Case studies

No vibes. Just numbers.

Every study lists the before/after, what we did, and how to verify it. Data comes from GA4, Search Console, Lighthouse and ad-platform dashboards — raw exports available under client authorisation.

Featured · Hong Kong e-commerce brand

Six months: ROAS from 2.1 to 5.3

A Hong Kong home-goods e-commerce brand: slow legacy site, stalled rankings, rising ad costs. We rebuilt on Next.js, fixed Core Web Vitals, and restructured SEO/GEO content and campaigns.

Before → after

Lighthouse performance score
4192
Largest Contentful Paint (LCP)
4.8s1.6s
Monthly organic clicks
baseline+168%
AI-search citations (of 10 category prompts)
03
Return on ad spend (ROAS)
2.15.3

What we did

  • Next.js rebuild: server rendering, image optimisation, script diet — LCP from 4.8s to 1.6s.
  • Technical SEO: structured data, hreflang, internal-link restructure — organic clicks +168% in six months.
  • GEO content: FAQ and comparison pages rewritten for AI search — cited in 3 of 10 category prompts (was 0).
  • Campaign restructure: ad groups by product margin, negative-keyword cleanup — ROAS from 2.1 to 5.3.

How to verify

Figures above come from GA4, Google Search Console, Lighthouse reports and ad-platform dashboards. The client is anonymised under NDA; prospective clients can request the raw records in a meeting.

Want to be the next case study?

Start with a free SEO/GEO audit and get baseline numbers for your site — improvement starts with measurement.

Free SEO audit