Monthly at the same price, plan included

Stakes first. A single report tells you where you stand with ChatGPT, Gemini, Claude, Perplexity and Grok on the day it ran. It cannot tell you whether the page you published last week moved anything, whether the roundup you got into was read, or whether a competitor's new comparison table just took your place on the question that matters, because those are all differences between two runs and a single report is one run. The work is monthly, so the measurement has to be. Track, Grow and Command on the fixed ladder already re-run every month with a delta, at $249, $495 and $950; the monthly custom report is the same idea for a report you shaped yourself.

What the monthly is

Build a report in the configurator: the engines, the model version each answers on, 20, 50 or 100 questions, any competitor reports. Then, at step five, choose monthly instead of once. The same report re-runs every month with the same questions on the same models, the 90-day action plan is rebuilt on every run from that run's losses, and the report carries a movement delta against the previous month, so you can see which questions flipped from lost to won and which went the other way. It's charged monthly by card, cancelled from your dashboard, and runs to the end of the period when you do.

The price is the one on the rail. A report that costs $445 once costs $445 a month, and the plan that would be a $150 line on the single report is included, so a buyer who wanted the report and the plan every month pays less by subscribing than by buying both twice, and a buyer who wanted only the report pays the same either way. That's the entire pricing structure, and it fits in one sentence on purpose.

Why no discount

Most subscriptions are cheaper per month than the one-off because the vendor wants the commitment and prices the second month as a reward for it. We didn't, for a reason a buyer should hear from us: a monthly re-run costs us exactly what the first run cost. Every question is asked again, three times per engine or once if you bought the single pass, with live search on, at the retail rates we pay the vendors. Nothing is cached, nothing is reused from last month, and nothing is cheaper the second time, so a discount would have to come out of the measurement, and the measurement is the product. We won't sell a cheaper second month that's quietly a smaller one.

Why is the plan free on the monthly and $150 on the single, then? Because the plan is worth most when it's refreshed against a new run, and least when it's a one-off document nobody grades. On a single report we charge for it as the separate piece of work it is. On the monthly it's the reason the subscription exists, so it rides along, and the next run tells you whether the last plan worked. An opinion, and one that costs us the $150 line on every subscriber: a plan that isn't re-graded monthly is a PDF, and we'd rather not sell PDFs.

What refreshed means

Not edited. Rebuilt. The plan reads the stored answers from the run it's attached to, the questions you lost that month and the pages the engines cited over yours, and produces positioning, the 30-day quick wins, content briefs tied verbatim to those questions, the third-party pages worth being on, the technical checklist, and how to measure whether it moved. Next month it reads next month's answers. A question you won drops off; a question you lost for a new reason comes on with the new reason attached. Set two plans side by side and you can read the delta as a to-do list: what you did that worked, and what you haven't done yet.

One thing stays fixed on purpose: the questions. A delta only means something if both runs asked the same thing of the same models, so the monthly holds the questions and the model versions steady and the report says which they were. Change the shape, adding Grok or moving from 20 questions to 50, and that is a new report with its own first month rather than a continuation, because a delta between two different instruments is a coincidence dressed as a trend, and we would rather start your count again than sell you one.

An admission: a monthly delta on 20 questions is noisier than we'd like. AI answers vary between runs, three samples per question is how we report that variation as volatility rather than hiding it, and a single question flipping on a single engine can be the market or can be the machine changing its mind. The plan does not pretend otherwise. It never promises a ranking, it says what the engines said and not what's true, and we'd rather you read a small honest movement than a large flattering one. Search on, or you measured the past is the longer version of why.

If you'd rather we did the work

The plan is written so your team can run it. If you'd rather we did, we implement it at our agency rates, quoted by conversation rather than as a line on the invoice, because the work depends on the plan and the plan depends on the run. The content, the placements, the schema: we do it, and the next monthly report grades us on it, on the same questions, in the same delta you'd use to grade yourselves.

Subscribe from the configurator, approve the questions we suggest, and the first run starts within a minute. The first delta arrives with the second report, one month later, and that's the moment the subscription starts earning its price: the same questions, re-asked, with every gap that closed and every one that didn't, and a fresh plan for the ones that didn't.