A practical guide to making your organization easier for people and AI to find (and trust)
Our teams have spent a lot of time lately testing how brands appear in AI answers, reviewing the sources behind those answers, and helping organizations understand what AI seems to know about them. The thing that continues to surprise me each time is that most companies do not have an AI visibility problem because they lack content, but it is often because they are not sharing their most useful content publicly.
At its simplest, GEO, or generative engine optimization, is the work of improving whether and how an organization appears in AI-generated answers. The best GEO strategy begins by understanding what your organization knows, what its buyers genuinely need, and where you have something credible to add. On that note, GEO becomes less a technical exercise and more one that is focused on making your value clearer for both people and AI systems alike.
AI is changing more than where brands appear
Search engines have helped us find digital sources for decades, but with AI search, we can now pull ideas from several places, compare options, and get recommendations without ever entering a website. We are already seeing that shift in user behavior. Pew Research Center found that people clicked a traditional search result in 8% of visits when a Google AI summary appeared, compared with 15% when one did not. That means the first impression of a company may now be formed inside an AI answer.
However, one of the easiest misinterpretations is that you can simply “fix your AI ranking.” GEO does not work that neatly, even though many of the same foundations as SEO still apply. A brand may appear when someone asks for well-known companies in a category, then disappear when the question becomes more specific. The answer can change with the wording of the question, the buyer’s need, the AI platform, and the sources available to support it. Even the same prompt can produce different answers from one run to the next.
Your website is just one signal, not the whole story
A company’s website absolutely still matters. It is often the clearest place to explain what the organization does, who it serves, and why its work is different. Strong structure also helps both search engines and AI systems interpret that information.
But the website is only one part of the public record. AI systems can draw from news coverage, leadership posts, podcasts, partner websites, and other third-party sources. Together, those sources shape how a company is understood and what it is known for.
That wider footprint appears to matter. In an analysis of 75,000 brands, Ahrefs found that branded mentions across the web had the strongest observed correlation with visibility in Google AI Overviews. Correlation is not causation, but the finding supports a practical point: what others say about a brand matters alongside what the brand says about itself.
What your organization knows in the AI era
What your organization knows goes beyond the expertise of any one person or team. It includes what the organization is genuinely qualified to be known for and notably where it has a meaningful perspective or advantage.
Thinking this way leads to a more useful question than, “What can we post about?” It asks what your organization can credibly contribute to the conversation you are trying to be a part of.
Answering that question requires separating three things that are often treated as one:
- What you know: The knowledge your organization has gained through real experience and can support with evidence.
- How you share it: Whether that knowledge is clear, accessible, and useful to others.
- What others can confirm: What customers, partners, and others can support based on their own experience with your organization.
Many organizations possess more useful knowledge than their public presence suggests. It may live in sales conversations or leadership discussions rather than anywhere buyers can find it. GEO reveals that gap, so it’s important to identify what deserves to be shared and express it in a form buyers can use.
The strongest GEO content will often begin by asking questions like these:
- What are we genuinely qualified to be known for?
- What do customers value or repeatedly misunderstand?
- Which questions require our team’s real experience to answer?
- What can we substantiate with evidence?
- Where do we have a meaningful perspective or advantage?
- Which buyer situations matter enough for us to invest in?
This changes content from an output exercise into a knowledge-sharing exercise where you are actively sharing, not just promoting.
Publish to improve understanding
That said, good GEO content should help someone understand a decision, even if that person never becomes a customer. That means answering the real question, supporting important claims with proof, and being honest about limitations, tradeoffs, and fit.
This is where responsibility belongs in the GEO conversation. To me, it is not a separate ethical layer placed on top of the strategy. It is a quality standard for the strategy itself.
A company can repeat the same claim across dozens of pages, place it in friendly publications, and make it easier for AI to retrieve. But repetition does not turn a weak claim into useful knowledge. Princeton University’s GEO study reached a similar conclusion: reliable citations, relevant statistics, and credible quotations improved visibility in its tests, while keyword stuffing did not. Most importantly, none of these tactics creates market trust if the underlying claim cannot be supported.
The better question is not only, “How do we appear in this answer?” It is, “What can we add that would make the answer better?” That creates the conditions for a better GEO strategy and a more trustworthy public record.
A practical starting point for GEO
GEO can quickly become a long list of tactics, but here is where I’d start if you’re curious about establishing a baseline for your organization:
- Start with the position. Decide what the organization should credibly be known for and which buyer questions matter most.
- See what AI understands today. Test a range of real questions across AI tools. Look for patterns in visibility, accuracy, comparisons, and recommendations.
- Find where that understanding comes from. Review the sources and signals shaping those answers. Separate what the company says from what others can verify.
- Look for knowledge gaps. Identify useful expertise that is missing, unclear, unsupported, or trapped inside the organization.
- Improve the public record. Publish better explanations, strengthen proof, connect related pages, and make the information easy to interpret. Then test again.
Technical work still matters too. Clear headings, useful page structure, internal links, fresh information, and accessible content can all make knowledge easier to find and interpret. For ChatGPT Search specifically, OpenAI says a site must allow OAI-SearchBot if it wants to appear as a source. Schema can still support broader search visibility, but Google says it is not a special requirement for generative AI search. These tools cannot decide what the organization should stand for or create expertise that is not there. Clarity has to come first.
You still have time
We have a line we share with everyone when speaking about GEO: “none of this should create internal panic.” At the time of writing, GEO is still relatively new and changing quickly as models and search experiences evolve. Use this time to understand your own perspective, then share what you believe others will find genuinely interesting and helpful.
If you’re curious about learning more about GEO, contact us here or attend one of our upcoming GEO webinars.
— Strahinja Spasojević