Data Analytics Manager in Kota Jakarta Selatan at bukuwarung

About BukuWarung

BukuWarung builds the financial infrastructure for micro and small businesses in Southeast Asia. We serve merchants through payments (BukuAgen and BukuPay), credit (BukuModal), savings and banking (BukuSimpan), and a growing hardware fleet of EDC terminals and Bukupe QRIS soundboxes. Four business lines, one merchant.

The next phase is data native: real time decisioning, thin file underwriting, fraud prevention, and a distribution engine that runs on evidence rather than instinct.

Why this role matters

We are scaling from roughly 25k devices deployed per month to 55-60k by December, from 164k to 420k active devices under support, from 8M to 80M transactions monitored monthly, and a lending book growing about 3x against a tighter loss bar. Every one of those curves is a decision problem before it is a modelling problem.

This role owns the layer where data becomes a decision. You define the numbers, certify them, put them in front of the people who allocate capital and headcount, and build the team that keeps doing it as we go multi-country.

What you will own

Metrics and single source of truth

• Define and certify the metric layer across Agent/EDC, Retail/BukuPay, Lending/BukuModal and BaaS/BukuSimpan. One definition of active merchant, activation, TPV, take rate and MRR that Finance, GTM and the board pack all use

• Own the semantic layer and data contracts with Product and Engineering, so an event change cannot silently break a certified metric

• Own metric change control: who signs off, what happens to history, how the business is told

Business partnering

• Embed analysts into the business lines and run them as one pod, not five orphans

• Sit in the weekly GTM reviews per business line and the monthly Business-Finance-GTM forum as the person who owns the numbers on the page

• Turn recurring questions into instrumentation, not into more analyst hours

GTM, channel and incentive analytics

• Channel level economics across OTS, digital and partnerships: CAC, payback, contribution, retention by cohort

• Measure whether the 2026 incentive framework across MSE, RSE and MEE is buying the behaviour we paid for, and detect gaming

• Merchant cohort and LTV models that inform acquisition targeting and budget allocation

Device and operations analytics

• Activation quality, transaction velocity post activation, dormancy and contribution per device across EDC and Bukupe soundbox, including refurbishment and 3PL cost

• Leading indicators of device dormancy and merchant churn that trigger a named field intervention

• Channel and partner fulfilment visibility for Ops: dispatch timelines, serial mapping accuracy, SLA adherence

Lending portfolio analytics

• Vintage curves, roll rates, delinquency heatmaps and early warning reporting for Arunesh and Ben, weekly

• Collections analytics as the function stands up: repayment intent signals, bucket level performance, collector and agency ROI

• Scoring and underwriting models stay with the data science team under Prakash. You supply the evidence and own how performance is reported

Experimentation

• Test design, holdouts and readouts for product, pricing and field programmes

• Quasi experimental methods where randomisation is not possible: geo splits, staggered field rollouts, synthetic controls

Self serve and AI

• Make routine questions self serve. Build the semantic layer and data quality that lets agents answer them under the AI Shift programme

• Retire manual reporting on a published schedule, starting with the weekly review packs

Team

• Hire, develop and assess analytics leads and analysts. Depth plus business judgment, people who can go from a query to a board slide

• Build the capability map and the bar for the function

Requirements

Must have

• 8-10 years in analytics, with 3+ leading analytics teams of five or more

• Owned a certified metric layer for a multi line business, including the political work of retiring competing definitions

• Fintech, payments, marketplace or on demand at scale. Comfortable with transaction level data

• Fluent in SQL and Python. Hands on with a modern BI and semantic stack (dbt or equivalent, Looker, Metabase or Superset)

• Experimentation depth, including designs for cases where clean randomisation is not available

• Unit economics fluency: MDR and take rate, CAC payback, LTV to CAC, contribution margin per device or per merchant

• Track record of killing manual reporting and shadow spreadsheets, with evidence

• Executive communication. Translates a result into a decision and a board level narrative

Nice to have

• Indonesia or Southeast Asia, with exposure to Bank Indonesia and OJK reporting requirements

• Lending portfolio analytics: vintage, roll rate, early warning systems, collections

• Field force or agent distribution analytics with 500+ frontline staff

• Device telemetry or IoT fleet analytics

• LLM or agent driven self serve analytics in production

First 12 months

• Certified metric layer live across the four business lines, reconciled to Finance, with change control in place

• Weekly GTM, OKR and monthly Business-Finance-GTM packs generated from one source, with manual assembly retired

• Channel and incentive economics that visibly move next quarter budget allocation across MSE, RSE and MEE

• Device economics live: activation quality, dormancy and contribution per device, each tied to a named intervention owner

• Lending portfolio reporting standing weekly with Arunesh and Ben, and collections analytics live as the function stands up

• Analytics pod hired and embedded, with self serve adoption measured rather than asserted