For a long time, digital marketing gave us a fairly comfortable model. Someone searched for something, Google gave them a page of results, and our job was to work out how to get as close to the top as possible.
Watch the whole recording here.
AI has made that picture a lot messier, and we think that is a good thing.
The question we hear constantly now is, “How do we optimise for ChatGPT?” or “How do we get our business mentioned in AI?” They are reasonable questions, but we think there is a better one.
Where is AI getting its information from in the first place?
Before an AI platform can recommend your business, compare your product with a competitor or tell someone which company they should consider, it needs evidence. It goes looking across the web and decides what information appears credible enough to use.
So rather than obsessing over the final AI answer, we need to start paying attention to everything that happens before it.
One question we like businesses to consider is this: if ChatGPT disappeared tomorrow, would there still be enough independent evidence across the internet for someone to conclude that your business is one of the best in your space?
That changes the conversation completely.
It is no longer just about what your website says about you. Of course your website says you are good. It should. The stronger signal is when customers, experts, industry publications, communities, review sites and other independent sources are saying similar things.
That is where AI starts to see consensus.
A simple habit can tell you a lot about where you currently stand. The next time you ask an AI platform an important question about your industry, don’t just read the answer. Open the sources behind it.
Do that repeatedly across the questions your customers are actually asking and patterns start appearing. You might notice the same Reddit communities coming up. Perhaps it is YouTube, LinkedIn, comparison articles, review platforms or particular industry publications.
That is your real search landscape.
The important point is that there is no single platform every business suddenly needs to dominate. Reddit might matter enormously for one question and barely feature for another.
Think about someone researching a bull bar. If they ask how to install one, YouTube is incredibly useful because they want to see how it is done. If they ask whether a particular bull bar is actually any good, a four-wheel-drive forum or Reddit discussion becomes more valuable because they want experiences from people who have bought one. If they ask who the major manufacturers are, industry publications may carry more weight.
The question determines the authority.
We demonstrated this by searching for the best power rack for a home gym. Google surfaced a mixture of products, community discussions and comparison content. When we put the same question into an AI platform, comparison articles became particularly prominent.
Then we looked at alternatives to HubSpot. Again, comparison articles appeared repeatedly, alongside community discussions in some search experiences.
That tells you something useful. If you want to be considered in that conversation, perhaps you need a genuinely useful comparison page. Perhaps you need to be included in an existing third-party comparison. Perhaps there is an active community conversation where customers are already discussing you and your competitors.
The opportunity is in understanding the environment before deciding what content to create.
Community conversations are particularly interesting because they contain something brands struggle to manufacture: honest disagreement.
People say things like, “The product is great, but the customer support isn’t,” or, “I tried both and this one worked better for my situation.”
That information is marketing gold, even before we think about AI.
It can tell you what customers value, where competitors are strong, what objections keep appearing, what people misunderstand about your product and where there might be gaps in the market. A niche forum with relatively little traffic can sometimes contain more useful commercial intelligence than a huge generic website because the people participating have genuine experience.
The mistake would be jumping into those communities and trying to manufacture praise.
The goal should be to earn the mention, not fake it.
Your happiest customers already have something incredibly valuable: genuine experience. Make it easier for them to share that experience. If there is a relevant conversation happening, you can point them towards it and invite them to contribute their own view. Don’t script it. Don’t tell them what to say. Give them the opportunity to participate.
The same principle applies to reviews and testimonials. People are much more likely to contribute when you remove the friction.
While all of this is happening outside your website, your owned content becomes more important too, not less.
Your website needs to contain information people and machines can actually use. Clear product and service information matters, but so do case studies, technical documentation, fitting guides, FAQs, comparisons and original research.
We particularly like information nobody else can provide.
Instead of saying you are a leading provider, prove what you have done. Instead of making another generic claim about quality, provide evidence. Create useful data, research, calculators, guides or insights that a journalist, customer, industry expert or community member might genuinely want to reference.
That creates what we think of as informational gravity.
Your website establishes your expertise, while the rest of the web provides independent evidence that supports it. AI then repeatedly encounters your business in different contexts rather than relying on one perfectly optimised page telling it how wonderful you are.
There is another shift coming that makes this even more important.
Search is moving from “tell me” towards “do it for me”.
Instead of searching through dozens of pages to compare four CRM platforms, someone can increasingly ask an AI agent to compare them against specific requirements, shortlist two and potentially take the next action.
In that world, clever keywords alone won’t be enough. The AI needs to find evidence. Does this company actually have experience with Shopify? Are there strong reviews? Are there relevant case studies? Is the information consistent across the web? Can the website be easily read and understood by machines?
That is why we don’t see this as replacing SEO, Google Ads, PR, content or community building. It connects them.
A practical place to begin is with the commercially important questions your customers are already asking. Take 15 or 20 of them and search across Google and the major AI platforms. Look at what sources keep appearing, where your competitors are mentioned, what people are saying about them and where your own brand is absent or misunderstood.
Then start closing those gaps.
And don’t assume the answer today will remain the answer forever. Different AI platforms use different sources, and those weightings continue to change. That fragmentation creates an opportunity for businesses willing to pay attention now.
Stop trying to manipulate the final AI answer. Build the evidence that deserves to shape it.
If this has you wondering what AI is currently finding about your business, start with those customer questions and follow the sources. You might be surprised by what you find, and if you want to unpack it further, we’re always happy to continue the conversation over a coffee.