From counting shelves to answering the committee

Mandate

Make group data dependable at the point of capture and accountable for the personal data partners hold on our behalf.

Outcomes

• Replaced a proposed shop-level ordering module with route-day forecasting
• Reduced chilled soya returns from 4.1 to 2.3 per cent of loaded volume in 2024
• Coded bonus cartons and payment type at the counter across 26 depots
• Moved 52,000 distributor-served household records onto one delivery system

Decisions in context

The appointment separated data from finance and gave it a direct line to the group managing director. I used the change to bring forecasting, field data and data protection under one set of definitions. My five heads cover engineering, forecasting, data quality, field systems and the data protection office; the data protection officer also has direct access to the committee chair. The three cases on this site belong to this period, and each began with a field observation rather than a technology proposal. I present to the Audit and Risk Management Committee each quarter on data quality, personal-data incidents and exceptions to our sharing rules. The committee has asked hard questions about the cost of the household transfer, and I have learnt to bring the operating colleagues who carried that cost into the room with me.

Reason for the move

Current appointment; seeking one compatible independent seat.

Mandate

Bring sales, finance and supply data onto shared definitions and one reporting platform.

Outcomes

• Consolidated four sales and finance reporting systems onto one platform by 2021
• Rebuilt household water demand planning by delivery zone after the 2020 movement restrictions
• Established the first data quality measures reported to the audit and risk committee
• Appointed the company's first full-time data protection lead in 2021

Decisions in context

Reporting to the chief financial officer gave the data function credibility with finance and a habit of reconciliation I have kept. We moved four reporting systems onto one platform and agreed definitions for an outlet, a call, a carton and a return. The Movement Control Order in March 2020 tested all of it. Outlet sales collapsed in some districts while household water orders rose beyond our delivery capacity. For several weeks we planned from depot managers' daily calls and told the board that our models were unreliable. Afterwards we built a household demand view by delivery zone, which the operations team still uses. I also asked for a full-time data protection lead in 2021, after an internal audit found household records in more places than anyone had listed. That finding began the work described in my third case.

Reason for the move

Appointed chief data officer when the board created the role reporting to the group managing director.

Mandate

Replace invoice books with handheld devices and give sales management a reliable outlet record.

Outcomes

• Replaced written invoice books with handhelds across 24 depots during 2015
• Redesigned the handheld screens after field rides, removing nine fields
• Built an outlet register of 21,000 van-served shops with checked locations
• Introduced route-day volume reports used for van loading

Decisions in context

The handheld programme was the largest change I had led. Until 2015 our van salespeople wrote orders in invoice books that depots keyed overnight, and errors surfaced days later. We bought devices for 360 vans and wrote the screens at head office. Within two months salespeople were skipping most optional fields, and I learnt that a form designed far from the counter would be completed carelessly at it. We rode with crews in six depots, removed nine fields and made three mandatory. The outlet register became the most valuable product of the programme: 21,000 shops with a checked location, owner contact and trading pattern. My team also began reporting volume by route-day, the measure that later underpinned our forecasting. Reporting to the commercial director, I learnt to defend data work in sales terms.

Reason for the move

Asked by the chief financial officer to lead a combined data and analytics function.

Mandate

Produce depot sales reporting that sales and finance would both accept.

Outcomes

• Reconciled weekly depot sales with finance's cash records until variances were explained
• Evaluated about forty promotion schemes a year against matched depots
• Produced the first monthly report of returns by product and depot

Decisions in context

I arrived from Jakarta knowing distribution but not manufacturing. Telagawangi's orders were written in books and keyed at depots overnight, and sales and finance each kept their own figures. I spent my first year reconciling depot sales with the cash finance had banked, week by week, until each difference had a reason. The sales director began to use our reports only when they matched the money. My six analysts also evaluated promotions by comparing depots that ran a scheme with similar depots that did not. Several popular schemes added little volume, and I learnt to present that finding with the area managers who designed them rather than to them. The first monthly report of returns by product and depot came from this period, and it later shaped how we treated chilled soya. It was also where I first worked with Hairul, then in area sales.

Reason for the move

Promoted to lead route-to-market analytics ahead of the handheld programme.

Mandate

Turn salespeople's invoice books into reports that branch managers could act on.

Outcomes

• Cut the delay between a sale and its report from nine days to two
• Introduced a weekly anomaly list reviewed with each branch manager
• Separated cash, credit and returned-goods records in branch reporting

Decisions in context

Jakarta taught me the economics of the small shop. Our salespeople visited warung and grocers on foot and by motorcycle, collected cash, extended short credit and wrote everything in invoice books. My team keyed those books for nine branches, and in my first months the reports arrived nine days after the sale. Shortening that delay mattered less than I expected. What branch managers valued was a weekly list of anomalies: shops whose credit had grown, returns that did not match deliveries, a salesperson whose cash was consistently short on Mondays. I learnt to treat each explanation from the field as evidence. Wahyu Prasetyadi, the commercial director, insisted that every anomaly list be discussed in person before any action. Working in Bahasa Indonesia, I also learnt how much is lost when a report is designed in one language and used in another.

Reason for the move

Wanted to work for a manufacturer and closer to Singapore; Telagawangi recruited me to Klang.

Mandate

Collect and check shelf prices, stock and sales estimates from a fixed shop panel.

Outcomes

• Audited around forty shops a fortnight in Johor Bahru and western Singapore
• Introduced a closure and change-of-owner log for panel shops
• Rebalanced the Johor sample after seventy shops left the panel in 2000

Decisions in context

I began with a clipboard. Every fortnight I visited about forty provision shops and minimarts, counting stock, copying shelf prices and estimating sales from deliveries. The forms assumed a shop's price was its label. Many owners priced by relationship, charging regular customers on account less than passers-by, and the form had nowhere to record that. As a panel analyst I checked the returns from 1,150 shops and learnt how easily a sample drifts. When seventy Johor shops left the panel in 2000, mostly through closures and changes of owner, district estimates moved before anyone noticed why. I introduced a simple log of closures and new owners, which my manager adopted across the panel. I left because I wanted to see the decisions our clients made with the figures, from inside a business that sold to these shops.

Reason for the move

Moved to Jakarta to work inside a distributor rather than measure the trade from outside.