The central line

When James Daunt became managing director of Waterstone's the last thing I expected to read, just a few weeks later, was that he was beginning "to move from local to central buying". Could the founder of the chattering classes' most loved bookshops really be adopting a system which a few years earlier prompted one of the biggest postbags of complaints the OFT and Competition Commission had ever seen? Come back, Scott Pack, all is forgiven.

Having started my career as a central buyer I fully appreciate that it's a system with major drawbacks. It improves control, and crucially puts the purchasing power of the chain in the hands of one individual. But at its worst it delivers a lowest common denominator range which rides roughshod over local tastes. Who could forget the first UK Borders, with many more baseball titles than football titles?

Much better, surely, to use that "army of knowledgeable retailers more than capable of editing the almost infinite selection of books in print down to a set of recommended reads", as Ian Vince recently described Waterstone's booksellers. Branch-based buyers cared about the titles they bought, and therefore sold more copies. Waterstone's retained higher calibre booksellers than the wages merited, as it was an interesting, fulfilling job. That produced a better customer experience. There was also an inbuilt democracy; just about every book—no matter how bad—was subscribed somewhere, and exposed to the risk of sale.

But giving buying decisions to individual stores is no panacea either, often leading to wildly different outcomes. Decisions can be based on a somewhat anecdotal understanding of the local customers, but rarely with the data, analytical tools and objectivity to make the most of that insight.

Details are yet to emerge of precisely how Waterstone's new central buying will work. But as I reflect on decades of directly (or indirectly) running retail buying there was one hybrid system that combined the best of local flair and insight with central expertise and purchasing leverage: the W H Smith "TCR" (Travellers Call Requisition) system, still used in the 1980s. Reps visited individual stores to show forthcoming titles to book department managers.

The aggregate of these "orders" then informed the central buyer's decision on whether, and how many, to buy centrally. If the title was bought centrally the provisional branch order was overridden by the central scale out. But for the many titles not bought centrally, branches simply received their original order. It was administratively cumbersome (who could forget the endless piles of pink slips?) and had very muddy accountability for stock management, but it allowed a cadre of knowledgeable, enthusiastic branch staff to have influence (and to feel enrolled) in central decisions.