all? If your article is just an aggregation of the content on the Internet, how are you differentiating yourself from everybody else out there?
The response above is marketing 101. Perhaps fine for a GCSE paper, but not good enough if you want to attract real customers (our job here). With follow up questions, I could definitely get more out of ChatGPT, but here are some of the things that are missing:
It’s very generic. How would this be different for your industry?
It’s not “Of the moment”. What’s new in the industry? What’s happened in the last month or two?
It doesn’t help with prioritisation. What needs to be done first? For a mature org vs. a startup?
How is the flywheel model different to other models? What is it similar to? Where shouldn’t you use it? Does it work the same in B2C as it does in B2B?
What’s the underlying strategy for this model? If somebody asked you “Why does it look like this?”, could you answer?
How is this different from the funnel model?
And so on. So again, for certain marketing tasks it is great. If I had an afternoon to write an FAQ about content marketing, this is where I’d start. But most of us aren’t under those time pressures – you should be writing quality over quantity. Have talked to some customers. Find out what their pain really is. Ask them why they bought from your competitor instead. Talk to industry experts. What you write from your own expertise will always be better from what ChatGPT comes up with.
New Year, new marketing plans. Hopefully by now you’ve kicked off various activities and you’re waiting to see how those early campaigns are working out.
The other thing I see in marketing departments at this time though is burnout. Everyone is trying to do everything either because there’s no real strategy there (“let’s throw everything at the wall and see what sticks”). Or it could just be bad planning (“the start date for every campaign is the 1st of January”).
Either way, you might soon be revisiting the strategy discussion. Specifically, why are we doing activity A? Can we kill activity B? Is activity A working yet? That activity can quickly turn into navel gazing, when what you need is focus and a way of choosing what you should be really worrying about. To that end, I’ve been using the flywheel model below for years now. The point of the model is that you have a list of metrics and activities you can look at to check whether you would actually doing them and doing well. As a simple example: if nobody is coming to a website to talk, what should you do? Should you hire a content writer? A designer for the website? A product marketer? The diagram and notes below give what I think the “next best” activities, based on splitting the marketing flywheel into five stages.
The marketing flywheel for senior decision-makers (SDMs)
This first diagram is for senior decision-makers (SDMs). There are no hard and fast rules here, but generally these are people who are less likely to be actually using the product themselves but certainly influence the buying decision heavily.
For each part of the flywheel, I’ve put what I think the most impactful activities. If I only have time to do one thing, what is it? Looking at the first diagram, if you’re brand-new into a market (nobody knows about you) and you’re trying to sell to senior people, where should you spend your money? You must create awareness of the brand first. I nothing else will work without this first. So your first activities have to be things like PR, analyst relations, thought leadership, at some budget for LinkedIn. If you were spending money on complex lead qualification processes, when you have no leads to qualify!, then you’re burning money.
The marketing flywheel for end users
This second diagram spend users. Meaning you are advertising to the people who are actually going to be using the products. That means they’re likely to be more junior and have very different requirements (for example they’re likely to care about usability and less likely to care about long-term financial benefits in the organisation).
Here, the marketing is different. End users don’t read the same things as senior decision-makers. They’re far more likely to do a Google search for a particular problem they’ve just hit a than for an in-depth analyst report.
What does this mean for marketing budget? If there is a community of users, then you need to reach out to them. If not, PPC and SEO crucial. Either way as you get further through the flywheel the product has to be amazing (for end users, there’s nothing you can do in marketing that will overcome an unusable product).
Hopefully this is useful as a way of making sure you’re making a big impact to the start of the year without burning everybody out and without burning through your whole budget by Valentine’s Day. If your plan is to “do everything” then that’s not strategy, that’s a recipe for employee burnout and empty pockets.
A lot has already been written about how the old marketing funnel model is no longer as relevant in modern B2B organisations as it used to be, and how a flywheel model is more appropriate for how customers really buy (as an old colleague said to me “The person who invented the marketing funnel should be shot!”. I think that’s excessive, but it makes the point…). Here’s how HubSpot describe the alternative approach: https://www.hubspot.com/flywheel.
But rather than just repeat the work of others, I wanted to write an article on a model that I’ve been using for a few years now and which I found very helpful when making practical decisions about marketing activities. Models are all well and good but the reality is we have to make calls every day on where to spend our marketing budget and for this we need a robust model to help guide us in the right direction. Should we spend more on PPC ads? On brand building? On events? What? That’s more importantly, why? You will be asked why you decided to spend all of the budget on activity A rather than B. You need to have answers to those questions.
Here is the short version of my PowerPoint that goes through this model (there is an 80 slide version of this that I’m working on as well). To summarise the context for this:
It’s focused on B2B marketing; a small to medium sized businesses; in the tech space; in the English speaking world. It certainly won’t be relevant to everyone – take the best, leave the rest.
It’s really only about marketing. For this to really work it has to be part of an overall marketing/sales/product plan.
It assumes you already have a little traction in the market. If you’re starting from zero brand awareness, that’s a different job.
There isn’t really anything here about the order in which to do things. I.e. what’s the strategy to do the right things first to make the biggest impact earliest? That requires a diagnosis of your company’s current performance and challenges, which is different for every organisation. But that’s the most interesting and difficult part of the job. If you try to do everything in this deck all at once you will fail – the trick is getting the ordering correct.
The plan relies on and assumes a relentless focus on the customers’ real problems. This is a cliche for a reason – the best marketing plan in the world will fail if you don’t have a deep understanding of the customers’ problems.
It uses the “Pincer” GTM strategy, where you are selling to both the senior decision maker and the end user at the same time. If you only need to reach one of those people, your job is much easier 🙂
But to repeat, having some sort of model for how marketing works is crucial. Otherwise how can you make decisions about which levers to pull? About what effect you think you’re having? Of course sales is the end goal here but you need to understand what happens before the customer makes the purchase if you want to influence what happens at the end. And in a B2B process that can take months if not years from initial awareness to signing on the dotted line. Understanding the drivers between these two stages is the key to knowing how are you can influence the process and increasing the chances of hitting the revenue goals.
A final note – all of this is, of course, based on the work of many others and I’ve blatantly taken diagrams and paragraphs from other marketing practitioners. Almost all of the ideas here from other people, all I’ve tried to do is bring it together and find the links and similarities. Any errors are mine.
