Personalized pricing and offer transformation

01

The business question

Customers differ in their willingness to borrow, the amount they use and how they respond to price. Applying a small set of broad offer rules makes operation straightforward, but limits how closely the offer can reflect those differences.

The business question was how to improve customer response and portfolio economics while staying within approved risk appetite and affordability constraints. This required a decision about pricing and limits together, with a way to evaluate whether the change created value.

02

My role and the team

I led the business roadmap and cross-functional execution of the initiative, connecting product objectives with the analytical work, rollout decisions and performance evaluation. My responsibilities span the period as Senior Product Manager and my current Product Transformation Lead role.

Today I lead a four-person team spanning data science, pricing optimization, business performance analytics and product transformation. The wider delivery involves Product, Risk, Finance, CRM, Sales, Data and Technology. Risk retains its underwriting and risk-policy responsibilities; model development, implementation and business decisions rely on their respective owners.

03

Decisions that shaped the work

• Define value beyond response alone. Evaluate predicted customer behavior alongside limit utilization and lifetime economics.
• Treat affordability and approved risk appetite as constraints in the offer design. A higher response rate is not sufficient evidence of a better portfolio outcome.
• Connect pricing and limit decisions. Their effects on ticket size, installments and customer take-up need to be considered together.
• Use comparison groups and monitoring to distinguish treatment effects from changes in demand, eligibility and channel mix.

04

From pilot to recurring operation

The initiative progressed from a digital pilot into broader channel rollout. The work included defining which decisions the optimizer would support, coordinating changes with the delivery teams, and evaluating results before wider adoption.

Version management, monitoring and executive review are part of this work. They give teams a way to discuss a proposed change, understand its expected economics and compare the evidence after implementation. Commercial response can be visible before credit performance has matured, so the timing of evaluation matters.

05

Trade-offs and evidence

More granular offers introduce complexity in measurement and operation. A decision that appears attractive in aggregate may depend on its customer mix or on another intervention running at the same time. This makes comparability, clear ownership and the ability to investigate changes essential.

The rollout delivered meaningful commercial uplift within approved risk appetite. This public account does not present a single uplift percentage across different cohorts or periods. Keeping risk appetite unchanged describes the decision constraints; it is not a claim that realized credit losses can never change.

06

What this work taught me

The leadership task is to make the objective, constraints and evidence understandable across specialties. Data science can improve the choices available; product leadership must connect those choices to customer needs, portfolio economics and decisions the organization can implement.

A useful test is whether the teams can explain why an offer policy was chosen, what would make them revise it and who owns that decision. Those questions remain relevant after the first rollout.