Glow & Co.
Conversion 1.6% → 4.2%
Glow & Co. was paying for traffic that bounced. We rebuilt the hero, ran 11 A/B tests across the PDP, and collapsed checkout to two steps — taking conversion from 1.6% to 4.2% on the same ad spend.
What was broken
Glow & Co. had spent two years optimising acquisition and almost nothing on the on-site funnel. Paid CAC was creeping up every quarter, the PDP had four competing CTAs, and the checkout asked for an account before payment. Bounce on mobile was 71% and add-to-cart from PDP was a flat 3.4% — even on returning visitors. The founders were considering pausing Meta ads entirely.
What we did
- 1
Ran a 5-day teardown: session replays, funnel maths, heatmaps, and a heuristic review of the top 6 PDPs. Found 14 leaks, ranked by expected revenue impact and dev cost.
- 2
Rebuilt the hero around a single value-prop and a one-shot CTA. Killed the carousel — replaced it with a static lifestyle shot and a benefits strip above the fold.
- 3
Ran 11 A/B tests on the PDP across 6 weeks: review placement, sticky ATC, size selector, social proof copy, urgency framing, and bundle pricing. Shipped only winners (statistical significance ≥95%).
- 4
Collapsed checkout from 4 steps to 2 — guest by default, Shop Pay first, address autofill, and a single shipping rate above the free-ship threshold.
- 5
Wired GA4 + server-side events so Meta and Klaviyo finally saw the same conversion numbers.
How it unfolded
Funnel audit, 14 leaks ranked, sprint plan signed off.
Hero rebuild and IA cleanup shipped to 50% of traffic.
11 A/B tests live in parallel. 6 winners shipped.
2-step checkout, Shop Pay default, address autofill.
Conversion stabilised at 4.2%. Same ad spend, 2.6× revenue.
“We thought we needed more traffic. Turns out we needed to stop leaking the traffic we had. PromoGent paid for itself in 18 days.”