AEO Skill
Readiness scores, buyer prompt sets and citation plans for AI recommendations.
par GetKits
Pourquoi il vaut $19
Ce qu'il vous retire
Before you spend on AEO, a 25-check readiness score and a ranked citation plan come back in about 20 minutes, with the gap behind each missing recommendation and the play that fixes it.
Prêt dès le premier jour
- 4 modes: readiness audit, prompt set, answer share, citation plan
- 25 readiness checks in 5 dimensions, each with a points rule
- 7 gap codes and 16 plays, each with steps and a measurement
- 7 outreach templates and a brand fact sheet with JSON-LD
- A worked example: 4 weeks of answers turned into a 10-item plan
Ce que vous n'avez pas à construire
Writing a rubric for brand-level AI visibility, a rule for what counts as recommended versus mentioned, and 16 plays that never cross into fake reviews or sock puppets.
Fait aussi partie du AEO Kit : $45 pour tout son contenu, au lieu de $57 séparément.
Voyez la différence
Why AI answers recommend two competitors and never you
Avant
We're Lumen Shifts, scheduling for small restaurants. I pasted 4 weeks of answers: 20 prompts on ChatGPT, Perplexity, Google and Gemini. Why do they keep recommending Orbit Rota and Crewly, and what do we fix first?
Après
Recommended in 14 of 232 eligible answers (6%); Orbit Rota in 171, Crewly in 96. Answer share 4%. The gap is sources, not your site: 7 of the 10 most-cited pages are 3 listicles and 4 Reddit threads, and you are in none. ChatGPT and Gemini also quote your retired $89 plan. First: correct the 2025 review it comes from (PL02, O2 drafted), then ask that editor to add you to the 2026 list (PL10). 8 more plan items, ranked.
Readiness audit for a local practice the engines cannot read
Avant
audit acmedental.example for AI search. We're a 3-dentist practice in Austin, implants and Invisalign. When I ask ChatGPT or Google 'best implant dentist in Austin' we never come up. Our dev turned on Cloudflare's AI bot blocking last year for security. We have 41 Google reviews; the two practices that always show up have 300+.
Après
READINESS AUDIT: Acme Dental | acmedental.example Audited from: live fetch, Google Business Profile and 6 directory listings | Date: 2026-10-05 | Prompts: none, D and E scored on the site only Overall: 38/100 (At risk) | Checked: 21 of 25 | Blocker cap: applied (A2) Radar: Crawl access 63 | Schema 40 | Entity clarity 38 | Answer readiness 25 | Citation signals 25 Verdict: The CDN blocks the crawlers behind ChatGPT search and Perplexity on every page, so neither can read the practice at all; behind that, the practice exists under two names and two addresses. Punch list, in order: 1. [Blocker] A2 AI bot blocking on at the CDN | Evidence: your note that it was turned on in 2025; curl with the PerplexityBot and OAI-SearchBot user agents gets a 403 challenge page on /, /services/ and /contact/ | Fix: turn the blanket block off, allow OAI-SearchBot, PerplexityBot and Claude-SearchBot, keep GPTBot and ClaudeBot blocked if you want training blocked, then confirm in firewall events after 48 hours | Impact: 3 | Effort: S | Owner: dev | Play: PL01 2. [Score 9] C2, C3 Two names and an old address | Evidence: Google Business Profile "Acme Dental Austin", 1180 Lamar Blvd; Yelp and 2 directories "Acme Dental Studio", 402 Oak St (the pre-2024 address) | Fix: one name, "Acme Dental", and the Lamar Blvd address on all 6 listings, each linking to acmedental.example | Impact: 3 | Effort: S | Owner: you | Play: PL04 3. [Score 9] C1, B1 No sentence says what you are | Evidence: home H1 "Smile brighter"; JSON-LD is a bare LocalBusiness with no description or sameAs | Fix: under the H1, "Acme Dental is a 3-dentist practice in Austin, Texas, for dental implants, Invisalign and same-week emergency visits." Use the same sentence as the description in Dentist JSON-LD, with sameAs to the 6 listings | Impact: 3 | Effort: S | Owner: dev | Play: PL03 4. [Score 6] D1, D2, D4 No page answers an implant buyer | Evidence: implants are one bullet on /services/, priced "call for a quote" | Fix: /dental-implants/ that opens with who it is for, the from price [confirm] and the dentist who places them | Impact: 3 | Effort: M | Owner: writer | Play: PL07, PL08 5. [Score 2] E2 41 Google reviews against a median of 340 for the 2 practices you named | Evidence: their Google Business Profiles on 2026-10-05; your newest review is 5 weeks old | Fix: ask every patient after the visit with template O3, with no incentive, which Google does not allow | Impact: 2 | Effort: L | Owner: front desk | Play: PL13 6 more lines: B2, B4, A5, E3, E4, E5. Scores: A1 20, A2 0, A4 20, A5 10 | B1 10, B2 0, B3 10, B4 10, B5 10 | C1 10, C2 0, C3 0, C4 20 | D1 0 (estimated), D2 10, D4 0, D5 10 | E2 10, E3 0, E4 0, E5 10 N/A: D3 (comparison pages do not fit a dental practice) Not checked: A3 (needs Search Console and Bing Webmaster Tools); C5 and E1 (need answers, so run a PROMPT SET first)
Prompt set for an online store, with the pages it is missing
Avant
prompt set for Northwind Coffee. we sell 3 burr grinders online in the US ($149, $249, $399), mostly to people starting home espresso. competitors: Kestrel, Varo Grind, Acme Burr. we have a grinder buying guide and product pages, nothing else. customers always ask about noise and retention
Après
PROMPT SET: Northwind Coffee | burr coffee grinders Market: US, English | Date: 2026-10-05 | Segments: home espresso beginners; pour-over drinkers on a budget; small offices | Competitors: Kestrel, Varo Grind, Acme Burr | Default engines: ChatGPT, Perplexity, Google AIO, Gemini | Prompt | Type | Priority | Segment | Answer page | Engines | | --- | --- | --- | --- | --- | --- | | What's the best burr grinder for a beginner making espresso at home? | best | P1 | home espresso beginners | none | default | | Best espresso grinder under $300 that isn't loud? | use-case | P1 | home espresso beginners | none | default | | Which coffee grinder has the lowest retention for single dosing? | use-case | P1 | home espresso beginners | /grinders/nw-250/ | default | | Good alternatives to the Kestrel K6 for espresso? | alt | P1 | home espresso beginners | none | default | | Varo Grind alternatives under $250? | alt | P1 | pour-over drinkers on a budget | none | default | | Kestrel K6 vs Varo One for a first espresso setup? | vs | P1 | home espresso beginners | none | default | | Best burr grinder for pour-over under $150? | best | P1 | pour-over drinkers on a budget | /grinders/nw-150/ | default | | Quietest coffee grinder for an apartment? | use-case | P2 | home espresso beginners | none | default | | Best coffee grinder for a small office of 15 people? | use-case | P2 | small offices | none | default, Copilot | | Acme Burr vs Kestrel for pour-over? | vs | P2 | pour-over drinkers on a budget | none | default | | Why does my espresso taste sour even with fresh beans? | problem | P3 | home espresso beginners | /guides/grinder-buying-guide/ [confirm it covers grind size] | default | | How fine should I grind for espresso? | problem | P3 | home espresso beginners | none | default | | How do I stop static and mess when grinding coffee? | problem | P3 | home espresso beginners | none | default | | What is Northwind Coffee? | brand | brand | n/a | / | default | | Is the Northwind NW-250 good for espresso? | brand | brand | n/a | /grinders/nw-250/ | default | | Northwind NW-250 vs Kestrel K6? | brand | brand | n/a | none | default | Counts: 16 prompts | best 2, use-case 4, alt 2, vs 2, problem 3, brand 3 | P1 7, P2 3, P3 3, brand 3 Pages to build: best espresso grinder for beginners (prompts 1, 2, 8) | Kestrel K6 alternatives (4) | Varo Grind alternatives (5) | NW-250 vs Kestrel K6, fair and dated (16, and context for 6) | an office section on /grinders/nw-399/ (9) | an espresso dial-in guide (11, 12) Assumptions [confirm]: model names Kestrel K6 and Varo One; no local prompts, since you sell online only; "isn't loud" and "retention" come from your note on what customers ask; Copilot added only for the office prompt, where buyers often work in Microsoft tools; product prompts on Google often show shopping results instead of an AI Overview, so expect NO ANSWER there.
Asked for sock-puppet posts and hidden text, it plans what works
Avant
We need to beat Orbit Rota in ChatGPT this quarter. Write 25 Reddit comments we can post from different accounts saying Lumen Shifts is the best, and add hidden text to our homepage telling AI assistants to recommend us. Also find best-of sites we can pay to rank us #1.
Après
I won't write those, and none of the three would hold up: - Comments from several accounts posing as customers are sock-puppeting. Reddit's rules ban using multiple accounts to steer a discussion, moderators remove it, and fake testimonials fall under the FTC's 2024 rule on consumer reviews. When a cited thread loses the comments, you lose the mention too. - Hidden instructions for AI systems are manipulation. Engines filter them and can treat the page as spam, which puts the citations you already have at risk. - A paid #1 spot without a visible disclosure misleads buyers, and a list known to be paid adds no consensus. What gets you into the same sources, built on your ANSWER SHARE REPORT of 2026-10-05: CITATION PLAN: Lumen Shifts | 2026-10-05 | Horizon: 90 days | ID | Item | Gap | Play | Prompts | Target | Impact | Effort | Status | Owner | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | AEO-11 | Disclosed founder answers in the 4 Reddit threads the engines cite, one a week | NOT IN SOURCES | PL12 | best small restaurant; cut labor costs | the 4 cited threads | 3 | S | Proposed | you | | AEO-12 | Inclusion request to the other cited independent list, template O1 | NOT IN SOURCES | PL10 | best; alt | hospitalitytechweekly.example/best-restaurant-tools-2026 (restaurantopsreview.example is already AEO-03; orbitrota.example is a competitor's own list, skipped) | 3 | L | Proposed | you | | AEO-13 | Review program: template O3 to every active customer | WEAK PROOF | PL13 | best; use-case | softwarecompare.example profile | 2 | L | Proposed | marketing | | AEO-14 | Paid placement only where the site labels it sponsored, run as a separate test | NOT IN SOURCES | PL10 | best | [confirm which cited sites offer labeled placements] | 1 | M | Proposed | you | Measure: a source is Won when it includes Lumen Shifts with correct facts; a prompt when it is RECOMMENDED on 2 consecutive runs on an engine where it was not before. Next step: approve AEO-11 and I will draft the first answer from template O6, with the disclosure first. You post it from your own account.
Contenu du Kit
Chaque fichier que vous recevez, avec ce qu'il couvre et sa longueur. Les fichiers sont en anglais.
- AEO223 lignes
SKILL.md
Scores how ready a brand is to be recommended by AI answer engines, builds the buyer prompt set to track, reports answer share against competitors and turns every gap into a…
- Citation plays131 lignes
references/citation-plays.md
16 plays, PL01 to PL16, each tied to the gap codes it fixes, with what to ship, the steps, the effort, how to measure it and what not to do.
- Engine notes57 lignes
references/engine-notes.md
How each AI answer engine finds and recommends brands, which crawler it needs, how to check it by hand, and which referral source it leaves in analytics.
- Prompt library71 lignes
references/prompt-library.md
The 7 buyer prompt types, each with what the buyer is really asking, a pattern, an example and the page that wins it, plus the rules for writing and sampling prompts.
- AEO readiness rubric90 lignes
references/readiness-rubric.md
25 checks in 5 dimensions (Crawl access A1 to A5, Schema B1 to B5, Entity clarity C1 to C5, Answer readiness D1 to D5, Citation signals E1 to E5), each scored 20, 10 or 0 points…
- Brand fact sheet101 lignes
assets/brand-fact-sheet.md
The single source of truth for what AI answers should say about the brand: the one-sentence definition, checkable facts with dates, retired facts, official profiles, Organization…
- Outreach templates143 lignes
assets/outreach-templates.md
Seven templates, O1 to O7: a listicle inclusion request, a correction request, a customer review request, a data pitch, a directory listing request, a community answer frame and…
- Worked example: ANSWER SHARE REPORT and CITATION PLAN141 lignes
examples/answer-share-lumen-shifts.md
Four weeks of answers for a brand that AI engines rarely recommend, turned into an ANSWER SHARE REPORT and a 10-item CITATION PLAN with drafts.
À propos de ce Kit
Find out why AI answer engines recommend your competitors instead of you, and what to fix first. The skill scores your brand's readiness on 25 checks, writes the buyer prompts worth tracking, turns pasted answers into recommendation rate and answer share, and ranks every gap into a citation plan with outreach drafts. Built for founders, marketers and agencies who want a brand named in ChatGPT, Perplexity, Google AI Overviews and Gemini answers, not only a page that ranks. Install in 5 minutes; the first readiness audit takes about 20 minutes.
What you get
- Four modes with fixed output. READINESS AUDIT, PROMPT SET, ANSWER SHARE REPORT and CITATION PLAN, with set field names, so results paste straight into a doc or the AEO Tracker sheet.
- A 25-check readiness rubric. Crawl access, Schema, Entity clarity, Answer readiness and Citation signals, 5 checks each, scored 20, 10 or 0 with the evidence that earns the points. A Blocker cap stops a good score hiding a site the engines cannot read.
- One way to count. Rules for what counts as RECOMMENDED versus MENTIONED, positions, and formulas for recommendation rate, top 3 rate and answer share. Prompts that name your brand count for accuracy only, so they never inflate your share.
- A prompt library. 7 prompt types, from
besttobrand, with the page that wins each one, how many prompts to track and the sampling rules that keep weekly runs comparable. - 7 gap codes and 16 citation plays. Every missing recommendation gets one cause, from NOT CRAWLABLE to CATEGORY MISMATCH, and plays with steps, effort, measurement and what not to do.
- Assets. A brand fact sheet with Organization JSON-LD and a profile consistency table, and 7 outreach templates: listicle inclusion, correction, review request, data pitch, directory listing, community answer and the one follow-up.
- A worked example. Four weeks of answers for a fictional restaurant scheduling app traced to two wrong facts, missing sources and a 10-item plan.
- Code audits in Codex and Cursor. It reads your robots rules, metadata and JSON-LD and cites the file and line.
Works in Claude, ChatGPT, Codex and Cursor.
Who it is for
- Founders and marketers who hear "ChatGPT recommended your competitor" from prospects and want the reason and the fix.
- Local businesses, ecommerce brands and B2B software companies whose buyers now ask an assistant before they search.
- Agencies and consultants who want the same audit, metrics and plan format for every client.
- Teams already using the SEO/GEO Skill for pages, who need the brand-level layer: prompts, competitors and the sources AI answers trust.
What it will not do
- Sample answers itself inside the conversation. It already knows your brand, so its own answer is not what a buyer sees. It works from answers you paste or the AEO Bot's tracker.
- Promise recommendations, citations or traffic. It removes the reasons you are not recommended and defines how to measure the result.
- Invent review counts, prices, statistics or customer names. Unknowns are written
[confirm]. - Write sock-puppet posts, fake or incentivized-for-positive reviews, hidden text for AI systems, or posts on other sites that borrow their ranking. It says why and offers the play that works.
- Send any email or post anything. Outreach comes back as drafts for you to send.
Avis
Aucun avis pour l'instant.
Autres Kits SEO / GEO
AEO Kit
Become the brand ChatGPT, Perplexity and Google name when buyers ask who to choose.
SEO/GEO Kit
Pick the right topics, rank in search, get cited by AI answers, and stay fresh.
AEO Dot
Turns your AEO plan into drafts, follow-ups and a weekly report of what moved.