Cadence, a B2B CRM, went from appearing in 12% of AI answers to 78% in a 30-day sprint by shipping answer-first comparison pages, aligning entity descriptions across the web, and re-testing prompts weekly across the four major models.
- 12% → 78% — AI visibility score
- 3.2× — mention rate
- 4/5 — answers recommending
We watched the models go from never naming us to recommending us first. Pipeline followed within the quarter.
The starting point
Cadence, a B2B CRM for early-stage startups, ranked respectably in classic search but was invisible in AI answers. Across 120 buyer-intent queries, the brand surfaced in 12% of answers — competitors owned the shortlist.
What we did
A 30-day sprint: mapped the query space, shipped answer-first comparison and "alternatives to" pages, aligned entity descriptions across every profile and directory, and re-tested across ChatGPT, Perplexity, Gemini, and Claude on a weekly cadence.
The result
By day 30, Cadence appeared in 78% of tracked answers and was actively recommended — not just mentioned — in 4 out of 5 shortlists for its core query set.
Want your own number?
A free thirty-minute consult ends with a written map: your zones, your unit counts, and a total. No obligation to book anything.