I'd love to see some proper data on Goodreads usage. Or, more specifically speaking, how many people use it, how frequently they use it, and how closely it approximates the actual reading population. Generally speaking, a platform of this size probably reflects 'real readers' better than most others. But, again, who knows?
What's particularly interesting is how Amazon haven't changed Goodreads since they first got it. We're not seeing Goodreads reviews or ratings ported over to Amazon. We're not seeing Amazon ads all over Goodreads. It may be that they're doing nothing at all with it, just squatting on the site to prevent a publisher takeover (unlikely). Or, given Amazon's propensity for data mining, they're studying the hell out of the recommendations, lists and shelves, with an eye of improving the Amazon algorithm.
And that's the thing with Amazon, the one 'weakness' - it is crap to browse. There's no serendipity; no inspiration. Users go to Amazon with a product in mind. The purchasing decision has already been made elsewhere; Amazon is just (let's face it) the best means of fulfilling it. The closest approximation to browsing that Amazon can provide is tactically-targeted impulse sales - increasing the size of the basket - the user's buying something, so what else can they be convinced to pick up at the time?
One means, which Amazon already exploits, is by bundling - 'buying x? Get y, and save 30%!' The other, which Amazon is desperately keen to improve, is the whole 'if you're interested in x, you'll probably like y' algorithm. And that's where sources of data like Goodreads are invaluable. And if you consider the volume of purchases made on Amazon every day, just being able to improve this 'nudge' by 1% - or half a percent - would mean a lot of money.
Amazon aside, what's actually interesting in the chart above? Goodreads' use as a book finding platform is, at best, tertiary. Certainly 13.9% isn't a small number, especially at volume. And it'd be foolish for publishers and authors to ignore it. But 'finding new stuff' is a distant bronze after 'bragging about the stuff I've finished' (25.3%) and 'telling the world what I'm already doing' (22.5%).
That's probably a scathing indictment of online culture as a whole - we're much more interested in telling people what we think than reading the opinions of others. But, aside from disdaining the downfall of society, as a matter of blogger/reviewer survival, this is a scary stat. If the Goodreads math is indicative of reviewing culture as a whole, there are twice as many people giving opinions as reading them. What does your review offer that makes it worth reading?
All that from one survey... imagine what Amazon's doing with everything.
[Aside: a separate look at Goodreads also went up this morning - Mark Lawrence comparing sales data to reviews. All the normal sample size caveats apply, but it makes for an interesting study with a very compelling conclusion.]