passcite
Free report
How it works

Find, fix, approve, then prove it

A loop, not a dashboard. It starts with the questions your client’s buyers ask an assistant, and ends with the same questions asked again after the changes are live — so what you hand over shows what moved, not what shipped.
The loop

Four steps, and a before/after

Each one says what it produces and whether it is running. 2 of 4 are.

01

Find

Running today

One client website in. Buyer questions out, with the name of whoever won each one.

We generate the questions a buyer actually types before choosing — comparative, place-bound, price-bound — and ask them of ChatGPT, Perplexity and Gemini. 12 of them on a free report, 30 on a paid one. Most never mention your client by name, because buyers do not know the name yet; that is the whole point of asking them.

Every answer is recorded with the model that produced it and the domains it cited. Nothing is averaged into a score. What comes back is a list of questions, and for each one the businesses and directories the engines named instead.

What comes back, from reports that ran.

02

Fix

Running today

The changes that would answer those questions, each with the questions it answers.

For each lost question the report states what the site would have to say for an engine to be able to quote it — literally, because engines quote claims and skip pages that only imply. Those statements cluster into actions: a comparison page, an FAQ, schema describing the business, a listing to claim.

The report splits them into two columns: what we publish for you on the left, and what only you can do on the right — claiming a listing, fixing a Google Business Profile, asking for reviews — each with steps and a checkbox. The test for being an action at all is whether it can be finished.

Both columns, item by item.

03

Approve

Being built

Nothing reaches a client's site without you seeing it first.

Every write is a draft in an approval queue. You read it, you approve or reject it, and only then does it publish. Anything that went live has one-click undo. Per-client rules let the client who trusts you with metadata auto-pass while the one whose pages you want to read first keeps waiting.

A fully automated competitor sells the absence of this step as the feature. That is a real trade and it is a real preference — if you would rather nobody looked, we are the wrong tool.

Every platform says which half it is offering — WordPress is where publishing lands first.

04

Prove

Being built

The same questions, asked again, after the change is live.

Publishing triggers the re-test. Same question set, same engines, and a report that puts the two runs side by side question by question — including the questions that did not move, which is the half that makes the other half worth reading.

It follows a publish rather than a calendar. Re-running weekly measures how much the engines drifted, not what your work did, and it would spend more on engine calls per month than the plan costs.

The other list

What this does not do

Published because the alternative is you finding out in month two. Each is a capability a tool in this category could reasonably be expected to have, and a decision not to build it.

  • Backlinks

    A different business with different risks, and the reports do not show links deciding these answers.

  • Social posting

    Engines do cite Reddit heavily. Posting into it on a client's behalf is not something we will automate.

  • Paid ads

    Nothing here touches a budget or a bid.

  • Weekly rank tracking

    That is a different product — Peec and Profound build it. The re-test follows a publish so the comparison means something.

  • Our own index of engine answers

    Every question goes through the official APIs, per run, and the model is recorded with the answer.

  • Rewriting your existing copy

    The first version adds pages and fields. Your published voice is not ours to edit.

Why the loop is the product

Every part of this exists somewhere already. Tools that measure AI visibility are a crowded category; tools that generate pages are a more crowded one. What almost nobody closes is the distance between them — between a finding and a published change, and back to the same finding asked again.

That gap is where the work actually is, and it is also where the uncomfortable question lives: does adding a comparison page and some schema really make an engine change its answer? Nobody in this category prints a before and after, ourselves included — not yet. The fourth step is how that stops being a rhetorical question, and it will be as capable of embarrassing us as of vindicating us.

FAQ

Questions

See who the engines recommend instead

12 buyer questions across ChatGPT, Perplexity and Gemini. You get the questions your client loses, who won them, and what is missing from the site. No score, no signup wall.

Get your free reportOne client website, no signup. Paid plans will ask the same questions 2.5× deeper.