FREE TOOL · NO ACCOUNT · THE AUTOMATIC SAMPLE NEEDS AN EMAIL
The Baseline Kit: audit yourself.
This is the DIY version of the audit we sell — same protocol, your clicks. Answer four questions and the Kit generates about 100 questions your buyers actually ask AI, grouped into named clusters, each with a run on all seven engines: ChatGPT, Google AI Overviews, Google AI Mode, Perplexity, Gemini, Claude and Grok. Score what you see and the Scorecard turns it into the full report.
Run as many or as few as you like. Every question on every engine is up to 700 runs, and nobody does that in one sitting — the report only ever counts what you actually ran, and says so on every number. Start with one cluster; the maths is honest at any size.
STEP 1 / GENERATE
Four questions about your business.
STEP 2 / RUN & SCORE
Pick a question. Press run. Every engine.
One click puts the question you choose to all seven engines and fills its row in for you. Then read the Scorecard. Every other row is yours to run by hand below — or ours, all hundred, three times over, for $149. You can overrule anything the machine scored; every number stays editable.
ONE QUESTION, EVERY ENGINE, ONE CLICK
Checking which engines are switched on…
Five engines are asked through their makers’ APIs with web search on, and each chip names the model it asks. Google’s AI Overviews and AI Mode have no API, so those two are read through a SerpApi capture. An answer through an API stands in for the app your buyers use and can come out differently, so every question below also links to each app.
HAVE US RUN THE REST
The rule that matters more than any of that: start a new or temporary chat every time, and don’t run these from an account that has spent the week researching your own company. A chat with memory of you will name you, and what it shows is your account’s view. The market sees a fresh chat. Every affordance above was checked by hand on 18 September 2026.
HOW TO SCORE AN ANSWER
Read the answer, then press every chip that describes it. A chip stays pressed until you press it again, and you can change any score later.
STEP 3 / YOUR BASELINE
Per platform. Never blended.
Cross-platform overlap in AI answers is tiny — in our measured runs, most questions come back different on every engine. One blended score would hide exactly the thing you need to know: which platform you’re losing on, because that decides which fix you need.
STABILITY SPOT-CHECK — BEFORE YOU BELIEVE ANY OF IT
AI answers regenerate every time. We ran 100 buyer questions on seven engines three times over, and a pass is one full run of every question on every engine. When an engine named a business in one pass, the next pass named it again on 95 of 174 pairs. So: re-run three of your questions on ChatGPT right now and record whether your result repeated.
WHAT THIS BASELINE IS — AND ISN’T
You now have the before-photo. The photo doesn’t do the work.
What you just built is the honest starting line: where you’re named, where you’re not, per platform, with a stability check on top. The loop that follows it runs on the monthly plan. The AI writes your Business Profile fields and your review ask from your own site and baseline for you to paste, a note every Monday, Wednesday and Friday morning hands you one task, and the same questions run again so you can see what moved. Done For You, $2,500 to start and then $1,995 a month, is where we apply the work inside your own accounts after you approve it.
Prefer to ask a person first? hello@commandmodule.com
NO PLACEMENT GUARANTEES — NOBODY HONEST CAN MAKE ONE. WE PROVE WHAT CHANGED INSTEAD.
Working Reddit yourself? The free Reddit Assistant finds answerable threads and keeps your pacing inside the lines that get accounts banned.