This is an old, old publishing parlour game and one that certainly hasn't got any easier in the last five years.
However,this weeks report by Bowker is the best bit of news the chain will have had for a while. 45% of undecided purchases, i.e. where the buyer did not know what they wanted to read, were made as a result of bookshops.
However, while it is certainly encouraging news for physical book retail, the big question is how long will that last? Will that be just 30% in two years time? Or does it represent a stable rump? Sadly it is hard to feel that the digital high water mark has been reached.
However, despite the fact that the Internet is in theory a fantastic tool for communication, it's Library of Babel qualities more often than not make it seem like a white noise amplifier. Which is why it sometimes seems it is only those companies with REALLY LOUD VOICES, especially the digital big four of Amazon, Apple, Google and Facebook, that thrive.
And one of the upshots of that is that discoverability is something the Internet is as yet not very good at. The Internet is great at niches - it is great if you know what you are looking for. It is brilliant at giving the illusion of throwing the world open because it makes it easy to stumble across all sorts of weird and wonderful things. It is great at huge phenomena - like Fifty Shades. It is lousy at surprising us with things we actually want.
Just as books were the ideal prototypical product for phase one of e-commerce - the perfect combination of a deliverable size and a connector to key ABC1 demographics, so they are I suspect going to be the key battleground for phase two of the digital marketplace which is going to be all about adding meaning to the connectivity of the internet.
If book discoverability is all about the ability to communicate meaning.
There is a huge section of the book buying market that seriously does want to read dozens of Mills and Boon romances every month - who, having read every word of the Fifty Shades trilogy sets out to read Sylvia Day without pausing for breath. Amazon and Good Reads are terrific for people like that.
However, a huge part of the market lies outside of the genre channels. These are the people who make the really big bestsellers. These are either the huge number of people who maybe only read four or five books a year, these are the people who just like to read widely.
No one expects TV audiences to just watch comedy, or just watch soaps. The mass market intelligently and rightly likes variety.
Of course in the world of books, with their limited marketing reach, that has always been a bit of a problem and it has got greater as the Internet has bitten into the physical retail sector because the Internet is so bad at intuitive, lateral connections between books. It is exactly the sort of thing that bookshops used to be really good at - generating word of mouth bestsellers, handselling: all of those arcane 'discoverability' skills which the book business needs so badly.
But if I was Waterstones I wouldn't hold out much hope that the Internet won't discover a solution to this problem - it probably will and that means it is going to need to find something else to save its bacon unless it can work out a way of being better at discoverability than the internet not simply by default, as it is now, but as an intrinsic part of the way it does business.