Order exactly what you want

You know your buyers better than a pricing table does. You know whether they ask ChatGPT or Perplexity first, whether Claude matters in your category, whether Grok is where your competitor's loudest fans live, and whether twenty questions covers your market or barely scratches it. So we built the report to be ordered the way you'd order it if you were sitting across from us: line by line, with the price of each line beside it, and nothing in the box that you didn't put there.

Here's what you get to decide. Which of the five engines judge you: switch on ChatGPT, Claude, Gemini, Perplexity and Grok in any combination, from one to all five. Which version of each does the judging: the newest flagship, the model every paid tier answers on, or the fast rung, each named by its real version and priced on its own line, so "ChatGPT" is never a mystery box. How many buyer questions they get: 20, 50 or 100, and each button shows the price the engines you've chosen would make of it, so the difference between coverage and a sample is a number you can read rather than a feeling. Whichever rungs you choose, every question is asked three times per engine with live search on, or once if you take the single pass at a third of the engine lines, and every answer is stored with the model that gave it. The page says why we recommend three: one answer is a draw, and three is the smallest number that can show the disagreement.

And nothing you didn't switch on runs. There is no default bundle that quietly includes an engine you didn't want, no question count that rounds up to the next tier, and no add-on that arrives already ticked. The configurator opens on the three-engine, 20-question shape the $180 audit sells, because a page should open on a price rather than a blank, and from there every change is yours: the rail rewrites itself on every click, the buttons for 50 and 100 questions carry the price your engines would make of them, and the server prices the same choices from the same catalogue when you pay, so the total you read is the total you are charged, and a checkout whose amount disagrees is refused. The questions are ours to suggest and yours to change: once the payment lands, the set we drafted from your brand, category and buyers is on the page to edit, cut or add to, and no engine is asked anything until you approve it.

The lines you can add, and the one you can't

Name up to four competitors on the form and each one becomes a $50 line you can take or leave. A competitor report reads the same stored answers with the other name as the subject: their mention rate, their share of voice, where they were recommended and where they were passed over, on the same questions, engines and model versions as yours. Nobody is asked again and nobody is contacted. Add the 90-day action plan for $150 if you want the dated version of the fix list, with content briefs tied verbatim to the questions you lost. Or take the same report monthly at the same price and the plan comes included, refreshed on every run; we will explain why the monthly costs the same as the single one in its own post.

The one line you can't order is a model we haven't run ourselves. It stays on the page so you can see it exists, marked not yet available, and it stays unbuyable until our adapter has answered on it, because a report you pay for shouldn't rest on a rung we've only read about. A rung we've run but not yet probed for cost sells at the vendor's list price with a margin, and is labelled list price beside the measured ones. That is a refusal that costs us a sale now and then, and we'll keep making it.

Why it's an order form and not a quote request

Because the alternative is a call, and we'd rather spend the call on your results than on your invoice. We give you the whole price before you pay by pricing each engine from what a grounded answer costs on that model, measured on our own account or taken from the vendor's list and labelled so, the passes you chose and the extraction included, and by letting the page do the arithmetic in front of you instead of behind a sales desk. If the number is wrong for you, change a line and watch it move: drop Grok, move Claude from Fable 5.1 to Sonnet 5, go from 50 questions to 20. You're the one holding the pen, and the total follows it.

Design agencies: the same form has a white-label switch, and with the cover set once under Settings, Report branding, every report you order lands with your name and colours on it and ours nowhere. Your clients' reports become your reports, and each client can have a different shape, priced on its own.

An admission, since you'll notice it anyway: the form asks more of you than a tier button does. You have to know your category, your buyer and your competitors well enough to type them in, and you have to make five or six choices before you see a total. We think that's the right trade, because a report shaped by someone who knows the market is worth more than one shaped by a plan name, and those choices are the part we can't make for you honestly. Someone will disagree and want the button. The fixed ladder is still there for them.

Then we run it

Pay, approve the questions, and the engines start within a minute. The report opens on your dashboard and is emailed to you the moment it lands, competitor reports alongside it, the action plan a button away if you took it, and every number on it traces to a stored answer you can read. If a call fails and never answers inside the window we give it, that call comes off a one-time bill on its own, pro rata, without you asking; on the monthly the same tally is credited to the period by hand.

And if you'd rather we did the work the plan describes, book a call or email us from the configurator and we'll implement it at our agency rates; either way the proof shows up in the next report, on the same questions, with every gap that closed and every one that didn't. We hope the first one lands with more of your name in it than you expected, and we'd be glad to help make the second one better than the first.