JTP Analytics helps data-mature organisations understand what drives growth, where marketing is genuinely working, and how to make better decisions under uncertainty.
Run independently by Jake Piekarski, JTP combines rigorous statistical modelling with practical marketing experience. The focus is not simply producing another dashboard or model: it is building measurement that teams can understand, challenge and use.
Marketing measurement built around your decisions
JTP is best suited to organisations with established data, meaningful marketing activity and questions that standard reporting cannot answer. This includes direct-to-consumer businesses, established brands, in-house marketing teams and agencies strengthening their own measurement offering across the UK and US.
The methods are designed to remain useful as access to user-level tracking and third-party cookies becomes less reliable. Aggregate modelling, experiments and causal inference provide a more durable foundation for decision-making.
Services
Bespoke modelling
Build a model around the commercial question, from Marketing Mix Modelling and budget optimisation to product adoption forecasting.
Incrementality testing
Design and analyse experiments that separate activity which caused an outcome from demand that would have happened anyway.
Model audits
Review an existing model’s assumptions, validation, implementation and suitability before important decisions are made from it.
Team education
Build confidence across technical and marketing teams through training, workshops and practical measurement guidance.
Independent expertise, explained clearly
JTP is led by Jake Piekarski, a Principal Data Scientist at PyMC Labs with three years of agency and consultancy experience in marketing analytics. Jake has worked across Fortune 100 consumer packaged goods, airlines, manufacturing and consumer goods, food and beverage businesses, and charities.
Alongside client work, Jake contributes to the open-source PyMC-Marketing ecosystem and has taught the Bayesian Marketing Analytics course. His work combines the statistical depth needed to build credible models with the communication needed to make them useful.
Start a conversation
If you have a modelling problem, an existing model that needs a second opinion, or a marketing measurement question that has become difficult to answer, send a short note below.
For budgets, timelines and more detailed requirements, use the full project enquiry form.