Model Research IQ, our AI visibility platform
What AI models know and say about a company, how that compares with its rivals, with search and social, and with what people think, all in one place and all kept as a dated record.
What Model Research IQ is
Model Research IQ is the platform we use to produce every reading. For one company and its rivals it brings together what AI models answer, what people search for and say online, and what people tell surveys, and it holds all of it as one record with the same companies and the same measures throughout.
The record is the point. Every answer, every reading, with its date and the model's version, is what lets you measure change instead of guessing at it, and how fast that change is happening. A tool that shows you today's answer cannot tell you whether the last fix worked.
The platform is built to be AI and platform agnostic. It reads any model and any search engine and holds the result the same way, so a new model is a new column in the record, not a new tool. The calculations are done outside the AI wherever possible, and where a model's judgement is used the figure says so.
AI Overviews, search and social in one record
The record keeps the text of Google's AI Overview for every question, beside the search results, the social conversation and the answers from the AI models. One question, every place it is answered, so you can see where a brand is named, where it is only cited and where it is missing altogether.
Access is by invitation, and our details are on the contact page.
The measures Model Research IQ reports, and what you do with each
AI visibility
How often do AI models name the company when asked about its category? See whether you are in the answer at all.
Named and picked
When the company is named, is it the one recommended? Know where you win and where you lose to a rival.
Head-to-head preference
When a person asks "this company or that one", who does the answer favour, and for what? Be preferred overall, and know which point of difference you are preferred for.
Cited but not named
How often is the company's own page the source of an answer that does not name the company, and who is named instead? Turn a citation into a mention. The gap is the brief for the content team.
Claim accuracy
Do AI models get the company's facts right? Fix the incorrect facts and the pages that carry them.
Reasons
Why do AI models recommend the company? What do they hold against it? Lead on your strengths. Answer the objections.
AI against people
Do AI models see the company the way people do? Find where AI is ahead of or behind your customers.
Knowledge and search
What does a model already know? What does web search add? Decide whether to change what is learned or what is found.
Change over time
What was gained and lost since the last reading? See whether what you did worked.
Search and social trends
What are people searching for and saying, and does AI follow? Spot a topic in social before it reaches search and the AI answer.
Questions people ask us
Is Model Research IQ a dashboard or a report?
Both. The platform holds the record and the charts, and a reading is a dated report drawn from it. Most clients read the report at the interval they have chosen and use the platform when they want to ask a question the report did not answer.
Can we load our own data into Model Research IQ?
Yes. Your brand tracker, your search data and your social listening go in beside the AI answers, matched to the same companies and measures. That is how the AI against people reading is built.