This is a documented walkthrough of how we took a service business from zero AI visibility to being the #1 recommendation on ChatGPT for its primary target phrase in seven days. Every fix is listed. The timeline is exact. We've anonymised the client — the specifics below describe a vehicle valuation business based in New Zealand — but every step is reproducible on any client site your agency runs.
The point of this post is not to claim a magic formula. The point is to show that AI visibility moves fast when you work the right list in the right order. If your agency is considering offering AEO as a service and you want a template to plan the first engagement around, this is it.
The starting point
The business was a vehicle valuation service operating in one country. Good website, clean design, reasonable SEO foundations — ranking well on Google for several of its target keywords. The team had been running SEO for years and were comfortable with their organic channel. They came to us because they had started noticing a trend they couldn't explain: traffic was soft despite stable rankings, and direct enquiries had shifted in tone. More prospects were arriving saying things like "ChatGPT suggested I look at you" — or, worse, arriving with a competitor's name already in mind because that's who ChatGPT had mentioned first.
The symptoms are what we cover in our post on why traffic is dropping even when SEO isn't broken. Good Google performance, shifting customer behaviour, AI agents recommending someone else. The diagnosis is always the same: ranking on Google and being recommended by ChatGPT are no longer the same thing.
Day 1: The audit
We ran a full AI visibility scan on day one. The output was stark.
Visibility score: 0/100. Across eight target queries that customers actually ask when looking for a vehicle valuation service, ChatGPT didn't mention the business a single time. Not in the primary answer, not as a runner-up, not as an honourable mention. The business was invisible.
Competitors being recommended instead. Every one of the eight queries returned a ranked recommendation. The winners were predictable — two direct competitors the client already knew and was actively trying to beat on Google, plus one adjacent-category player the client had never considered a competitor but which was being presented by ChatGPT as a valid answer. The AI was making choices the client had no visibility into.
The specific issues surfaced by the audit. The tool gave us a prioritised diagnostic list. Paraphrased, it said: missing structured data at the site level (no LocalBusiness or Service JSON-LD); no FAQ schema anywhere on the site; meta descriptions written as marketing taglines rather than clear business descriptions; service-page copy that led with positioning rather than with what the business actually did, who it served, and where; a thin third-party presence — the business had a Google Business Profile but several of the fields were incomplete, and there was no presence on the directories ChatGPT was citing for competitors.
This is the classic profile. A well-run site that reads beautifully to humans and says almost nothing concrete to an AI agent. If you're about to run this process on a client, assume you'll see a version of the same list on the first scan.
Day 2–3: Structured data and schema fixes
The first wave of fixes was structured data. The reason to start here is that schema gives AI agents explicit, machine-parseable information about the business — and the lift is concentrated. A few hours of developer time translates directly into a measurable change in how AI agents perceive the site.
LocalBusiness JSON-LD. We added a LocalBusiness schema block to the homepage and to every service page, with the explicit properties AI agents care about: business name, address, phone, service area, opening hours, a description field that described in one paragraph exactly what the business did and who it served, and individual Service entries for each service offered with a clear description of each. Nothing clever — just complete, accurate, machine-readable facts.
FAQ schema. We wrote twelve FAQ entries and wrapped them in FAQPage schema on the most-visited pages. The questions were not generic "what is a vehicle valuation" filler. They were the exact phrases customers asked when talking to ChatGPT — the phrasing the scan had surfaced as high-intent target queries. Each answer was 60–120 words, direct, and named the specifics an AI agent would look for: service area, turnaround time, pricing indicators, specialisations, differentiators.
Meta description rewrites. The meta descriptions on the key pages read like taglines ("Trusted vehicle valuations. Accurate. Fast."). We rewrote every one of them into a flat, factual summary an AI agent could parse: who the business is, what they do, for whom, where, and what distinguishes them. No adjectives that didn't carry information. No marketing rhythm. Just the facts, in the order an AI would look for them.
By the end of day 3 the schema and metadata layer was in place. We ran a partial rescan to sanity-check the markup was valid and being picked up. It was.
Day 4–5: Content improvements
Structured data tells AI agents what the business is. Content tells them why it's the right answer. The second wave of fixes targeted the pages customers actually land on.
Rewrote key service pages. The two most important service pages had opened with brand positioning. We reordered them. New structure: a one-paragraph summary at the top that stated what the service was, who it was for, how it worked, and how to access it — the information an AI agent would extract to answer a direct question. The positioning copy stayed on the page but was pushed below the informational block. Humans still get the brand. AI agents get the facts first.
Added the specifics AI agents need. Every service page got explicit sections for: service areas (named regions, not "across the country"); pricing indicators (ranges or "from $X" rather than "competitive pricing"); specialisations (specific vehicle categories or valuation types rather than "all vehicles"); turnaround expectations; and what distinguished the business from the two direct competitors — expressed as features and facts rather than as adjectives.
Stated the obvious in every page template. Every page header now stated, within the first paragraph, what the business did, who it served, and where. This feels redundant on a page the customer navigated to on purpose. It is not redundant to an AI agent crawling the site as a standalone document.
The content work was the longest part of the seven days. Not because it was technically hard — it was mostly rewriting — but because the editorial judgement about what to cut and what to lead with required a pass through every major page.
Day 6: Third-party presence
On day 6 we turned to the off-site footprint. AI agents cross-reference multiple sources. A business that shows up consistently across its own website, its Google Business Profile, industry directories, and review sites reads as more credible than a business that only exists on one canonical URL.
Google Business Profile. We completed every empty field on the profile — business description, service categories, service area, hours, attributes. We uploaded a fresh set of photos. We wrote the profile description with the same discipline as the meta descriptions on the site: factual, machine-parseable, specific about service area and specialisations.
Directory consistency. We audited every directory the business was listed on and fixed inconsistencies — mismatched business names, outdated phone numbers, different service areas described in different ways across sites. Consistency matters because AI agents treat divergent information as a credibility signal against the business.
New third-party presence. The scan had surfaced the platforms AI was citing when recommending competitors. Two of those platforms — a specialist industry directory and a review site — the client had no presence on. We created profiles on both, populated them with the same information structure as the rest of the footprint, and ensured the language matched what was on the site.
Day 6 work was clerical more than creative. But it closed the last credibility gap AI agents were picking up on.
Day 7: Rescan and results
On day 7 we ran a full rescan. The output was a different business.
Visibility score: from 0 to the high 70s. Every one of the five scored dimensions — structured data, content clarity, authority signals, AI readability, off-site presence — had moved substantially. Structured data and content clarity saw the biggest jumps; off-site presence lagged slightly because third-party signals take a few more days to fully propagate.
Mentioned in the majority of target queries. Of the eight target queries, the business was now recommended on six. Two were still won by entrenched competitors — a legacy of their longer-standing review footprint that a seven-day sprint couldn't fully close.
#1 recommendation on the primary query. For the single highest-value query — the phrase the client had identified as their primary commercial target — the business was now the top recommendation on ChatGPT. The competitor who had held that position on day 1 had dropped to third.
Competitor leaderboard flip. The competitor leaderboard had inverted. On day 1 the client was nowhere on any query. On day 7 the client was in the top three on six queries, and the two direct competitors were now competing for the spots the client wasn't already holding.
What made the difference
Seven days is fast. It works because the fixes compound and because AI agents re-evaluate quickly. Looking back at the engagement, four things drove most of the impact.
Structured data was the single biggest lever. Adding LocalBusiness and Service JSON-LD with complete, explicit fields took the site from invisible to parseable. AI agents don't guess at what a business does — they read the signals. Explicit beats implicit every time, and structured data is the most explicit signal available.
FAQ schema had outsized effect. Because AI agents are answering questions, content that's already structured as question-and-answer is unusually easy for them to use. FAQ schema is also forgiving to write — you don't need to predict every phrasing of every question, you need to cover the most common forms. The ROI per hour of writing time is high.
Third-party presence confirmed legitimacy. AI agents cross-reference. A business that exists consistently across its own site, its Google Business Profile, and a handful of external directories reads as a real business. One that only exists on its own URL reads as less certain. Closing the third-party gap wasn't a huge amount of work, but it was load-bearing.
None of the fixes were expensive or exotic. Every fix on the seven-day list was something an in-house agency team could ship without specialist hires. Schema markup, metadata rewrites, content reordering, directory listings, Google Business Profile completion. The total developer time involved was in the low single-digit hours. The leverage came from doing the right things in the right order, not from any one fix being particularly clever.
The takeaway for agencies
The point of documenting this engagement is not to show off the result. The point is to show that the process is repeatable. Every client site your agency runs can go through the same audit-fix-monitor cycle. The specific fixes vary — a Shopify merchant's list looks different to a local service business — but the workflow is consistent.
This is a repeatable process, not a one-off win. Agencies that treat AEO as a service line rather than as a special project get compounding returns. The first engagement teaches the team the workflow. By the fifth client you've systematised the fix implementation. By the twentieth you're running a production line.
The results are fast and visible. Unlike SEO, where clients might wait three to six months to see keyword rankings move, AI visibility scores shift within days of landing the fixes. Clients can see their score climb in a quarterly review, which makes retainer renewals a conversation about trajectory rather than price.
The margin is high. Tool cost on an agency plan is roughly $10 per site. Client pricing for AI visibility monitoring and reporting typically runs $200–$500/month. The fixes take hours, the monitoring is automated, the reports are white-label. The retainer is justified by the score movement and by the ongoing competitive intelligence. If you want the broader economics, we laid them out in the agency tracking guide.
Every client has a version of this report waiting for them. Most agency clients are in the same shape as the business in this engagement on day 1. Good website, reasonable SEO, and no AI visibility. The first scan is where the conversation starts.
Your next move
The fastest way to see whether this engagement is replicable on your clients is to run a baseline scan on one of them. Pick the client whose quarterly review is coming up, or the one who's already asking about AI visibility. Run a free scan, read the report, and see whether the fix list looks like something your team could ship in a week. In our experience the answer is always yes — the question is which five clients you run it on first.
If you're evaluating the tooling, our comparison of GEO and AEO platforms walks through what every tool on the market does and what it costs. AgenticLens is the one we built for this workflow — purpose-built for agency AEO delivery with white-label reports, per-site pricing, and the diagnostic depth that makes seven-day turnarounds like this one reproducible.
Run a free scan on a client site at agenticlens.io — then plug it into your agency workflow and ship the first fix list this week.
See how AI agents see your business.
Run a free GEO scan. No account required. Results in under a minute.
How Agencies Can Track Client AI Visibility (Without Building Anything)
Your clients are asking if ChatGPT recommends them. Here's how to add AI visibility monitoring to your agency's service offering using existing tools — no development required.
The Agency's Guide to AEO: Adding Answer Engine Optimization to Your Services
AEO is the next service your clients will pay for. Here's what it is, why it matters, and how to start offering AI visibility monitoring as a service today.