What agri-food can learn from the keiretsu supplier model

Buying and selling in food often follows a simple script. The buyer calls several suppliers, takes the lowest price that meets the specification, and calls again next season. Suppliers know the rules, guard their costs and expect to be dropped the moment a cheaper offer appears. The script is clear and keeps prices honest. It also leaves a lot of value on the table, because neither side has much reason to spend money on the other.

Japanese manufacturers, especially carmakers, built a very different model over the second half of the twentieth century. It is usually discussed under the word keiretsu. This page explains what the term means, how the supplier version worked, and what parts of it might travel to farms, processors and retailers.

What keiretsu means

Keiretsu is a Japanese word that roughly means a series or a lineage. In business it describes groups of firms tied together by long relationships, shared shareholdings and regular dealings. Two kinds are usually distinguished.

The horizontal groups grew out of the large family-controlled combines that were broken up after the Second World War. Firms that had once belonged to the same combine drifted back together, this time through banks and mutual shareholdings rather than family control.

The vertical groups are the ones that interest supply chain managers. A carmaker such as Toyota sat at the top of a pyramid. First-tier suppliers made whole systems, such as seats, brakes or electrical parts, and they in turn bought from second- and third-tier firms making smaller components.

How the supplier model worked

The features that drew attention from Western managers in the 1980s and 1990s can be laid out in order, from the start of a new model to its production run.

1. Few suppliers, chosen early. Instead of putting every part out to bid, the carmaker picked one or two suppliers for each system and brought them into the design work well before production.
2. Target costing. The carmaker set a target price for the finished car, worked back to a target cost for each system and asked the supplier to help meet it, rather than simply asking for its lowest bid.
3. Shared engineering. Engineers moved between the firms. Suppliers suggested design changes that saved money, and the carmaker sent staff to help suppliers improve their own plants.
4. Steady price reductions. Contracts often assumed that costs would fall each year as the supplier learned. Savings were split between the parties rather than seized by one.
5. Long horizons. A supplier that performed well could expect the next model's business too. That security justified investment in tooling, training and new methods.
6. Supplier associations. Suppliers met regularly as a group to share quality and productivity practices, under the carmaker's guidance.

The result, by most accounts, was a system that combined lower costs with high quality and fast development. It was not built on trust alone. Carmakers kept a second source for many parts and rated suppliers closely, so the threat of losing business never vanished. But the default was to fix a problem together rather than to switch suppliers.

The contrast with price-only buying

The difference is less about friendliness than about what each side is willing to invest. Under adversarial buying, a supplier will not spend money on equipment or know-how that suits one buyer, because the buyer could walk away next season and take the benefit with it. Under the keiretsu model, the expectation of a long relationship makes those investments sensible, and the buyer shares information that a price-only buyer would keep to itself.

An illustrative example from food

Consider an illustrative processor that buys 50,000 hogs a year from 20 farms. The numbers are round and invented.

Under price-only buying, each farm sells to whichever plant pays most that week. The processor gets carcasses of uneven weight and leanness, and loses perhaps $3 per hog in trim, missed specifications and scheduling problems. That is $150,000 a year.

Under a longer-term arrangement, the processor shares carcass data with each farm, agrees on genetics and target weights, and pays a premium for hogs that fit. Farms spend more effort on sorting at shipping. If half of the $3 is recovered and split evenly, each side gains $37,500 a year, and the farms gain steadier sales besides.

The same logic can run between a processor and a retailer: shared sales data, joint planning of promotions, agreed packaging that suits both the plant and the shelf. Readers interested in how such chains are organised will find more on the value chain hub and under value chains.

Limits and risks

The model has real costs, and Japanese experience in the 1990s made some of them plain. Long slow growth exposed groups whose cross-shareholdings and loyal dealings had protected weak firms. Some companies began to sell shareholdings and trim their supplier lists, and in 1999 one large Japanese carmaker, after taking on a foreign partner, announced plans to cut its supplier base sharply.

For agri-food, four risks stand out.

• Lock-in. A farm that invests to suit one buyer may have nowhere else to go if the relationship sours.
• Power imbalance. A large processor or retailer facing many small farms can capture most of the savings, which turns partnership into a polite name for dependence.
• Thin markets. If most animals or crops move under private arrangements, fewer trades happen in open markets, and the public prices that everyone else relies on become less reliable.
• Competition law. Close coordination between firms, especially competitors, can cross legal lines on price fixing or market sharing.

None of these rules out closer supplier relationships. They do suggest that the parts of the keiretsu model worth borrowing are the habits, such as early involvement, shared data and joint cost reduction, rather than the cross-ownership and the closed circle. Quality is a large part of what such partnerships deliver, which links this topic to the wider question of quality management in food businesses.

A good test for any proposed partnership is to ask what each side will put into the relationship that it would not risk on a one-season deal, and how the gains from that investment will be split. If neither answer is clear, the relationship is likely to fall back to price-only buying at the first sign of strain.