Everything outside your own site that determines whether you show up
AI engines build answers from the pages they retrieve. Many of those pages are reviews, listicles, directories and forum threads. We find the ones that decide your category, check if you are named on each, and show you where you are missing.
Acme in scheduling software
Pages engines cite for the category
- ABest scheduling tools for small teamslisticle.exampleListicleactive
- BScheduling software reviewsreviews.exampleReview platformactive
- CWhat do you use for team scheduling?forum.exampleCommunityactive
- DScheduling apps directorydirectory.exampleDirectorystale
- EHow to choose a scheduling toolmagazine.exampleEditorialactive
Illustrative data. Acme and every .example site are made up.
What we measure
We run the collection. You get the gaps.
One engagement covers the corpus, your presence on it and the pages where you are missing.
Corpus
Pages engines cite for your category
The corpus, as a list of real pages
Listicles, review sites, directories and community threads that engines cite for your category. Each page has a state and a last-checked date.
Sat 17 Oct
2:51 AM
Email from ranq
A new page entered your corpus
Email from ranq
Scheduled run finished
Movement since the last run
Every page state and presence check carries a date. The report lists pages that entered or left your corpus and pages where your presence changed.
Ask anything
Roundups and review sites name your competitors first [1][2]. A forum thread also mentions your brand [3].
References [1] a listicle · [2] a review site · [3] a forum thread
Every answer, stored raw
We keep the full response from every run, with every citation. When we change how citations are extracted, we recompute your history from the stored answers.
A gap list, walked through with you
We go through the ranked gaps with you and explain why each page matters. You leave with the pages to work on first.
Engines we support, scoped per engagement
Our collector supports ChatGPT, Perplexity, Gemini, Google AI Overviews and Google AI Mode. We agree which ones your engagement runs, then run the collection for you. You do not configure providers or manage keys.
Prompt set
We write the questions your buyers ask engines. The set is weighted toward unbranded category queries, plus branded and comparative ones.
Corpus
We build the set of third-party pages that engines cite for your category. Pages enter, get checked again, go stale and drop out.
Presence checks
We fetch each corpus page and check whether your brand is named on it. Every check is stored with the date it ran.
Gap scoring
Pages where you are absent are ranked by how often engines cite them and by page type. A listicle and a forum thread need different work.
Brand matching
We look for your name, your aliases and your domain. If your name is also a common word, that word alone does not count.
Read and recommend
We measure and recommend. We do not write content or contact site owners. Your team decides what to act on.
How it works
How an engagement runs
We define the category, build the corpus, check presence and re-run. You get the gaps and how they move.
[your brand]
Category [category]
Prompts
Keywords
Prompt set
best [category] tool
[category] for startups
[brand] review
is [brand] any good
[brand] vs [rival]
[rival] alternatives
Unbranded
Branded
Comparative
Scope
We measure. You decide what to do.
Your team owns every action. We show you where you are missing and why each page matters. Then we answer your questions about it.
What you get
- A corpus of the pages engines cite for your category
- A presence verdict for each page, with the date we checked it
- A gap list ranked by page type and by how often engines cite each page
- A citation history we can recompute from every stored engine answer
What we leave to you
- We do not write content
- We do not contact site owners or submit listings
- We do not act on your behalf
- We do not resell engine output
How we collect data
Engine answers come from a data provider. Then we check the pages.
Here is what each step does, so you know what a gap rests on.
- ChatGPT
- Perplexity
- Gemini
- Google AI Overviews
- Google AI Mode
The engines our collector supports. Each engagement runs the ones its brief needs.
Engine answers
We send your prompts to the engines through a third-party data provider. We store each full response with a checksum of the bytes as they arrived.
Citations
We pull out only the pages an answer cites. We strip tracking parameters and unwrap redirect links, so one page counts once.
Page checks
We fetch each corpus page and read it for your brand, with a rendered browser pass when a page needs one. A page we cannot read is never counted as missing.
Recomputable history
Citations are extracted from the stored responses, never from a fresh call. When the extraction changes, we recompute your citation history from source.
Some engines prohibit automated querying in their terms. Read how we handle that.
Engagements
Priced per engagement, against the data it needs.
Single brand
One brand in one category, tracked on a schedule we agree with you.
Priced per engagement
- Category definition and prompt set
- Corpus built from what engines cite
- Presence checks across the corpus
- Gap list ranked by citation and page type
- Scheduled re-runs with a movement report
Portfolio
Several brands or products, each measured against its own category.
Priced per engagement
Everything in Single brand, plus
- A prompt set and corpus per brand
- A schedule set per brand
- One report across the portfolio
- A competitor set per brand
We price each engagement against the data providers your brief needs. More engines, more prompts and a tighter schedule cost more to collect. The price follows that cost.
Email us at [email protected] or
A different brief
If your question does not fit either shape, tell us what you need to know. We scope an engagement around it.
What you should know before we talk
What is presence rate, and why not share of voice?
Share of voice counts how often engines name you in their answers. It tells you that something changed, not where to act. Presence rate counts how many of the pages that decide your category name you. Each page that does not name you is a place to act.
Where does the data come from?
We get engine answers from a third-party data provider. Our collector supports ChatGPT, Perplexity, Gemini, Google AI Overviews and Google AI Mode, and we agree which ones your engagement runs. Then we fetch the pages they cite and check each one for your brand.
Do you write content or do outreach for us?
No. We measure and recommend. We do not generate content, submit listings or contact site owners. You get a ranked list of pages and the reason each one matters. Your team decides what to do with it.
How do you decide that a page names us?
We look for your brand name, your aliases and your domain, as whole words. If your name is also a common word, the word alone is not enough. We mark that page as unclear and do not count it. A page we could not fetch is never counted as missing.
How often do you re-run?
On a schedule we agree for each engagement. A tighter schedule costs more to collect, so we set it against how fast your category moves.
What happens to the raw data?
We store the full response from every run, not only the metrics we parse from it. When we change how citations are extracted, we recompute your history from the stored responses.
Is it legal to query AI engines programmatically?
Not always. The terms of some AI engines prohibit automated querying, and there is active litigation in this area. We collect for measurement only. We never resell raw transcripts. You receive our analysis, not a copy of engine output.
Why is there no login?
Because we run ranq for you. We configure the providers, keep the corpus current and send you the results. There is no dashboard to learn and there are no keys to manage.
Find the pages where
you are missing.
Tell us your brand, your category and who you compete with. We reply with how we would scope the engagement.
Or copy