How it works
Identify, diagnose, direct.
Three steps, run in that order, ending in a ranked plan and a route to getting it done.
The questions your customers actually ask.
We build the question set from how people really phrase things, not from keyword exports. Which brand is best for a given sport or game. Alternatives to a named competitor. Which vendor an operator should shortlist. What is worth using in a particular market.
Each question is run repeatedly across ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews, and separately in every market you care about, because the same question returns different names in different places.
For every answer we record who is named, in what order, how often, and which sources the answer was built from. Running a question once tells you almost nothing. Repetition is what separates a stable position from a lucky one.
Every gap is a measured value.
Knowing you are absent is not useful on its own. The diagnosis works out why: which competitors hold the question, which third-party sources the models keep returning to, whether you appear in any of them, and what on your own site stops a model describing you with confidence.
Your brand is benchmarked against the real bar for your category and market, so a finding is never a matter of taste. Where you sit, where the segment sits, and how far apart those numbers are.
Opportunities are treated as first-class findings. Where answers vary between runs and cite no consistent source, no brand has claimed the question yet.
| Signal | Yours | Segment | Verdict |
|---|---|---|---|
| Structured data present | None | 78% of peers | Missing |
| Answer-shaped content | 556 wds | median 961 | Thin |
| llms.txt guidance file | Missing | ~50% publish | Below |
| Trade-press citations | 31 | median 1,263 | Far below |
| Page performance | 89 | p90 = 74 | Above |
Illustration. Brands shown are fictional.
A plan, in the order you should do it.
Findings are sequenced by what will move your position soonest for the least effort, each carrying the reasoning that produced it: the questions affected, the engines it showed up on, and the work involved.
That ordering is the part that makes it usable. Every audit produces a list. Very few tell you which three things to do first and why those three.
Findings come in two shapes, and the difference matters. A named competitor already holds the question, which means taking a seat off someone. Or no brand holds it at all, which is an open slot and usually the cheaper win.
From there it is a question of getting it done, which is the next section.
A competitor is cited ahead of you on payout questions
Kingfisher is named in 11 of the 14 answers about withdrawal times in your markets, drawing on three review sources that do not mention Northgate at all.
No brand owns this question yet
Answers vary between runs and cite no consistent source. Weak incumbents make this the fastest ground available to you.
Your category pages are unreadable to a model
Nothing on the page states plainly what you offer or where, so the model has nothing quotable to lift.
Illustration. Brands shown are fictional.
Identify what the answers actually say
We run the questions your customers and buyers really ask, at scale, across ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews, in the markets you care about. We record who gets named, in what order, how often, and which sources each answer was built from.
Diagnose the problems and the opportunities
Then we work out why. Which questions you are absent from. Which competitors own them, and what they have that you do not. Which third-party sources the models keep leaning on, and whether you appear in them at all. Questions with weak incumbents get flagged as opportunities, because that is usually the fastest ground to take.
Direct the work, in priority order
You get a prioritised plan. Specific changes, sequenced by what will move your position soonest for the least effort, each with the reasoning attached. Where your platform supports direct updates we apply them for you, with your approval. Where it does not, your team runs the same actions through our MCP.
From the finding to the change itself.
The analysis is only worth what gets done with it, so the plan is built to be executed rather than admired. If you are weighing this against a visibility score, the two are set side by side here.
- The analysis needs no access at all. Visibility checks, benchmarks, gaps and the ranked actions all run read-only, from what is publicly visible. Connecting a platform is a separate decision you make later, and only if you want the changes applied rather than handed over.
- Nothing is written silently. Every change is previewed as a literal diff and waits for your approval before it goes anywhere. New pages arrive as drafts. A change that spans several edits is bundled, so it applies as one piece or not at all.
- Anything applied can be undone. Every change is logged with who made it, a before and after snapshot, and the check that ran once it was live. Access is granted by you and can be withdrawn at any point.
- Bespoke platforms connect through our MCP. Most iGaming brands run proprietary stacks, so the same analysis and the same actions are available inside Claude Code or another AI tool, driven by your own team.
A report on your brand, then a session on what to do with it.
IGAIV works by appointment. There is no signup, and no self-serve dashboard.
You request a report
Tell us the brand, the markets that matter, and who you consider your real competitors. That is enough to start.
We run the analysis and send it to you
Prepared for your brand and your markets, not a template with your logo on it. You read it before we speak.
We walk you through it
You book a session and we go through what the answers show, where the ground is softest, and the order we would tackle it in.
Is this just SEO?
No, and it does not replace it. The two share a foundation: a fast, crawlable, well-structured site with clear and current content, and being talked about on sources others trust. The difference sits on top. SEO competes for a position in a list of links, where the customer still chooses. An AI answer is the choice. It names three or four brands and the rest of the market does not appear. You can hold page one and still be absent from the answer.
Will IGAIV get us recommended?
No one can promise that, and we do not. IGAIV shows you where you stand today, what is driving it, and what to change. The work and the outcome are yours.
Do you make the changes for us?
Where we can, yes. If your platform supports direct updates, we apply the changes through the connection, with your approval on each one. Most iGaming brands run bespoke or proprietary stacks, and for those the same actions are available through our MCP, so your own team can run them inside Claude Code or another AI tool. Either way the plan carries enough detail to be executed without coming back to us.
Which AI tools do you cover?
ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews. The same question returns different names in different places, so every question is run per market and reported that way.
Where to go next.
AI visibility for casino, sportsbook and lottery brands
What AI assistants say when someone asks where to play, and who gets named instead of you.
AI visibility for iGaming suppliers and B2B vendors
Whether AI names you when an operator shortlists aggregators, studios, platforms or payments.
IGAIV set against a visibility score
What each one gives you, and which suits the job you actually have.
What AI visibility means in iGaming
How answer engines decide which brands to name, and where it parts company with SEO.
Find out who AI is recommending in your markets.
Request a report on your brand and see what the answers actually say.