Decision log
Three pieces of work from BTPN Syariah, Jakarta. In each I was product owner of a scrum squad of about ten: a scrum master, a tech lead, engineers and QA. I owned the roadmap and priorities, and dealt with business, compliance and users.
01 · Payments
Fixing bill payments that failed quietly
Senior Product Manager, transactions squad
2024 to 2025
90%fewer failed payments, roughly
- Problem
- I took over bill payments, transfers and top-ups in SiTepat, the bank's customer app. Some products could never succeed, because the provider had dropped them and the app still listed them.
- Decision
- We built dashboards for every product and provider, audited the whole product list, and set one rule: reroute when a provider is unreliable, hide when the fault is temporary, remove when the product is gone.
- Trade-off
- Rerouting was sometimes the less profitable option. We chose the payment that works.
- Result
- Failed payments fell from about 20 in every 100 to about 2, and the app sold only what it could deliver.
02 · Onboarding
A shorter sign-up for customers we already knew
Product Manager, onboarding squad
2021 to 2023
4 or 5steps cut from app sign-up
- Problem
- In the SiTepat pilot, every new user went through the same long sign-up, whatever the bank already knew about them.
- Decision
- After 40+ interviews, observations and A/B prototype tests, we added an ID scan that recognised people already registered as agents and gave them a shorter route.
- Trade-off
- Two routes are harder to build and maintain than one. Everyone else still got a simpler flow, with 4 or 5 steps removed.
- Result
- More people who started sign-up finished it. The pilot, run in three provinces, reached about 1,000 customers transacting every month.
03 · Field tools
Replacing paper forms for 10,000+ field officers
Product Manager, acquisition squad
2016 to 2019
40%less time to enrol a customer
- Problem
- Field officers enrolled new customers on paper. Every form was then typed into the system a second time.
- Decision
- We built a mobile tool that let officers enter details once, on the spot. I wrote the playbooks and trained staff across regions for the rollout.
- Constraint
- It was the officers' first digital tool, and it could not depend on a good signal in the field. We built it offline-first, so entries synced automatically once officers were back at the office.
- Result
- Enrolling a customer took 40% less time than on paper, measured from start and finish times. The second round of typing went away.
