Montaic had 222 URLs in its sitemap, a working product, and paying customers. The site structure was solid. But when we ran the audit, the search foundation was missing. Not broken. Missing.
The pricing page title was 17 characters. Canonical tags were missing on every key page type. Google was conflating "Montaic" with "Monti," "Monte," and "Montana" in SERPs. The brand had no entity recognition. And across ChatGPT, Perplexity, Gemini, and Google AIO, Montaic appeared in zero responses to the eight queries their customers actually ask.
Meanwhile, a competitor called ListingAI was cited in all eight.
We ran the standard six-deliverable audit, produced the 90-day roadmap, and started executing the same week. Two phases: foundation first, then content.
For the fourth pillar article, we ran the same 53 Nashville listings through four different writing paths and graded every output on the same rubric. This is the data that became the centerpiece of the content strategy.
| Path | Avg Score | 7-8 Range | Delta vs Originals |
|---|---|---|---|
| ChatGPT gpt-4o-mini | 3.6 | 0 of 53 | -1.0 |
| Zillow originals (agent-written) | 4.6 | 1 of 53 | -- |
| Montaic default path | 5.1 | 3 of 53 | +0.5 |
| Montaic voice-locked | 6.9 | 42 of 53 | +2.3 |
The biggest gain was not specificity. It was cliche avoidance. The voice calibration layer, trained on just five writing samples, eliminated every major filler phrase: zero "stunning," zero "dream home," zero "hidden gem." The writing sample taught the model what to skip more than what to say.
Category-level lift over default: cliche avoidance +3.3, emotional appeal +2.6, structure and flow +2.0, specificity +1.8.
Five hundred dollars. Six deliverables. Forty-eight hours. Yours to keep whether you hire us after or not.
Start with the audit