Chief Product Officer / Japan
What each product decision gave up
- Group revenue
- ¥48.6bn
- Year to March 2026; commercial and institutional kitchen equipment
- Product-division colleagues
- 212
- Planning, design, certification and service-parts engineering
- Repair-parts period after production
- 10 years
- Five core product families, extended from seven years in 2022
Decisions & consequences
01Lowering the rice line's loading point and giving up a tenth of its output
In 2022 our continuous rice-cooking line was due for its first full redesign in twelve years. Operators loaded stainless pans of washed rice and water onto a conveyor at 1,050 millimetres, and a loaded pan weighed about 13 kilograms. During visits to customers' central kitchens, my planners found that loading crews in several school-lunch centres were mostly over sixty. Two centres could no longer fill the loading position from their part-time recruits. The redesign would set working conditions in those kitchens for fifteen years.
Our sales director wanted the new line to match a competitor's advertised output in the same floor area, because municipal tenders score capacity per square metre. Our largest contract caterer agreed. Engineering proposed the old pan plus an automatic loader at about ¥6.8 million per line. I recommended lowering the loading point to 820 millimetres and reducing the loaded pan to 9 kilograms as standard, with the automatic loader as an option for the largest kitchens. With lighter pans, hourly output in the same footprint fell from 3,600 to 3,250 servings.
Before the design freeze we tested prototypes with twenty-seven operators aged fifty-eight to seventy-four in three centres, timing each loading cycle and changing the rail profile twice. The president approved the recommendation with the sales director's objection set out in the same paper.
The line launched in April 2024. By March 2026 we had sold forty-one lines against a plan of forty-six and lost two municipal tenders, worth about ¥390 million, decided on capacity. Centres with the new line report that the loading position can again be staffed from their ordinary rota. I claim no injury reduction; customers record strain differently, and the period is short. We now publish loading height and pan weight in every tender response.
What I would do differently is test with crews before the capacity figure entered the sales plan. Once 3,600 servings appeared in a forecast, every reduction looked like a loss rather than a choice.
Extending repair parts to ten years and keeping the dearer controller
When I became CPO in 2021, our stated period for holding repair parts was seven years after a model left production. Each product generation also brought a new control board that could not be fitted to earlier machines. Dealers told us that care homes and small restaurants were retiring working fryers and dishwashers because a single board was unavailable. In that year's dealer service records, an unavailable part explained 6.8 per cent of unrepaired calls. At the same time, purchasing proposed a custom controller from a supplier whose component commitment ran for only five years, saving ¥2,900 a unit.
Finance and the parts warehouse objected to a larger inventory that would include parts nobody might order. The sales director pointed out that a repairable machine defers a replacement sale. I recommended extending the parts period to ten years for our five core families, designing one controller family able to replace the two previous generations, and staying with the general-purpose microcontroller and its longer supply commitment. We gave up the cheaper board and accepted that some replacement sales would move later.
The board approved the change in early 2022. Designers fixed connector positions and mounting points across the families, and service-parts engineers wrote retrofit instructions for each earlier model. We supplied wiring diagrams and service manuals to all 140 authorised dealer service companies. Parts inventory rose by ¥260 million over three years, and we wrote down ¥34 million of slow-moving stock in the year to March 2025.
By the year to March 2026, unavailable parts were cited in 2.1 per cent of unrepaired calls. Dealers fitted the new board to about 1,900 older machines. The sales director estimates that deferred replacements cost several hundred million yen a year; I report that as his unverified estimate.
I would set the rule for writing down slow stock at the start, not after the first year's surprise. I would also state each family's parts period in the original sales documents instead of a later service notice.
Holding the connected oven until its record admitted its gaps
In 2024 we were preparing a connected version of our combination steam oven. A network module sent cavity and core-probe temperatures to a dashboard that customers could use within their HACCP-based hygiene records. A restaurant chain was tendering for 180 ovens with delivery from October, and our sales team wanted to describe the product as recording the customer's hygiene control automatically.
Testing in two customer kitchens showed why that wording was unsafe. When the kitchen network dropped, the module kept only two hours of readings. More seriously, the record showed a core temperature even when the probe hung outside the food, reading the air in the cavity. A manager reviewing the dashboard could see a satisfactory figure for a dish that had never been probed.
The president wanted the product launched for the tender, and the sales director argued that competitors made similar claims. I recommended holding the launch until the oven stored sixty days of records locally, marked every gap in the record and flagged core readings taken without an inserted probe. I also rewrote the claim: the oven records its own temperatures, and the customer's hygiene plan decides what must be checked. The quality-assurance director supported the recommendation, and the president accepted a four-month delay once he had seen the test records.
We lost the chain's tender, worth about ¥560 million. The oven launched in February 2025. By March 2026 we had sold 1,140 connected ovens, and forty-one per cent of buyers had activated the record service. Two customers have used the probe flag to correct their own procedures, which I consider the most useful result. Every claim about a connected product now passes a review by quality assurance and an outside food-hygiene adviser before release.
What I would do differently is bring customers' hygiene managers into the claim wording before the engineering freeze. We discovered the probe problem in testing; their experience would have shown it at the specification stage, and the delay would have been shorter.