Verified against ChatGPT · 2026-08-05
Turn your AI Visibility Checker score into a 30-day fix plan
Turns a raw AI Visibility Checker score and its failed checks into a plain-English explanation and a ranked 30-day fix plan, ordered by score points recovered per unit of effort.
The prompt
Ready to copy — highlighted parts are example details you can swap.
You are an AEO consultant. A visitor ran the AI Visibility Checker and got a score with specific failed checks — turn that into a plan they can actually execute, not a restatement of the number. SCORE: 54 / 100 SITE TYPE: B2B SaaS marketing site FAILED CHECKS: GPTBot: Blocked (Disallow: / under User-agent: GPTBot). No /llms.txt found. FAQPage schema missing on homepage. The score is weighted: crawler access 40%, structured data 20%, page basics 20%, llms.txt 10%, sitemap 10%. TASK 1. Identify which weighted bucket above is most responsible for the current score, based on the failed checks given. 2. For every failed check listed, give one plain-English sentence on what it actually means, and the concrete fix — not "improve SEO," the literal change (for example: "remove Disallow: / from the GPTBot group in robots.txt"). 3. Produce a ranked 30-day plan, ordering fixes by estimated score points recovered versus how much work each takes. 4. Close with one honest sentence: a higher score means the site is eligible to be cited by AI engines, not guaranteed to be — content quality and topical authority still decide the rest.
Customize the highlighted detailsoptional — the prompt above already works
Why this works
The prompt reasons directly over the checker's own published weighting (40/20/20/10/10) instead of duplicating or second-guessing its scoring logic, so the fix plan uses the same vocabulary as the report itself. Ordering fixes by score points recovered per unit of effort converts a diagnostic list into an actual work order a non-technical site owner can hand to a developer. The closing caveat is load-bearing, not boilerplate — it's the tool's own stated limitation (eligible to be cited, not guaranteed to be), and repeating it here stops a rising score from being mistaken for a guarantee the fix plan can't actually make.
What you get back
Most responsible bucket: crawler access (40% weight) — GPTBot is fully blocked, which alone caps the score well below what structured-data and llms.txt fixes could recover. 1. Remove `Disallow: /` from the `User-agent: GPTBot` group in robots.txt — recovers the largest single chunk of the crawler-access weight, ~15 minutes of work. 2. Publish /llms.txt listing your top 5 pages — recovers the full llms.txt weight, under an hour. 3. Add FAQPage schema to the homepage — recovers structured-data points, half a day including review. A higher score means AI engines are able to fetch and parse the site, not that they will choose to cite it — that still depends on content quality and topical authority.
Verified against
ChatGPT GPT-5.1 · 2026-08-05
Claude Claude Sonnet 4.5 · 2026-08-05
Changelog
- 2026-08-05 — Initial publish, built directly around the checker's published 40/20/20/10/10 weighting so the fix plan uses the same scoring language as the report itself.
Building this for real?
This is a free starting point. If you'd rather have SEO built and running for your business, that's Scult's day job.
EXPLORE SEO
