Loading page…
Audit whether clients “exist” in ChatGPT/Gemini's recommended answers, deliver a report + fixes, bill monthly
client went from invisible to recommended in 8 weeks
“People ask ChatGPT for recommendations every day — if you're not in the answer, you're losing business, and you don't even know it.” That line alone is the closing pitch.
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
[Video] AEO = Answer Engine Optimization. The service: test whether a client's business “exists” in the recommended answers from ChatGPT/Gemini/Claude/Google — run “best XYZ in XYZ”-style queries and see whether the client is mentioned, recommended, or completely invisible; deliver an audit report + a remediation playbook. One client pays a $2,000/mo retainer. The result: within 8 weeks the client went from completely invisible in ChatGPT to being recommended as a top option. Built with Claude Code, with multiple agents each owning a slice of the audit.
One client at $2,000 a month means this business has proven its pitch rather than its pipeline. The pitch is unusual because it can be run before anyone agrees to a meeting: feed the system a business name, run "what's the best XYZ in XYZ" across ChatGPT, Gemini, Google and Claude, and the output is either the client's name or an empty space where it should be. That audit costs a handful of API calls and lands as a fact about the prospect rather than a claim about the seller. For a solo operator selling to clinics, restaurants, online stores and service businesses — the same buyer set the AI sales agent already serves — the fastest route from one client to ten is to run the audit unasked across fifty businesses in a single category, send each one the answer they are missing from, and name the competitor who is in it. The opening line is already written: people ask ChatGPT for recommendations every day, and if you are not in the answer you are losing business and do not know it.
The first money was $2,000 a month, billed as a retainer rather than a project fee, which is the right shape for this deliverable because a position in an AI answer is not something you fix once. What was sold is an audit plus a playbook: the system runs the queries, identifies exactly what is wrong, and produces the list of fixes. The proof that justifies the second invoice arrived inside two months, with the client going from completely invisible in ChatGPT to recommended as a top option in 8 weeks. Pricing lands at this level because the buyer measures it against an SEO retainer rather than a software subscription — $2,000 a month is an ordinary agency line item and small against one recovered patient, booking or CRM seat. The first sale in this category almost always follows a free audit, because the audit converts a vague worry into a specific absence the buyer can see. An 8-week window with a named before-and-after is the natural opening offer: short enough to buy on a hunch, long enough for movement to show.
The model closest to your capabilities: you have Claude Code, understand SEO logic, and can build tool sites. The classic “charge for the service first → productize” path, right in the middle of the AEO awareness dividend.
Paste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself. Timestamps jump to the second each figure is said.
Members read the exact words said here
7:38 · hear it in the videoThe exact line behind each figure is in the Unicorn Club. Read the quote, then jump to the second it's said. Join the Unicorn Club
Data credibility: Creator's first-hand account of his own projects. All revenue figures are founder-reported or estimated by the video creator — unaudited.
Two loops turn here. The first sits inside delivery: the whole audit runs on multiple Claude Code agents, each handling a different part, so auditing the eleventh client costs roughly what auditing the first one cost. The report is sold, the machine is kept. The second is the before-and-after itself. "Completely invisible to recommended as a top option in 8 weeks" becomes a case study the moment it exists, and in a category with no established benchmarks, the first documented result carries the entire pitch. The point where this stops being consulting is when the audit goes self-serve — enter a domain, get a score and a fix list — because the free audit that was a sales tool becomes a funnel that runs unattended, and the $2,000 retainer turns into the upgrade for clients who want the fixes done rather than listed. There is a second lever available in the existing book: an operator already running an AI sales agent for clinics, restaurants and online stores worldwide is one offer away from a second revenue line inside accounts that already reply.
The no-feedback risk here is not silence from the market, it is silence from the metric. Answer engine optimization has no rank tracker with a decade of history behind it, and the answer a model gives shifts between runs, so proving a fix worked means running the same query matrix against the same prompts repeatedly until the pattern is unarguable. That is part of why the 8-week frame matters: it is roughly how long new citable pages and structured data need to be crawled, indexed and start surfacing in answers, and anything measured inside two weeks reads as noise. The instrument that makes the wait tolerable is running the audit on the client weekly and charting mentions rather than revenue, because a business moving from zero mentions to intermittent ones across four engines is progress even while the phone stays quiet. The signal to change approach is the opposite pattern — fixes shipped, pages indexed, still absent after two months — which usually means the model's answer for that query is anchored to third-party directories, review sites and publications, so the work belongs off the client's own site.
[Video, verbatim] “People ask ChatGPT for recommendations every day — best restaurant near me, best CRM, best clinic. If you're not in the answer, you're losing business, and you don't know it.” [AI cross-check] Two independent sources point to the same trend: Outrank (Tibo) says “show up in search results and ChatGPT will recommend you”; MeetOscar is already seeing ChatGPT-referred traffic growing in its UTMs and optimizes for it specifically. Awareness on the business side lags far behind — and that lag is the information-gap window.
[Video] (1) An AI sales/support agent SaaS (hooks into WhatsApp/IG/FB, handles objections, remembers history, actually closes deals): $100K/yr; clients are clinics/restaurants/e-commerce worldwide; tech path GoHighLevel + Zapier → FlutterFlow → Cursor → Claude Code (“features that used to take weeks now ship in days”). (2) A government NFC access-control system contract: $50K single contract, Claude Code built 90% in about 2 months. (3) A deep-research cold-email agent: every paragraph of the email custom-written for the target company, built in 1-2 days. (4) A multi-agent automated dev pipeline: triage/investigate/fix/notify/review — “a 3 a.m. bug is fixed by the time I wake up, the client has already gotten a WhatsApp notification, and I just do QA.” His line: “Whatever you think Claude Code can do, multiply it by 10.”
[AI] Of all 26 cases, this is the one closest to your capabilities: you have Claude Code, you understand SEO/AEO, and you can spin up tool sites fast. The path: (1) build the audit tool (query matrix × target market → competitor comparison → automated report); (2) use free audit reports as the sales weapon, starting at a $1-2K/mo retainer; (3) after 3-5 clients, productize into a self-serve SaaS (enter a domain → AEO score + fix list). Caveats: service revenue isn't an asset, so productize early; single-client evidence is weak, but the model's logic is corroborated by multiple sources.
That was one of the 50 breakdowns open to everyone. The Unicorn Club opens the other 1,285 at the same depth — plus pain-point filters and 10 build specs a week.
Join the Unicorn Club — $5/moCancel anytime, in one click