AskBosco · Churn and lifetime value · 2024
Churnandlifetimevaluefor Penhaligon's (PUIG)
Turned a rule-based CRM into value-based targeting, worth €176k+ in win-back uplift in backtests.
€176k+Win-back uplift in backtests
01 / 03
Every customer, one point
A CRM with fixed rules treats these customers the same.
02 / 03
Grouped by value and churn risk
The models predict who is likely to leave and what each customer is worth.
03 / 03
Win back the few who matter
Targeting by value found €176k+ of win-back uplift in backtests.
The project
- Client
- Penhaligon's (PUIG)
- Role
- Data Scientist, AskBosco
- When
- 2024
- Discipline
- Churn and lifetime value
- Tools
- Survival models
- Classification
- CRM integration
- Vertex AI
The situation
Penhaligon's CRM targeted customers with fixed rules, rather than by how likely they were to leave or what they were worth.
What I built
- An end-to-end churn and lifetime value modelling system
- Value-based targeting across retention and acquisition
- Lookalike audiences built from the highest-value customers
What changed
Financial backtests showed €176k in incremental win-back uplift, a further €114k from high-value targeting, and lookalike audiences worth about €109 more per customer.
- €176k+Win-back uplift in backtests
- €114kFrom high-value targeting
- €109More per lookalike customer
Hiring for marketing measurement or customer ML in the UK?
Eligible to work in the UK, no sponsorship needed. Based in London.
The CV is a two-page PDF, rebuilt from this site whenever it changes.
Jihar Gifari · London · jigifari@gmail.com · +44 7985 687426