Areas of expertise
Experimentation strategy
Building and scaling A/B testing programmes — from experiment design and statistical rigour through to cultural adoption across product teams.
Product analytics
Defining metrics, building dashboards, running deep-dive analyses. Turning product data into decisions rather than reports.
Interim leadership
Stepped up as interim Experimentation Manager at Reach in 2022, leading the team through a delivery gap and helping recruit a permanent hire.
Background
Nine years at Reach plc — one of the UK's largest digital publishers — moving from digital production and business analysis into data and experimentation over time. The last four years were product-facing: designing and analysing A/B tests, investigating user behaviour with BigQuery and Athena, and turning results into recommendations for product managers and senior stakeholders. In 2022 I stepped up as interim Experimentation Manager, leading the team through a delivery gap and helping recruit a permanent hire.
Practically: I owned end-to-end experiment analysis across 65+ national and regional titles — producing go/iterate/stop readouts across registration prompts, content gating, and engagement initiatives, contributing to a ~5% uplift in yearly pageviews worth an estimated £50K+ in incremental ad revenue. I designed a painted door experiment to quantify subscription appetite at segment level, providing paywall strategy teams with concrete demand signals where none had previously existed. I evaluated 10–20 experiments per month, built Looker Studio dashboards for behavioural monitoring, and analysed feature rollouts — including a Bookmark feature whose case I helped make, which shipped and remains live today.
Since September 2026 I've been founder of Keniri, which launched in October and is now my full-time work, and before that of Socion — a full-stack matching app in React and Supabase — building on over two decades of independent Socionics research that has already produced Socionics Insight (a 372-page reference site) and a 17-book collection. I also built ChatBotCV, which turns a CV into an interactive chatbot so recruiters can question a candidate's experience directly instead of skimming a static PDF — the same idea behind the chat widget on this site — and ATS Defence, a browser tower defence game where you play the applicant tracking system: built half for the joke and half because it let me own an experimentation stack end-to-end for once — fifteen instrumented events into Supabase, a GrowthBook A/B test on the opening difficulty, and a pre-registered analysis that says up front the sample is probably too small to conclude anything. It plays on a phone as well as a desktop, with a board built for the shape rather than the desktop one shrunk to fit it. Both are parked now that Keniri has my full attention, but still online. I mention this because it's directly relevant to how I work: I'm used to owning problems end-to-end, directing tools rather than waiting for direction, and shipping things in production rather than in decks.
I'm comfortable directing AI tools for analysis work — guiding code generation and interpreting output — which is how most practical analytics work gets done now.
Want to see the work itself? Three experiments written up start-to-decision — the painted-door test, the engagement programme, and the Bookmark case — in the case studies, plus a fourth you can inspect end-to-end rather than take my word for.
Now
I'm building Keniri full-time, and running it the way I ran experiments at Reach: an instrumented funnel, a fixed measurement window agreed before looking at the numbers, and product decisions made on the readout rather than on instinct.
Based in London.