# The thin centre *The art of leaving well alone* --- Most acquirers that last are built on an operating system: a central machine that makes every business they buy a little better. Danaher is the great example, with a house method it runs through everything it owns. There is a quieter design that does close to the opposite. It keeps the centre tiny, lets the businesses run themselves, and treats deciding where the money goes as the only thing the top is truly expert at. It has been worked out independently more than once, on different continents, by companies that did not copy each other, which is the strongest sign it is a real design rather than one firm's luck. --- **It was invented at least twice.** In Chicago from the early 1980s, Illinois Tool Works under John Nichols grew by buying industrial businesses and deliberately keeping them small: larger acquisitions were split into pieces, each unit kept its own customers and its own P&L, and by 1996 ITW was running 365 of them. At almost the same time in Stockholm, [[Bergman & Beving]] was holding its group of autonomous trading businesses to a single ratio. Neither copied the other. The design has been rediscovered since, by the Nordic groups that followed (Lifco, Indutrade, Addtech) and by allocation-led American groups like Roper. **The centre stays small on purpose.** Lifco runs on a head office of fewer than twenty people. Indutrade is much the same, mostly finance and acquisitions, with full operating control left to its two hundred-odd companies. The businesses keep their names, their managers and their own profit and loss. The centre is good at two things: buying well, and choosing where the cash goes. Everything else is kept out of the middle on purpose, because the moment you build shared services and a group-wide playbook you have started to become the other kind of company, the kind this design is defined against. **The whole skill is allocation.** A business like this lives or dies on return on capital, and it judges every acquisition and every reinvestment against a hurdle it will not talk itself below. Buying at the right price is the entire game, because nothing at the centre can rescue an overpriced deal later: small, durable, niche businesses at sensible multiples, left to throw off cash, with that cash going into the next one. Run for long enough, the arithmetic does the work. **The design carries its own clock.** It earns its best returns while there is a deep pool of small, durable businesses to buy at sensible prices. As the group grows, each small deal moves the needle less, and the pull is towards bigger deals at higher multiples, which is exactly where the arithmetic stops working. ITW itself aged out of it in a different way: the group that proved the design let the unit count sprawl into the hundreds, decided the complexity had outgrown the benefit, and spent much of the 2010s consolidating those units into a few dozen divisions. Either way the realistic aim is to slow the drift, not escape it. --- The contrast with Danaher is the sharpest nuance: the same acquisitive growth by the opposite design, one improving what it buys from the centre and the other refusing to touch it. Both have worked for decades, which makes them an [[Attractive opposites|attractive opposite]] in their own right. --- Anyone planning to grow by buying has to decide which of the two they are actually running before they scale it. The thick-centre operating system earns its keep only if the centre genuinely makes businesses better. The thin-centre allocator earns its keep only if you are patient enough to buy well and leave things alone. The failure is the centre that is neither, interfering without improving and allocating without care. ---