Data Scientist

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.

Brands including Dyson, Gymshark, Adanola and Huel 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.

We're looking for a Data Scientist to join Fospha's Data Science team in London.

Fospha builds marketing measurement products for ecommerce brands — attribution, marketing mix modelling, incrementality testing, and brand impact measurement. Our Data Science team owns the models behind all of it, from methodology through to production code. You will work across our technical stacks and own maintaining and growing the codebases that power our product and solutions.

This role suits an established data scientist who wants to own things properly. You'll lead larger production projects, choose the modelling approach rather than being handed it, and talk to clients as the modelling expert in the room — supported by a team that reviews each other's work seriously, and a company that rewards high agency with ownership.

Team: Data Science

Level: Career (Data Science Career Development Framework)

Location: London

• Lead larger production projects end to end — scoping, building, and shipping production-level code, rather than working through tickets someone else has specified
• Choose the modelling approach — independently selecting and applying the right method within our suite, across attribution, marketing mix modelling, incrementality testing, and brand impact. You will be given time to keep up-to-date on the latest methodologies in the industry
• Resolve bugs and queries independently — including in parts of the codebase you didn't write, without needing to route them upwards
• Use AI as leverage, not as a crutch — solving coding tickets quickly, unblocking yourself on product and engineering dependencies, and building automation workflows that save the team time
• Work with QA properly — using our automated tooling efficiently and flagging the gaps in it rather than working around them
• Communicate as a modelling expert — confidently and independently, with clients and with colleagues, including when the message is that a number they like is wrong
• Help develop the people around you — code review, methodology critique, and hands-on support for less experienced colleagues. Breaking down tickets and helping those more junior is essential

• Solid commercial data science experience — typically 3–5 years, with a track record of shipping models into production
• Strong ML knowledge across multiple algorithm families, and the judgement to pick the right approach for the problem rather than the one you know best
• Strong Python and SQL, with the ability to lead on production-level code and set the standard others work to
• Able to debug and resolve issues independently across unfamiliar repositories, using AI tooling to accelerate rather than to guess, while still understanding the problem fully
• Strong AI fluency — you solve tickets quickly with it, you unblock cross-department dependencies with it, you build automation workflows with it, and you never send AI-assisted output without checking it
• Confident, independent communication with clients and stakeholders as the technical authority on the work
• Emerging mentorship — you're ready to develop junior colleagues, and you want to
• Initiative in accepting and planning your own work, rather than waiting to be allocated it
• Genuine attention to detail — much of this work involves noticing when a number is wrong

• Bayesian modelling experience, particularly hierarchical models
• Experience with AWS or comparable cloud tooling
• Familiarity with automated QA tooling and test coverage practices
• Experience with marketing, ecommerce, or advertising data

Experience with attribution methodology, MMM, incrementality testing, or Bayesian modelling is genuinely an advantage at this level — but it isn't a filter. Our stack takes time to learn regardless of what you arrive with, and we'd rather hire strong modelling judgement and teach the domain.

How you'll grow

We run a published Data Science Career Development Framework with six levels. You'd join at Career, where the expectations are:

AI Fluency & Tooling: Leverages AI to solve coding tickets quickly; consistently unblocks themselves on cross-department dependencies, especially product and engineering; never sends AI-assisted communication without checking quality; expert at using AI to build automation workflows.

Machine Learning & Modelling: Strong ML knowledge across multiple algorithm families; independently chooses and applies the right modelling approach within Fospha's suite.

Engineering & Codebase: Leads on larger coding projects and tickets with production-level code; resolves queries and bugs efficiently and independently; master at using supplied AWS tooling through AI; uses automated QA tools efficiently and flags gaps to the QA team.

Stakeholder & Communication: Confidently and independently communicates with clients and colleagues as a modelling expert; assists with the development of junior members of staff.

Job Complexity: Consistent contributor and respected knowledge holder, with emerging collaboration and mentorship abilities.

Supervision: Demonstrates initiative in accepting and planning work.

Progression to Advanced level is against explicit, published criteria — taking on almost any ticket efficiently, building and planning production repositories, collaborating with product on project outcomes and estimates, actively developing junior colleagues, and representing Fospha as a trusted external voice on modelling in senior client and partner conversations. You'll know what you're working towards from your first week.

Throughout, we look for the same core behaviours: concise communication, collaboration, problem solving, critical thinking, growth mindset, attention to detail, time management, and initiative.

• Real methodological depth. Bayesian attribution, MMM, geo lift testing, and causal inference are our day job, not a side project. You'll be working on all of it, not adjacent to it.
• Ownership at the methodology level. At this level you're choosing the approach, not implementing someone else's choice.
• Published career framework. No guessing what the next level requires or when you'll get there.
• A team that reviews each other's work properly. Code review and methodology critique from people who care about getting the model right.
• Direct client impact. The models you build and maintain drive real budget decisions at brands you'll recognise.