Data Scientist

DealerOn

14d•$86,100 - $128,700•Remote

About The Position

Serve as the embedded analytics partner to Product teams; define feature success metrics before launch and measure outcomes after. Design and execute platform experiments — page templates, form flows, merchandising treatments, search experience — including power analysis and preregistered success criteria. Deliver clear reads on whether launches performed, including negative results. Apply quasi-experimental methods where controlled tests are not feasible: staged-rollout cohorts, difference-in-differences, matched comparisons, holdouts. Produce the analysis behind OEM and enterprise dealer group business reviews, and present those findings in the meeting alongside Account Management and the Director, who own the relationships. Answer network-level OEM questions: dealer base performance against benchmark, lagging markets and their causes, digital program and co-op spend effectiveness. Build and validate predictive models: lead scoring, days-to-sell and inventory turn, renewal risk, market-level demand forecasting. Partner with Data Engineering to move models and pipelines into production. Build the cross-network benchmarks used by OEM and enterprise account teams. Partner with Marketing to produce original research for external publication — industry studies, benchmark reports, whitepapers — owning the methodology, the analysis and the accuracy of every figure. Serve as the technical author behind that work as it reaches a wider audience: sales enablement material, conference and webinar content, and press or analyst enquiries. Convert recurring analyses into dashboards, templates and documented pipelines. Escalate platform issues surfaced in client data to Product, including where the cause is DealerOn's.

Requirements

3–5 years in analytics, data science, or analytics consulting.
Advanced SQL — complex analytical queries against a data warehouse, written and debugged unaided.
Python or R, with pandas or dplyr and a statistical modelling library (statsmodels, scikit-learn, tidymodels).
Experimentation and causal inference: test design, power analysis, and quasiexperimental methods where a clean test is unavailable.
Applied predictive modelling with evaluation against real baselines.
Experience embedded with product managers and engineers, working to their cadence.
Experience presenting analysis directly to senior business stakeholders and defending it under questioning.
BI tool fluency: Tableau, Power BI, Looker or similar.
Effective use of AI coding assistants, with the judgment to verify their output.

Nice To Haves

SaaS, digital marketing, agency, or automotive retail experience.
Familiarity with OEM digital programs, co-op, or tier-3 marketing.
Marketing measurement depth: attribution, incrementality testing, media mix modelling, GA4.
Analytics engineering practice: dbt, warehouse modelling, orchestration.
Cloud warehouse experience: BigQuery, Snowflake, or Redshift.
Automotive experience is not required.

Responsibilities

Serve as the embedded analytics partner to Product teams; define feature success metrics before launch and measure outcomes after.
Design and execute platform experiments — page templates, form flows, merchandising treatments, search experience — including power analysis and preregistered success criteria.
Deliver clear reads on whether launches performed, including negative results.
Apply quasi-experimental methods where controlled tests are not feasible: staged-rollout cohorts, difference-in-differences, matched comparisons, holdouts.
Produce the analysis behind OEM and enterprise dealer group business reviews, and present those findings in the meeting alongside Account Management and the Director, who own the relationships.
Answer network-level OEM questions: dealer base performance against benchmark, lagging markets and their causes, digital program and co-op spend effectiveness.
Build and validate predictive models: lead scoring, days-to-sell and inventory turn, renewal risk, market-level demand forecasting.
Partner with Data Engineering to move models and pipelines into production.
Build the cross-network benchmarks used by OEM and enterprise account teams.
Partner with Marketing to produce original research for external publication — industry studies, benchmark reports, whitepapers — owning the methodology, the analysis and the accuracy of every figure.
Serve as the technical author behind that work as it reaches a wider audience: sales enablement material, conference and webinar content, and press or analyst enquiries.
Convert recurring analyses into dashboards, templates and documented pipelines.
Escalate platform issues surfaced in client data to Product, including where the cause is DealerOn's.