<p> </p> <p>Fospha is the measurement system enterprise retail and ecommerce brands run their business on. We give marketing teams one clear, daily view of what's actually working — across every channel and everywhere they sell, from their website to Amazon and TikTok Shop — down to the level of a single ad or piece of creative. It replaces guesswork and gut feel with a number marketing, finance and agencies can all trust and act on.</p> <p>Brands including Dyson, Gymshark and CarParts use Fospha up to 25 times a day to decide where budget should move next. We've spent over a decade building this, with more than $40 billion in marketing spend now optimised through the platform — and we're scaling fast across London, Mumbai and Austin.</p> <p> </p> <h2>About the role</h2> <p>We're looking for a <strong>Graduate Marketing Scientist</strong> to join Fospha's Marketing Science team in London.</p> <p>Fospha builds marketing measurement products for ecommerce brands — attribution, marketing mix modelling, incrementality testing, and brand impact measurement. Marketing Science owns the applied end of that: designing and delivering incrementality tests and MMM engagements for clients, and standing behind the numbers when a client challenges them.</p> <p>This is the entry point into the function, and it is a hands-on one. You'll work on live test and MMM delivery under supervision from the start, and you'll be the first person looking at a client's data when a number doesn't behave the way it should. It's a role for someone who wants to learn causal measurement properly, in a business where it's the product rather than a side project.</p> <blockquote> <p><strong>Team:</strong> Marketing Science</p> </blockquote> <blockquote> <p><strong>Level:</strong> Graduate — Entry (Data Science Career Development Framework)</p> </blockquote> <blockquote> <p><strong>Location:</strong> London</p> </blockquote> <hr> <h2>What you'll do</h2> <h3>Marketing mix modelling (MMM) & Testing Services</h3> <ul> <li><strong>Assemble and validate test data</strong> — geo-level spend and conversion series, checking pre-period parity between treatment and control, spotting the coverage gaps that invalidate a design before it launches</li> <li><strong>Support test design under review</strong> — market matching and control selection, power and minimum detectable effect sanity checks, and identifying contamination risks such as geo-targeting settings that don't behave the way the platform's documentation claims</li> <li><strong>Run analysis and read the results honestly</strong> — pre-treatment fit diagnostics, lift estimates with their intervals, and what a null result does and doesn't tell you</li> <li><strong>Qualify client data for MMM</strong> — spend coverage across channels, whether there's enough variation in spend to identify an effect at all, series length and granularity, collinearity between channels, and gaps that will bias the result</li> <li><strong>Assemble and validate model input datasets</strong>, and investigate the discrepancies that surface when you do</li> <li><strong>Support model runs and read the diagnostics</strong> — fit, residuals, convergence, and whether a channel's estimated contribution is plausible</li> <li><strong>Contribute to output-extension work</strong> under review — building on an existing MMM result, for example forecasting or budget scenario work derived from it</li> <li><strong>Compare results across methods</strong> — where MMM, incrementality, and platform-reported figures disagree, understanding why is the interesting part of the job</li> </ul> <h3>Model trust and diagnostics</h3> <ul> <li><strong>First and second line on client trust queries</strong> — investigating why a number changed, working in SQL against client data to isolate the cause</li> <li><strong>Distinguish a bug from a methodology change</strong> — attribution window changes, model recalibration, data feed gaps, and platform reporting shifts all look similar from the outside and have very different signatures underneath</li> <li><strong>Triage PSPs on model trust</strong>, resolve what you can, and escalate what turns out to be a genuine model problem with a clear diagnosis attached</li> <li><strong>Reconcile platform-reported figures against our measurement</strong> — why walled-garden ROAS disagrees with ours is the hardest recurring question in the business, and you'll be learning it from the inside</li> <li><strong>Log and tag incidents consistently</strong>, so recurring failure patterns become visible and can be automated away rather than repeatedly handled</li> </ul> <h3>Client communication and enablement</h3> <ul> <li><strong>Run templated explainer sessions</strong> under review, walking clients through how our measurement works</li> <li><strong>Draft documentation and presentations</strong> above the core explainer content, and feed recurring query themes back into the source material</li> <li><strong>Fact-check methodology claims</strong> in product marketing collateral before it goes out</li> </ul> <hr> <h2>What we're looking for</h2> <p>We’re looking for someone with a strong foundation in maths and stats with clear communication who is looking to growth their skillset.</p> <h3&g