> Markdown version of [/jobs/ext/4654333-data-scientist](https://www.wearedevelopers.com/jobs/ext/4654333-data-scientist). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Data Scientist - **Company:** Tails - **Location:** Richmond, United Kingdom (Remote available) - **Experience:** Expert - **Salary:** £47,778 - **Contract:** Permanent contract - **Skills:** A/B Testing, Airflow, Amazon Web Services, Data Analysis, Information Engineering, Data Infrastructure, Python (Programming Language), Machine Learning, SQL Databases, Snowflake, Core Data, Scikit Learn, Machine Learning Operations, Data Pipelines - **Published:** November 10, 2026 - **Apply:** https://www.adzuna.co.uk/jobs/details/5959384669 ## About the Role * 2-4+ years in a data science role, with a proven track record of deploying statistical and machine learning models on complex, messy, real-world data, and showing the business impact they had * Experience building customer lifetime value, survival, propensity or churn models, ideally for a subscription business and used to drive CRM or marketing activity * Strong experimental design and causal inference: A/B tests, holdouts and control groups, uplift measurement and incrementality * Experience feeding model outputs into paid media, such as value-based bidding, offline or server-side conversion events, or predicted-value audiences on Meta or Google * Experience with marketing mix modelling or attribution * A bachelor's or master's degree in a STEM related field * Advanced proficiency in Python and SQL with extensive experience working with core data science libraries (e.g. pandas/polars, scikit-learn) * Familiar with cloud platforms (e.g. AWS) and orchestration tools (e.g. Airflow, Dagster) in relation to operationalising data science products * Experience leveraging Snowflake ML and robust data pipelines to deliver scalable, production-ready data science solutions * A decision-first mindset: you start from the business decision and its owner, and you're comfortable recommending to stop when a test doesn't deliver * Highly Desirable: Subscriptions or D2C business experience ## Description The Data Scientist will own Tails.com's predictive modelling and machine learning work, with a primary focus on improving customer lifetime value (LTV) and driving more sustainable, profitable growth. They will build and develop models that help the business better understand which customers are likely to stay, spend and engage long-term, using these insights to improve acquisition strategies, customer retention and the overall customer experience. The role will work closely with Performance Marketing, CRM, Customer Experience & Retention, Product, Finance, Engineering and Data teams, taking ownership of projects from identifying opportunities and building models through to testing, implementation and measuring commercial impact. The focus isn't simply on building models, but on turning data science into practical business decisions that improve how Tails.com acquires and retains customers. 3 best things about the job 1. Ownership & influence: Own the predictive modelling and machine learning agenda, with the opportunity to shape the wider data science roadmap and influence business strategy. 2. Commercial impact: Work on high-impact projects around customer lifetime value, acquisition and retention, with a direct link between your work and business growth. 3. Cross-functional exposure: Partner with teams across Marketing, Product, Finance, Customer Experience and Technology, seeing your work implemented across the business. Measures of success - In the first few months, you would have: * Built strong relationships with key stakeholders across Performance Marketing, CRM, Product, Finance, Customer Experience and Technology. * Developed a strong understanding of Tails.com's customer lifecycle, existing data infrastructure, models and commercial priorities. * Taken ownership of the customer lifetime value modelling roadmap, identifying opportunities to improve existing predictions. * Started developing and implementing predictive models to help improve customer acquisition, trial conversion and retention. * Identified opportunities to embed data-driven insights into CRM and paid media activity, with clear testing plans to measure commercial impact. Roles & Responsibilities: Lifetime value, predicted lifetime value and acquisition signals * Own our lifetime value model, building on the models and definitions we already have rather than starting again, and agree one definition of lifetime value, CAC and payback with Finance and Analytics * Predict each new customer's lifetime value as early as possible after sign-up, from acquisition source, behaviour, dog and plan data and early engagement * Send that prediction to our acquisition partners (such as Meta and Google) so they optimise towards customers likely to stay rather than cheap sign-ups, and check with Marketing and Engineering that the events we send are complete, timely and matched to the right customers * Compare channels by the value their customers go on to create, not only cost per sign-up * As the value work matures, help build marketing mix modelling that shows how spend across channels drives valuable customers, building on the measurement we already have Conversion and early-life retention * Find where and why trial customers do not become paying customers, and test the changes that improve trial-to-paid conversion without relying solely on discounts * Decide which new customers need extra support in their first months, what kind, and while there is still time to change the outcome, matching risk and value to the actions most likely to work * Run controlled tests with CRM and Customer Experience & Retention that measure the extra conversion, retention and retained value they create Experiment standards * Set the minimum test and holdout standards for this work, so every result can show what it caused Production and monitoring * Deploy models to production with help from our data engineering team, and monitor their health, drift and retraining so problems are caught and fixed quickly