Redpanda: A/B Testing Program That Lifted Sign-ups 22%
I run the A/B testing program for Redpanda's website, a B2B SaaS data streaming platform. One homepage test lifted Cloud sign-up conversion 22%, and it changed how the team weighs content downloads against product sign-ups.
The homepage asked visitors to do everything at once
Developers, architects and buyers all landed on the same homepage, and sign-ups, sales forms and content downloads competed for the same attention.
The homepage is the top landing page on the site, so it carries more of the funnel than any other page. Visitors who were ready to try the product had no clear first step.
Rank the ideas, ship the clear fixes, test the big bet
I ranked ideas by expected impact and effort. Obvious fixes shipped straight away. The homepage itself was a bigger bet, so it got a proper A/B test.
- Hypothesis: a clearer path for visitors ready to try the product would lift sign-ups without hurting sales forms.
- Built the variant in Webflow.
- Primary metric: Cloud sign-up conversion. Guardrails: Tier-1 sales forms, Tier-2 content forms and bounce rate.
- Ran it to 6,622 visitors before calling it.
+22% sign-up conversion, and a trade-off worth making
The variant lifted Cloud sign-up conversion 22% and Tier-1 sales forms 14%, with a 5% better bounce rate.
Tier-2 content downloads fell 26%. We took that trade on purpose: a sign-up is someone using the product, while a download is a lead that still has to be nurtured.
What it led to
- Moving Cloud sign-up onto the marketing site: about +22% sign-ups in an early read.
- "Analyze with AI" buttons on blogs and guides, which have their own case study.
Other Projects
