The problem with agile publishing

This new publishing model – with a focus on user feedback, short lead times and successive iterations of the same product – takes its cue from software development. This is hardly surprising, since that's what a large part of book publishing is turning into.

But is it right for book publishing? Is it even right for software development, for that matter?

It's normal for successive iterations of software to be released in stages. Often, this relies on a community of dedicated beta testers, happy to put up with the various flaws (and even find fixes for them), just to get their hands on something new. The software itself is easily updated, as patches and new versions are released.

This is all very well. But I think there's a much greater tendency for users to show patience with small companies or with open source developers than they are with multi-million pound corporations. You only have to look at some of the reaction to Microsoft's upgrade from Hotmail to Outlook (purely as an example of how users can turn against a big company – I can't comment on whether the grievances are genuine).

And when it comes to book publishing, I think we have to be especially careful. People are already asking whether we need publishers. I happen to think that we do. But the reason we need them is that they provide exceptional quality control. We know that a well-published book from a good publisher has been through some painstaking processes to ensure that it's written, structured and presented in the best possible way. If it's a reference book, we feel we can trust the information in it.

A big part of agile publishing seems to be the idea that it doesn't have to be perfect first time. But I think an equally big part of book publishing is the idea that it does.