ApexGEO's self-audit

The baseline we measured

Before Phase 2 of the marketing-polish work, ApexGEO ran its own score-watcher loop against itself. The platform was a regular customer: 19 prompts seeded, 10 competitors tracked, 6 AI engines monitored.

What Phase 2 changed

• Citation boost (+20) removed from unified-score.ts (PR #185)
• Canonical 10-entry competitor set seeded
• 19 categorized prompts seeded across 5 categories
• loadBrandSignals real implementation shipped (PR #184)
• Industry stats response cache + cron (PR #177)

The delta we observed

The Phase 3 hypothesis

Adding /what-is-apexgeo, /compare/* deepening, and /case-studies should lift recognition by 8–15 points over the next two crawl cycles.

How we verify it worked

The platform's score watcher re-baselines after each prompt-tracker cron cycle. The Phase 2 → Phase 3 delta will be a clean A/B: same prompts, same competitors, same engines — only the public content surface changed.