Product Data Scientist

For an international organisation whose work has a clear social purpose, this is an opportunity to help shape how product decisions are made across its digital platforms.

You will work closely with Product, Design and Engineering to understand what users do, evaluate changes and make results useful for decision-making. The focus is on measurement, experimentation and reliable evidence, rather than building machine learning models for production.

What you'll do

• Design A/B tests and use causal methods when controlled testing is not possible.
• Define success metrics and check whether they measure the outcomes that matter.
• Analyse adoption, retention, funnels and user behaviour across different groups.
• Use statistical models to understand results, test assumptions and communicate uncertainty.
• Set clear standards for analysis and help colleagues assess the quality of their findings.
• Work with engineers to improve tracking, documentation and data quality.
• Evaluate AI tools that support analysis, including how their answers are checked.
• Turn complex findings into concise recommendations for product teams and senior stakeholders.

What you bring

• 3+ years of experience in statistical modelling or machine learning.
• Strong SQL skills and fluency in Python or R.
• Experience designing experiments and applying causal methods when randomisation is not possible.
• A solid understanding of regression, multilevel models and statistical uncertainty.
• Experience with product analytics, including funnels, cohorts, retention and adoption.
• A strong eye for data quality and the ability to work closely with Product, Design and Engineering.
• Clear communication skills, including the confidence to explain what the data cannot establish.
• Experience with BI tools, survey research or behavioural research methods is a plus

The challenge

A product change can improve one metric without improving the outcome that matters. Sometimes a controlled experiment is possible. Often, users are part of different groups or a change reaches them at different times.

Your challenge is to build evidence that teams can trust and use. That means choosing the right measures, understanding the limits of the data and making the findings clear enough to guide a decision.

How do you establish whether a change made a difference? Which comparisons are valid? And how do you help teams act on the results without overstating what the data proves?

Package

In addition to the competitive salary, the package includes an Individual Choice Budget (12%), profit sharing, a pension plan under the applicable CLA, remote work options and additional employment benefits.