Our method
Audit. Entity. Evidence. Measure.
There is no single AEO trick. The durable work looks less like gaming an algorithm and more like reducing ambiguity—then proving what you claim. This is the sequence we run on our own products, and the one we run on yours.
- 01
Audit
Find out what the engines say today.
We build the question set your buyers actually ask—category questions, comparison questions, problem questions—and run it across ChatGPT, Claude, Gemini, Perplexity, and Copilot. The output is a baseline: where you are named, where a competitor is named instead, which sources get cited, and where the description of your product is wrong.
- Question set and baseline scores
- Citation source map
- Accuracy and misstatement log
- 02
Entity
Make the company unmistakable.
A model connects entities, categories, attributes, and evidence. If your category is vague or your facts change from page to page, there are fewer reliable paths to your product. We settle the core language, then make it consistent across your site, structured data, profiles, directories, and founder bios.
- Category and positioning statement
- Structured data implementation
- Off-site profile alignment
- 03
Evidence
Publish what a model can quote.
Inclusion follows verifiability. We identify the claims you need to substantiate and build the assets that substantiate them: original data, precise definitions, comparison content, documented outcomes, and the independent corroboration that keeps your own site from being the only source.
- Definitional and reference pages
- Original research or benchmarks
- Third-party corroboration plan
- 04
Measure
Track the answer, not the ranking.
The baseline gets re-run on a fixed cadence. We report what moved, what did not, and what we changed our mind about—so the programme compounds into a clearer model of your market instead of a list of completed tasks.
- Recurring visibility report
- Decision record
- Next-cycle priorities
What we report
Five numbers that replace the ranking.
In an answer engine, your company can influence a decision without ever receiving a click. These are the measures that survive that shift.
Mention rate
How often the brand appears at all for a relevant question.
Citation share
Which sources the answer links to, and how often those are yours.
Answer accuracy
Whether the description of your product, pricing, and audience is correct.
Competitive displacement
Which competitors are recommended in your place, and on what basis.
Evidence gaps
The questions where no source—including yours—gives the model enough to work with.
Questions
Straight answers.
Including the ones where the honest answer is "no". If a vendor will not give you these, that is information too.
What is answer engine optimization?
Answer engine optimization (AEO) is the practice of making a company legible and credible enough that AI answer engines include it when generating a response. Where SEO optimises for a position in a list of links, AEO optimises for inclusion in a synthesised answer: clear entity definition, verifiable evidence, and consistent facts across the sources a model retrieves.
How is AEO different from SEO?
SEO asks whether a search engine can find, understand, and rank a page. AEO asks whether a source is clear, credible, and specific enough to shape the answer itself. The two overlap on technical accessibility and content quality, but they differ in the unit of value: SEO is measured in rankings and clicks, AEO in mentions, citations, and the accuracy of what is said about you—outcomes that often happen without a click at all.
How do you measure AI visibility?
By running a fixed set of buyer questions across multiple answer engines on a repeating cadence and scoring five things: mention rate, citation share, answer accuracy, competitive displacement, and evidence gaps. Because model outputs vary between runs, each question is sampled several times and reported as a rate rather than a single result.
Can you guarantee my brand will appear in ChatGPT?
No, and you should be sceptical of anyone who does. Answer engines are non-deterministic, they change without notice, and none of them expose a ranking dial. What can be controlled is the input: whether your entity is unambiguous, whether your claims are verifiable, and whether independent sources corroborate them. We commit to measurable movement on those inputs and to honest reporting on the outputs.
Does AEO mean I should stop doing SEO?
No. Answer engines still crawl, still parse structure, and still rely heavily on the open web as a retrieval corpus. Technical accessibility and useful content remain prerequisites. AEO adds a layer on top of that foundation rather than replacing it.
How long does an AEO programme take to show results?
Entity and structured-data corrections can register in weeks. Evidence and third-party corroboration compound over months, because they depend on material being published, indexed, and picked up by sources outside your control. We set the baseline in the first cycle specifically so that progress is legible rather than anecdotal.
Who is this for?
Founder-led B2B SaaS companies, typically seed through Series B, that have a real product and a real point of view but are not showing up when their category is discussed. It is a poor fit for companies without a defensible difference—no amount of structural work makes an undifferentiated product quotable.
Norbu Labs builds its own products. Is consultancy a side business?
Both tracks are real, and they are deliberately connected. We build and operate our own SaaS, which is where this method was developed and where it gets tested first. We then take on a small number of engagements each quarter—capped on purpose, because the studio work is not a placeholder between clients. If we are at capacity we will say so rather than stretch.
Phase one
Start with the audit.
Send us the five questions a good-fit buyer would ask before they know you exist. We'll run them and show you what comes back.
Start a conversation