Two minutes
What an AI citation actually is.
No jargon. One real category, one real question, and where the money goes.
We make sure a machine can tell what you sell and why you're the answer — so you turn up in the shortlist your buyer never sees you compete for.
WhatsApp is optional, and only ever used to send the call link. No follow-ups, no list.
$750, invoiced after a 20-minute call — so we only take it on if it will pay for itself. Or run ten checks yourself, free →
Read this before the form. We would rather lose the sale here than take $750 from someone it will not pay back.
Most agencies sell all three. Here is the honest split.
Entity resolution, structured data, answer-shaped pages, clean feeds. Every piece of it can be checked by you, on your own site, without taking anyone's word for it. Schema validates or it doesn't. Your entity resolves or it doesn't.
This follows from layer one. Measured monthly, with the rented share broken out so you can see which part of it would survive a change in how the engines retrieve.
Years of press and relationships build this. It isn't something we sell. You'll get an honest read on where you stand, and that's all.
Layer one is the precondition. A brand a machine can't read won't get retrieved or remembered, whatever it spends.
Ask ChatGPT to recommend something. It searches, finds a listicle you're on, names you. Looks like a win.
Ask again with search switched off. Most brands vanish.
One survives a change in how the model retrieves. The other doesn't. Every dashboard in this category reports a single citation number that can't tell them apart.
We run both and give you the split.
citrel.ai · 45 prompts, five engines · updated monthly, up or down
Every number we publish carries how we got it. Verified means you can re-run it yourself and get the same answer. A composite score inherits the lowest confidence of its inputs, and we never average across engines.
See the full index — all 45 prompts and the method →
Read it as: we're named in 8 of 45 prompts on ChatGPT, and 2 of those 8 survive when retrieval is switched off. Across all five engines we're named 58 times and hold 12 of them. 79% of our own visibility is rented. We're a new company. You can watch these move.
No jargon. One real category, one real question, and where the money goes.
The honest way to value this: citation is a multiplier on the search and paid placements you already fund, not a channel of its own. Price it that way and it makes sense. Price it as traffic and it never will.
You'll find all of this yourself within a week. Better you hear it now, with what we do about each one.
Otterly, Peec, Profound, Rankscale, Scrunch and Athena are good products and we use them. Some are starting to ship fixes too. The difference that lasts is not who diagnoses better — it is that a person here actually writes the schema, builds the entity and ships the pages.
Most clients run both. Keep the dashboard, and stop paying it to tell you the same bad news every month.
The fairer comparison for upkeep is not a dashboard, it is a retainer. SEO and content retainers in the US and UK run $2,500 to $8,000 a month. We are the disciplined cheap end of that, not the expensive end of tooling.
Nine of the ten agencies we looked at make you book a call to find out. Here's ours.
The audit fee comes off the build. Nothing here is a subscription you have to escape.
A small consumer brand in its first year. No traffic, no rankings, no audience, no ad budget. Category and country withheld at their request.
See exactly what $750 gets you — all 35 sections, explained → · or a redacted sample
Ours: the AI-surface impressions, the top-10 queries and the clean product feed are the machine-readability work. Not ours: the 101 orders came mostly from the founder's own community posting. It sits here as context, not as something to take credit for. Attributing orders to citations is exactly the claim we say nobody can make.
We don't name the brand publicly, and we no longer screen-share a client's live accounts — that is their revenue data and it isn't ours to show. What you get instead, on the call, is a live redacted walkthrough of the actual dashboards — figures visible, everything identifying removed. Ask for it and it is the second half of the twenty minutes.
Retrieval engines like Perplexity and Google's AI Overviews shift in four to eight weeks. ChatGPT pulls from training data as well as search, so anything that has to reach the training corpus takes six to eighteen months. The audit tells you which of your prompts sit in which bucket.
You keep all of it. The schema, the pages, the entity work, the measurement baseline. It's published on your domain in your name. There's nothing to switch off and nothing to hand back.
No. Models change, retrieval changes, competitors move. We guarantee the build gets done to a written spec and the reporting stays honest. If a number goes flat we say so before you have to ask.
Free audits attract people who were never going to buy, and we'd end up rushing them. $750 buys 45 prompts run properly across five engines plus the retrieval-off comparison. It comes off the build if you go ahead.
Citrel AI is a one-person practice run by Renu Gupta, registered in Agra, India, with the full address and a working phone number on the contact page. Our clients are in the US, UK, Europe, the Gulf and Australia. We overlap at least four hours with US Eastern, we publish our prices, and we publish our own index so you can judge us on what we show rather than where we sit.
The same $750, converted at your card or bank's rate on the day. We are an Indian sole proprietorship below the GST registration threshold, so no GST is charged and no input credit arises — the invoice is the whole cost. If that changes, the invoice will say so and the price on this page will not move without notice. For clients outside India this is an export of services, and no Indian GST applies either way.
Yes, and you can check. Our robots.txt, sitemap.xml and llms.txt are all live, our structured data validates, and our own visibility index is published monthly including the months it gets worse. Every file we would ask you for, we ship ourselves.
Nothing. The audit runs entirely on public data and your domain name. No Search Console, no analytics, no admin logins, no NDA required to get started. If you go ahead with the build we will ask for Search Console and whatever CMS access is needed to publish, and nothing beyond that.
No. It sits underneath whatever they do. They work on rankings and content; we work on whether a machine can resolve who you are and lift a clean answer from your pages. If you have an agency, we will send them the spec and they can implement it themselves. Plenty of clients do that.
It overlaps. The difference is what you optimise for. Search ranks pages and hands the user ten to choose from. An answer engine picks one and explains it. That changes what the machine needs from you: a resolvable identity, structured facts, and content in a shape it can lift.
45 prompts, five engines, five working days. $750 for the first ten clients.
WhatsApp is optional, and only ever used to send the call link. The fee comes off the build if you go ahead.
