| Metric | Month 2 | Month 3 | Change |
|---|---|---|---|
| Overall AI visibility | 9.0% | 14.1% | ▲ +5.1 pts |
| Bootstrap 95% CI | 3.5–15.8% | 6.5–22.0% | — |
| Rank among brands | #8 of 14 | #6 of 14 | ▲ 2 places |
| Mentions / runs | 71 / 792 | 122 / 864 | — |
| Iterations per prompt | 6 | 6 | — |
Top of the field, both scans. Brightway moved up past Anchor and Greenline while the leaders held station.
The gain wasn't carried by one model. Every assistant moved up, with the biggest single jump on ChatGPT, where Brightway had been weakest.
| Model | Month 2 | Month 3 | Change |
|---|---|---|---|
| ChatGPT | 11% | 18% | ▲ +7 |
| Claude | 12% | 16% | ▲ +4 |
| Perplexity | 8% | 13% | ▲ +5 |
| Google AI Overview | 5% | 9% | ▲ +4 |
| Overall | 9% | 14% | ▲ +5 |
Each cell is mentions out of iterations and the probability of appearing in an AI answer. Rows are shown for both scans so gains and the occasional dip are both visible.
| Prompt | Model · location | M2 | M2 prob | M3 | M3 prob |
|---|---|---|---|---|---|
| Who installs water heaters in Austin | |||||
| ChatGPT · Austin | 3/6 | 50% | 5/6 | 83% | |
| Perplexity · Austin | 2/6 | 33% | 4/6 | 67% | |
| Best plumber near me | |||||
| Claude · Austin | 2/6 | 33% | 4/6 | 67% | |
| Drain cleaning service near me | |||||
| Claude · Round Rock | 3/6 | 50% | 4/6 | 67% | |
| Emergency plumber near me | |||||
| ChatGPT · Austin | 1/6 | 17% | 3/6 | 50% | |
| 24/7 plumber Austin | |||||
| ChatGPT · Austin | 2/6 | 33% | 4/6 | 67% | |
| Repipe specialist near me | |||||
| Perplexity · Cedar Park | 3/6 | 50% | 3/6 | 50% | |
| Affordable plumber Austin | |||||
| Google AI · Austin | 2/6 | 33% | 1/6 | 17% | |
Short answer: the direction is encouraging, but treat it as a signal, not a confirmed result. Each scan issues a limited number of queries per model, so only large swings clear the bar. Testing the month-2-to-3 difference with the same query-level bootstrap used for the confidence intervals, every change for Brightway has a 95% interval that still crosses zero. What's notable is that all four models moved the same way, consistent direction across independent measurements is itself worth weight, even before any single move is individually significant.
| Scope | M2 | M3 | Δ | 95% CI of Δ | Verdict |
|---|---|---|---|---|---|
| Overall | 9% | 14% | +5 | [-6, +16] | Not yet significant |
| ChatGPT | 11% | 18% | +7 | [-9, +23] | Not significant |
| Claude | 12% | 16% | +4 | [-12, +20] | Not significant |
| Perplexity | 8% | 13% | +5 | [-11, +21] | Not significant |
| Google AI | 5% | 9% | +4 | [-9, +17] | Not significant |
The arrow reflects statistical significance, not the size of the raw move: ▲ / ▼ mark changes whose 95% interval excludes zero; ≈ marks changes that can't be told apart from noise even when the percentage moved.
| Brand | M2 | M3 | Trend |
|---|---|---|---|
| Brightway Plumbing & Drain | 9% | 14% | ≈ no significant change (Δ +5; CI crosses 0) |
| Reliant Plumbing | 47% | 45% | ≈ Δ -2; CI [-18,+13] |
| Lonestar Drain Co. | 43% | 41% | ≈ Δ -2; CI [-19,+14] |
| Hill Country Plumbing | 37% | 35% | ≈ Δ -2; CI [-18,+13] |
| Capital Plumbing Pros | 26% | 24% | ≈ Δ -2; CI [-16,+11] |
| Bluebonnet Plumbing | 20% | 18% | ≈ Δ -2; CI [-15,+10] |
| Anchor Plumbing Services | 14% | 13% | ≈ Δ -1; CI [-13,+11] |
| Cedar Park Plumbers | 11% | 6% | ▼ significant decline (Δ -5; CI [-9,-1]) |