For operators
Absent from the shortlist is not a lower ranking.
It is not existing in the conversation. When someone asks ChatGPT or Gemini where to play, the answer names three or four brands and the rest of the market does not appear at all.
Named in 7 answers. First in none of them.
Four competitors take almost every opening slot. Two of them are named in answers where Northgate is not mentioned at all.
Across 5 engines / 3 markets
Illustration. Brands shown are fictional.
The questions that decide where someone opens an account.
These are not obscure. They are the same questions people have always asked, put to an assistant instead of a search box, and answered with a handful of names rather than a page of links.
Every one of them is being answered right now, thousands of times a day, in every market you operate in. The answers differ by market and by engine, and no one on your team has read any of them.
Five kinds of question, all of them commercial.
The exact wording varies endlessly. The shapes do not, and each shape produces a shortlist.
- Which site is best for a given sport or game, asked plainly and answered with names.
- Alternatives to a named competitor, where the model lists who else to consider.
- Which brands are worth using in a given market, answered differently in every market.
- Questions about withdrawal speed, bonuses and terms, where one brand tends to own the answer.
- Comparisons between two brands, where the model decides which one to favour.
Where you stand, and what is putting others there.
Not a score. The answers themselves, the names inside them, and the reasoning behind a ranked list of changes.
- The shortlist your brand actually appears on, question by question and market by market.
- Which competitors are named instead, and what is putting them there.
- Which third-party sources the models keep drawing on for your category, and whether you feature in them.
- What on your own site stops a model describing you with confidence.
- The plan: what to change first, and why that order.
Specific enough to brief someone on Monday.
A finding names the question, the engines it appeared on, the competitor taking the slot, and the sources the answer was built from. Your team can verify every part of it before committing to the work.
That is the difference between a report you act on and a report you file.
Why is Kingfisher cited ahead of us on payout questions?
Three review sources carry Kingfisher. None mention Northgate.
Kingfisher is named in 11 of 14 answers about withdrawal times across your markets, on 4 of 5 engines. Each answer traces back to the same small set of independent review pages, and Northgate is absent from all of them.
Recommended sequence
- Publish a plainly worded, quotable page on withdrawal timings.
- Mark it up so the facts are machine-readable.
- Pursue presence in the three sources the answers rely on.
Illustration. Brands shown are fictional.
We already rank first on Google for these terms. Does that not cover it?
Not reliably. Ranking well is a good sign and it is not the same test. An AI answer names three or four brands and the rest of the market does not appear, and it is assembled from sources beyond your own site. It is common to hold page one and still be absent from the answer.
Our brand is well known in our markets. Surely the models know us?
Being known and being named are different things. A model reaches for whatever it can describe with confidence, and brand awareness among customers does not automatically translate into a clear, quotable entity for a machine. That gap is one of the more common findings.
Answers change every time we ask. How is this measurable?
Variance is part of the data. Running a question once tells you almost nothing. Running it repeatedly, across five engines and per market, shows you which names are stable, which are volatile, and which questions no brand has locked down yet. The unstable ones are usually the opportunity.
Do you need access to our site or our analytics?
Not for the analysis. It works from what is publicly visible, which is exactly what the models work from, so nothing has to be connected to find out where you stand. Access only comes into it later, if you want the changes applied directly rather than handed over, and that is your call to make once you have seen the findings.
Where to go next.
AI visibility for iGaming suppliers and B2B vendors
Whether AI names you when an operator shortlists aggregators, studios, platforms or payments.
What AI visibility means in iGaming
How answer engines decide which brands to name, and where it parts company with SEO.
How IGAIV identifies, diagnoses and directs
The questions, the benchmark, the ranked plan, and how the changes actually get made.
How AI answer engines choose which brands to name
Models do not rank you and reject you. They reach for what they can describe with confidence.
Find out who AI is recommending in your markets.
Request a report on your brand and see what the answers actually say.