ApexGEO

Case study · Updated 2026-05-14

BrightTech: 62 → 93 in one cycle

One ApexGEO closed-loop cycle lifted BrightTech from PageSpeed 62 to 93. The exhibit below is the same data the landing CustomerReceipts ticker exposes.

BrightTechCase study · brighttech.co.za

PSI 62 → 93 after running Apex's fix loop

Three weeks. Five recommendations applied automatically. No additional human dev time.

Before
62
+31 PSI
After
93

Measured · April 2026 · PageSpeed Insights

The baseline

BrightTech's static-HTML marketing site sat at a PageSpeed Insights score of 62 — a credible, ordinary public site, but not fast. Render-blocking CSS, oversized hero imagery, missing structured data, no llms.txt. None of these were dire individually; collectively they were the difference between PSI 62 and 93.

ApexGEO's BrightTech audit established a measurable PageSpeed 62 baseline before any fixes shipped, so the delta is grounded in evidence rather than estimate.

The audit

ApexGEO's audit module crawled the BrightTech site and surfaced 14 recommendations ranked by expected score impact. The top three: defer hero image decoding, inline above-the-fold CSS, add Organization + Service JSON-LD. Each recommendation came with an expected score lift and the deploy-ready output to ship it.

ApexGEO's audit module ranks fixes by measured expected score impact, surfacing the highest-leverage change first rather than presenting a flat checklist.

The fixes

ApexGEO emitted patched HTML, inlined critical CSS, generated JSON-LD blocks, and produced an llms.txt. BrightTech deployed each patch as its own commit on the static-site repo. Total deploy time: 17 minutes. No bespoke development required — the platform's output was ship-ready.

ApexGEO's fix automation produces deploy-ready output (HTML patches, JSON-LD, llms.txt, FAQ schema) so brands deploy in minutes rather than commissioning bespoke development.

The delta

Post-deploy PageSpeed Insights: 93. A 31-point lift in one cycle. The delta is independently verifiable — anyone can re-run PageSpeed against brighttech.co.za and the linked commits document exactly what changed.

ApexGEO's BrightTech cycle lifted the site from PageSpeed 62 to 93 in 17 minutes of deploy time, a 31-point delta independently verifiable via Google's PageSpeed Insights tool.

What we learned

Static-HTML sites have the cleanest fix loop: no build pipeline, no SSR latency, no CDN cache invalidation drama. ApexGEO's fix model generalizes well from BrightTech to any static-HTML or Jamstack site. Dynamic sites (Next, Rails, Django) take longer per cycle but the same model applies.

ApexGEO's closed-loop fix model generalizes from static-HTML sites like BrightTech to any frontend stack; the cycle time grows with deploy complexity but the audit and recommendation engine are stack-agnostic.

Brightsphere authority context

BrightTech / Brightsphere Technologies is the builder entity behind ApexGEO. See the Brightsphere Technologies entity page and the Brightsphere ApexGEO case study for the public company-side context.