Calm Senior Data Scientist - Retention & Lifecycle at Calm owns the science behind subscriber retention and lifecycle decisioning. The role involves building retention fact bases, diagnosing churn, and designing lifecycle experiments.
Responsibilities
As our Senior Data Scientist for Retention & Lifecycle, you will own the science behind how Calm keeps its subscribers. Retention is one of the biggest levers on Calm's subscriber growth, and retention science that feeds lifecycle decisioning is one of the data team's top company-impact priorities. You will build the retention fact base the business runs on, find where and why subscribers churn, and partner with Lifecycle Marketing to turn that understanding into targeted, measured interventions: save and winback offers, billing recovery, and messaging aimed at the users it will actually change. You will design the tests and permanent holdouts that tell us what each intervention is really worth, and help move Calm from campaigns judged on counts to decisions judged on incremental retained revenueThis role is ideal for someone who has worked on subscription retention before, likes being close to the people who act on the analysis, and balances statistical rigor with getting a good answer in front of a decision-maker on time.
Build the retention fact base. Own tenure-day retention curves, churn cliffs, cohort retention, cancel and pause outcomes, save rates and winback baselines (organic vs. lifecycle-driven, first vs. repeat), published in our semantic layer so Lifecycle's goals are measured against agreed baselines.
Diagnose voluntary and involuntary churn. Size involuntary churn by platform and partner with Engineering and Finance on retry and dunning timing, grace periods, payment-method tests and retention messaging to recover failed payments.
Design and read lifecycle experiments. Plan randomized tests for save, winback and re-engagement programs, set up permanent holdouts, and report incremental retained revenue, including offer cost, each quarter.
Build churn and uplift models with our ML partners. Develop churn propensity and uplift models trained on randomized lifecycle tests, delivered daily to our CRM, so campaigns target the users an intervention will actually move.
Shape lifecycle decisioning. Help design the rules, guardrails and holdouts behind save and winback offer decisions, laying the groundwork for next-best-action personalization across content, lifecycle and offers.
Model how users move between states. Contribute to Calm's growth model: engagement states (new, current, reactivated, at-risk, dormant) crossed with subscription states (trial, auto-renew on/off, grace, lapsed, won back), and the transition rates that drive bookings.
Keep leaders informed. Bring retention and lifecycle insights to the monthly Metrics Review, size retention goals in annual planning, and maintain lifecycle dashboards in Databricks AI/BI and Amplitude.
Partner across teams. Work day to day with Lifecycle Marketing, Product, Finance and Engineering, and with analytics engineers who own the behavioral, cohort and CRM audience data you rely on.
Qualification
Build the retention fact baseDesign and read lifecycle experimentsShape lifecycle decisioningModel how users move between statesKeep leaders informedPartner across teamsExperience with uplift modelingExperience with involuntary churnExperience with DatabricksMinimum Requirements5 years of data science experience
Required
Build the retention fact base. Own tenure-day retention curves, churn cliffs, cohort retention, cancel and pause outcomes, save rates and winback baselines (organic vs. lifecycle-driven, first vs. repeat), published in our semantic layer so Lifecycle's goals are measured against agreed baselines.
Diagnose voluntary and involuntary churn. Size involuntary churn by platform and partner with Engineering and Finance on retry and dunning timing, grace periods, payment-method tests and retention messaging to recover failed payments.
Design and read lifecycle experiments. Plan randomized tests for save, winback and re-engagement programs, set up permanent holdouts, and report incremental retained revenue, including offer cost, each quarter.
Build churn and uplift models with our ML partners. Develop churn propensity and uplift models trained on randomized lifecycle tests, delivered daily to our CRM, so campaigns target the users an intervention will actually move.
Shape lifecycle decisioning. Help design the rules, guardrails and holdouts behind save and winback offer decisions, laying the groundwork for next-best-action personalization across content, lifecycle and offers.
Model how users move between states. Contribute to Calm's growth model: engagement states (new, current, reactivated, at-risk, dormant) crossed with subscription states (trial, auto-renew on/off, grace, lapsed, won back), and the transition rates that drive bookings.
Keep leaders informed. Bring retention and lifecycle insights to the monthly Metrics Review, size retention goals in annual planning, and maintain lifecycle dashboards in Databricks AI/BI and Amplitude.
Partner across teams. Work day to day with Lifecycle Marketing, Product, Finance and Engineering, and with analytics engineers who own the behavioral, cohort and CRM audience data you rely on.
Preferred
Experience with uplift modeling, heterogeneous treatment effects, or causal inference for targeting
Experience with involuntary churn: payment retries, dunning, grace periods, or app store and web billing
Experience putting model scores into production in a CRM or lifecycle platform (e.g., Iterable, Braze) through reverse ETL
Experience with offer and discount economics: save offers, winback pricing, and measuring offer cost against retained revenue
Experience with Databricks, a semantic or metrics layer, Amplitude, or similar tools
Experience building growth or state-transition models of a user base
Minimum Requirements
5 years of data science experience, including at least 2 years on retention, lifecycle or CRM problems at a consumer or subscription business