Experience

Four roles, four sectors, and a fairly straight line through them. I started in life sciences consulting, turning contradictory market evidence into recommendations a client could act on. Then analytics at an automotive marketplace, where I ended up doing product work under a different job title and decided to make the move properly. Since then it's been data product ownership in beauty, and two years owning an enterprise system in heavy industry from discovery to production.

Each role below is written the same way: the problem I walked into, what I did about it, what I owned, and what changed. If you only read one line per role, read the last one.

  1. Nov 2024 to Sep 2026

    Industrial gases

    Product Manager Air Liquide

    Problem
    A legacy enterprise system used by field teams across Europe had to be replaced without interrupting the work that ran on it.
    Action
    Ran discovery with the teams who used it daily, wrote the PRDs, owned the backlog, and set up user acceptance testing across several European teams with test books that told testers what to check and what to flag.
    Product
    A replacement system taken from discovery to live, plus a role-based secure access feature that unblocked a contractual-data workflow nobody had picked up, cleared through Legal and IT governance.
    Outcome
    Zero-downtime cutover. System in daily use across European field teams. Monthly reporting automated into Power BI, cutting manual effort by around 40%.
    • Enterprise
    • Build and run
    • UAT
    • Change management
    • Power BI

  2. Jan 2024 to Jun 2024

    Beauty

    Data Product Owner L'Oréal

    Problem
    Market-share reporting across the Top 20 markets was fragmented, category definitions disagreed between consolidators, and a thousand-plus brands were being mapped to business categories by hand.
    Action
    Wrote the user stories and ran the backlog to bring Kantar, Nielsen and B2B sources into AWS, and reconciled category definitions across more than 20 markets into one comparable view.
    Product
    A standardised set of Power BI dashboards for the Top 20 markets, and an AI fuzzy-matching agent that allocates 1,000+ brands across four universes automatically.
    Outcome
    Manual brand mapping retired. Near real-time portfolio visibility for the business. A recurring blocker in executive revenue reviews removed.
    • AI agents
    • Data products
    • AWS
    • Power BI

  3. Feb 2023 to Jun 2023

    Automotive e-commerce

    Business Analyst Spinny

    Problem
    Buyers were dropping between showroom visit and token payment in the Hub app, and a quarter of deliveries were being cancelled after the deposit had already been paid.
    Action
    Ran discovery with showroom staff who used the app every day, then mined CRM leads data to rebuild the token-to-delivery journey and find where cancellations came from.
    Product
    A redesigned visit-to-token flow, and a ranked list of cancellation drivers handed to the closure team to run win-back against.
    Outcome
    Visit-to-token conversion up 10%. Delivery cancellations down 25%.
    • Discovery
    • Funnel
    • SQL
    • Consumer

  4. May 2021 to Jan 2023

    Life sciences consulting

    Associate ZoomRx

    Problem
    Pharma clients including Janssen, GSK and Sanofi needed to know how their brands were performing against named rivals and whether a new launch would win share or eat their own.
    Action
    Ran competitive benchmarking on brand reach, share of voice, message effectiveness and sales-rep effectiveness. Scoped go-to-market and launch studies, including cannibalisation analysis.
    Product
    Promotional recommendations and launch reads that clients used to decide spend and positioning.
    Outcome
    Two years of turning contradictory evidence into something a client could act on, which turned out to be most of the product job too.
    • Strategy
    • Life sciences
    • Go-to-market