What Your AI Visibility Report Shows After 30, 60 and 90 Days
A walk-through of a three-month pilot for a fictional insurer, from the baseline diagnostic to the third re-measurement. It shows what each cycle produces and how Starter and Growth outputs differ.
The example pilot
Example Insurer, a fictional regional auto and home carrier, runs a Growth pilot: 100 prompts across four categories (auto, home, bundled, claims experience), 3 AI engines, and four named competitors. The engines, prompts and competitors are fixed at the baseline and stay the same every cycle, so results are comparable.
| Item | Example setting (illustrative) |
|---|---|
| Tier | Growth: 100 prompts, 3 engines, fixed fee |
| Engines | Three engines chosen for the category |
| Competitors | Four named regional and national carriers |
| Cycle | Baseline diagnostic, then re-measurement at roughly 30, 60 and 90 days |
Scorecard across the three months (illustrative)
| Measure | Baseline | Day 30 | Day 60 | Day 90 |
|---|---|---|---|---|
| Mention rate (responses naming the brand) | 22% | 23% | 27% | 29% |
| Share of model vs named competitors | 11% | 11% | 13% | 14% |
| Recommendation rate (of brand mentions) | 38% | 40% | 41% | 44% |
| Mentions backed by a cited source | 31% | 30% | 36% | 37% |
| Open accuracy errors (confirmed) | 6 | 6 | 4 | 3 |
Illustrative data for a fictional brand. Not a client result, a benchmark figure or a forecast.
What the team changed, and when
Your team records each change and its launch date. The report sets those dates next to the results, so you can see whether movement follows a change or was already under way.
| Date (illustrative) | Change | Owner | What the next cycle showed |
|---|---|---|---|
| Day 12 | Corrected the outdated discount and state-availability details on the website and business profiles | Web team | Two of six confirmed errors no longer reproduced on repeat prompts by day 60 |
| Day 25 | Published a bundled home and auto comparison page with a direct opening answer | Content team | Bundled-category mention rate rose from day 60; consistent with the change, not proof of cause |
| Day 48 | Agency secured an updated listing on two independent comparison sites | Agency | More cited mentions by day 90 on the engines that cite sources |
What each month delivers, by tier
| Starter | Growth | |
|---|---|---|
| Weeks 1–2 (baseline) | Multi-Engine Diagnostic on 50 prompts, 3 engines: baseline, competitor gaps, initial accuracy review, prioritized recommendations | The same diagnostic on 100 prompts |
| Day 30 and day 60 | Re-measurement and share-of-model scorecard against the baseline | Scorecard plus new prioritized findings and recommended actions each cycle |
| Day 90 | Final scorecard against the baseline; tracks the diagnostic’s recommendations | Final scorecard, findings and a written summary of open issues and next actions |
| New recommendations after the diagnostic | No | Yes, every cycle |
| Accuracy follow-up | Confirmed errors re-tested with repeat prompts | Same, plus newly found errors added to the log |
Enterprise pilots add a quarterly deep-dive diagnostic, an AI traffic attribution review and an executive readout. Full tier detail is on the pricing page.
Reading the results honestly
- Small changes can be noise. A one-point move on 100 prompts is within normal run-to-run variation; look for movement that holds across cycles and engines. See how many prompts are enough.
- Engines change too. Model and retrieval updates between cycles can move every brand at once, including competitors. Competitor movement on the same prompts helps separate your change from a market-wide shift.
- Consistent with is not caused by. The report says when movement follows a dated change; it does not claim the change caused it.
- What’s still open matters as much as what moved: the day-90 summary lists remaining gaps and errors for your team or a renewal.
See this report for your own brand
Every pilot starts with a Multi-Engine Diagnostic and re-measures on the same prompts each cycle for three months.

