ApexGEO

Case study · Updated 2026-05-12

ApexGEO's self-audit

The platform measures itself. Real baseline, real Phase 2 fixes, real delta. Snapshot dated 2026-05-12.

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, 7 AI engines monitored.

ApexGEO ran its own score-watcher loop in 2026-05 to validate the closed-loop fix model on its own platform — eating its own dog food before asking customers to.

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)
ApexGEO's Phase 2 shipped five concrete fixes including the removal of a +20 citation-count uplift bug, a canonical 10-entry competitor set, and a per-prompt category dimension.

The delta we observed

AI Visibility score60+8
Visibility88+2
Recommendation22+2
SoM21%+3pp
ApexGEO's AI Visibility score moved from 52 to 60 over the two crawl cycles following Phase 2 — an 8-point lift attributed to the citation-boost removal and competitor-set canonicalization.

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.

ApexGEO's Phase 3 hypothesis is that pillar pages, deepened competitor comparisons, and case-study URLs should lift recognition by 8–15 points over two more crawl cycles. The next score-watcher re-baseline tests this directly.

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.

ApexGEO verifies recommendation impact by re-running the same prompts against the same competitor set after the change ships. Anything that doesn't measurably move the score gets rolled back or refined.