Data Scientist - Marketing Measurement & Causal Inference

Commitment: Approx. 3 days per week, with flexibility around project requirements
Location: Remote, UK
Rate: Competitive, dependent on experience
Start: ASAP
About InfluencyIQ
InfluencyIQ is a campaign intelligence platform built to help marketers better understand what is driving campaign performance.
Marketing data is increasingly fragmented across channels, platforms and business systems. InfluencyIQ brings those signals together within a single campaign environment, helping marketers connect activity with customer behaviour and commercial outcomes.
The platform is currently in development and approaching initial launch. We are now looking for a Data Scientist with strong statistical and measurement expertise to help develop and validate the analytical methodology behind our Campaign Intelligence capabilities.
The role
You will work directly with the Founder and Lead Developer to help turn an existing measurement framework into a robust, practical and production-ready analytical methodology.
This is not a traditional reporting or dashboarding role. You will be working on complex, multi-source marketing data and helping determine how statistical and measurement techniques can be applied credibly within a commercial product.
Your focus will be on developing and validating approaches across areas such as marketing measurement, time-series analysis, attribution, causal inference and statistical modelling, while ensuring outputs are appropriately qualified according to the strength of the available evidence.
You will own the statistical methodology and validation, working closely with our existing development team, who will own production engineering, platform architecture, integrations and UI.
What you'll work on
Review and challenge the existing InfluencyIQ measurement and analysis framework.
Define a production-ready v1 methodology for analysing campaign performance across multiple data sources.
Design and validate statistical approaches for establishing campaign baselines and evaluating performance changes.
Develop appropriate methods across time-series analysis, correlation, attribution, causal inference and broader marketing measurement.
Determine suitable analytical approaches for different data types, levels of data availability and sample sizes.
Develop practical approaches for dealing with confounding factors, overlapping activity and uncertainty within real-world marketing data.
Establish clear methodological standards for the strength, reliability and limitations of analytical findings.
Prototype and test analytical approaches where required.
Work closely with the Lead Developer as approved methods are translated into production functionality.
Validate outputs using test, synthetic and ultimately live campaign data.
Document the final methodology, assumptions, limitations and decision rules sufficiently for ongoing engineering and product development.
What we're looking for
You are likely to have a strong background in applied statistics, data science, econometrics, causal inference or marketing measurement.
We are particularly interested in candidates with experience across several of the following areas:
Causal inference and observational data
Time-series analysis
Marketing effectiveness or marketing measurement
Econometrics
Experimentation and incrementality
Attribution
Lagged relationships and intervention/event analysis
Statistical significance and multiple-testing considerations
Baseline modelling
Confounding and covariate analysis
Python and common statistical/data-science libraries
Translating statistical methodologies into production software
Experience with MMM, digital attribution, ecommerce, advertising platforms or multi-channel marketing datasets would be valuable, but is not essential if your statistical fundamentals are strong.
What matters to us
We are not looking for academic complexity for its own sake.
We need someone who can balance statistical rigour with the realities of building a commercial product: understand where strong conclusions are possible, where weaker evidence can still be useful if properly labelled, and where the platform should refuse to make a claim.
You should be comfortable making decisions, explaining statistical concepts clearly to non-specialists and collaborating closely with engineers to ensure the methodology is practical to implement.
The engagement
The initial engagement will run for approximately two months at around three days per week, with working time able to flex around key phases of the project.
The immediate objective is to deliver and validate a defensible production-ready v1 of the Campaign Intelligence methodology, rather than solve every future measurement problem within the first engagement.
There is potential for the role to extend as InfluencyIQ develops further capabilities around longitudinal learning, predictive modelling, forecasting and marketing investment optimisation.
How to apply
Please send your CV alongside a short note outlining any relevant experience in marketing measurement, causal inference, econometrics, time-series analysis or statistical product development.

Work arrangement
Yes