CRO at Scale, How 23 Experiments Generated a 45% QoQ Lead Lift
Most CRO programmes stall because they run tests without a system behind them.
When I took ownership of CRO at Multiplier, the function existed but didn't have a structured experiment pipeline. Tests were running, but results weren't feeding into a prioritised backlog, hypotheses weren't being formally documented, and there was no clear framework for what to test next or why. The volume was low and the compounding wasn't happening.
We fixed the system before scaling the tests. Using Optimizely, we went from 14 experiments in Q4 2025 to 23 in Q1 2026, covering UTM-based URL redirects, on-page HTML variants, IP-level personalisation, new URL structures, CTA positioning, and form format changes across paid and non-paid pages. Every result, win, loss, or inconclusive, fed back into a next-test backlog. Losing tests were treated as seriously as winners because a disproven hypothesis sharpens the next round of thinking.
The discipline around data interpretation mattered as much as the tests themselves. In B2B SaaS, it is genuinely tempting to celebrate any incremental lead movement. We didn't let ourselves do that uncritically. If a test was increasing time-on-page and also lifting leads, we went back and interrogated whether the leads were actually attributable to the CRO initiative or arriving despite it, from a campaign, a channel shift, or seasonal behaviour. CRO experiments should be targeted and accurate; if something is generating leads for reasons you don't fully understand, you haven't learned anything you can use for the next test.
The other area we were disciplined about was non-commercial pages. For content like employment guides and country hiring pages, the goal isn't always lead capture, it's trust and credibility. Keeping that content accurate across 150+ countries is operationally hard, but it's the foundation of why users convert on commercial pages later. Brand building and CRO aren't in tension if you sequence them correctly.
Q1 2026 outcome from CRO alone: 104 incremental leads, 55 additional MQLs, 28 extra SQLs, a 45% QoQ increase in lead volume with no incremental paid spend.
