Checklist
The 14-day AI visibility checklist
Every step of the 14-day plan on one page, with no email wall: what to do, how to verify it, and which tool to open. Thirty to sixty minutes a day. The reasoning behind each step lives in [[ET1|the companion guide]]; the technical side starts with the free scan.
How to use this checklist
Same questions every time. The question set from days 1 to 2 is the ruler. Never change it mid-cycle.
Dated records. One table per run. Engines drift, so dates tell you whether it was your change or the model update.
One change at a time. Between day 2 and day 14, notes first, edits second.
Six phases, fourteen days
Day 1 · Write your fixed question set
The ruler for everything that follows.
Do Write 8 to 12 questions a real buyer would ask about the problem you solve: recommendation, comparison, and problem shapes; half branded, half unbranded; no yes or no questions.
Verify A colleague outside marketing understands every question without explanation.
Open Any notes app. Keep it to one page.
Day 2 · Take the baseline
Table one: the before picture.
Do Run every question in ChatGPT, Perplexity, and Gemini, one fresh chat each. Record per question: named, yes or no; described accurately or not; sources cited.
Verify The table is complete for all engines and questions, with today's date and saved screenshots.
Open ChatGPT, Perplexity, Gemini.
Day 3 · Run the free technical scan
Find what blocks engines from fetching you.
Do Enter your domain in the free scan and let the seven checks run: reachability, meta basics, crawler access.
Verify Each check shows pass, warn, or fail; fails and warns are written down with the date.
Open The free scan on the audit page.
Day 4 · Fix retrieval basics
Pages return, pages rank, pages exist once.
Do Fix what the scan flagged: error pages, site:yourdomain.com gaps in Google, duplicate canonicals, robots.txt rules that block crawlers you want.
Verify A site: search shows your key pages; each URL returns one canonical address; a scan re-run shows fewer warns.
Open Google Search Console and a scan re-run.
Day 5 · Collect the cited sources
See who the answers already quote.
Do Rerun the question set and write down every domain the answers cite, across all questions and engines.
Verify A raw domain list covering six or more answers per engine.
Open ChatGPT, Perplexity, Gemini.
Day 6 · Sort sources into types
From a list of URLs to a map of where answers come from.
Do Bucket each domain: industry publications, review platforms, directories, community threads, comparison pages, your own site.
Verify Every domain sits in exactly one bucket; counts per bucket written down.
Open Any spreadsheet.
Day 7 · Pick two or three source types
The targeting decision for days 8 to 13.
Do Choose the two or three buckets that recur most and are realistic for you: publishable, joinable, or listable.
Verify A one-line plan per chosen bucket, written before day 8.
Open Your notes from day 6.
Day 8 · Draft the comparison page
One page that states your differences as facts.
Do Name real alternatives, state specifics (numbers, features, limits), and cite the source of every claim inside the page.
Verify Every specific number on the page has a source, on the page.
Open Your CMS or site editor.
Day 9 · Publish the FAQ page
Answer the exact questions from day 1.
Do Write one answer per question-set question, in the buyer's words, stating facts.
Verify Answers agree with the rest of the site and with your profiles.
Open Your CMS.
Day 10 · Publish the how-to page
Steps a reader can test.
Do Write the procedure your product or service implies, with concrete steps and expected outcomes.
Verify Someone unfamiliar with your product can follow the steps without asking you.
Open Your CMS.
Day 11 · Sweep entity facts
One name, one description, everywhere.
Do Compare name, one-liner, services, hours, and location across your site, search profiles, and major listings; fix contradictions.
Verify Five checked profiles show identical facts.
Open Your listing profiles.
Day 12 · First channel
Be present where your buyers already read.
Do Contribute one substantial answer, or get listed, on one platform from your day 7 selection. State facts, not promotional spam.
Verify The contribution or listing is live and would be worth reading even if AI never quoted it.
Open The community, publication, or platform you chose.
Day 13 · Second and third channels
Stop at two or three, not ten.
Do Repeat day 12 on the remaining chosen channels. Keep every stated fact identical.
Verify Two or three live placements in total; facts match your site.
Open Your day 7 plan.
Day 14 · Retest and compare
Table two, next to table one.
Do Run the same questions, same engines, fresh chats. Record mentions, accuracy, and cited sources exactly as on day 2.
Verify Table two is complete and dated; a short written diff exists: what moved, what did not.
Open ChatGPT, Perplexity, Gemini, and your day 2 table.
When the two weeks end
Run the free scan again, put the new table next to the old one, and decide from the numbers. If you want the deeper version, 30 buyer-style questions across three engines with the sources they lean on, the written diagnosis reports it without promises. The [[ET1|guide]] explains why each step exists.
Frequently asked questions
Is this checklist gated behind an email?
No. This page is the whole checklist. If you want the reasoning behind each step, read the [[ET1|companion guide]]; if you want the technical side measured for you, run the free scan.
Can I run this plan for any industry?
Yes. The steps are industry-neutral; the question set is where your vertical enters. A clinic, a devtools vendor, and a law firm each write their own 8 to 12 buyer questions, then follow the same loop.
What do I do after day 14?
Retest every two weeks with the same set and keep the dated tables. When the loop is stable and you want more depth, the written diagnosis extends the same measurement with more questions and sources.