It’s important to understand what your users do, how they use your product… but remember: data never paints the whole picture. User research, user testing, usability studies, behavioural analysis, usage statistice: it’s all important – understanding and making sense of this data is crucial for your product’s development – but this data doesn’t exist in a vacuum, and it’s not enough to (alone) drive a product.

An example: The first iMac was released in 1998 and apart from looking really funny it was the first personal computer sold by default without a floppy disk drive. There wasn’t even a place to build one in… you had to buy an external one and connect it with a usb cable. It caused an uproar – analysts and journalists called them crazy – you can’t ship a computer without a floppy disk drive!

When was the last time you even saw a floppy disk?

If apple had have run a survey with all their customers before releasing the iMac and asked the question: “would you expect a new computer to have a floppy disk drive?”, I’ll bet most of their respondants would have said ‘yes’. But product and market innovation doesn’t happen through user surveys or focus groups, and it doesn’t happen through analytics… It comes from visionaries who challenge expectations.

I had a similar conversation with a colleague recently as we were discussing including a new feature in maps.ovi.com. A couple of days after the discussion he came back to me and said: “we checked the application logs and site analytics, and we see very few instances of user behaviour that indicates that people want that feature, so we conclude that it’s a not feature that people want.”

Well of course there are not many instances of this behaviour in the logs. If your product interface doesn’t support a particular functionality (that is, if you don’t tell your users about it), how should they know to try it? And if someone did think to try it, and it doesn’t work, how many times would they try before they gave up?

User analytics is extremely important for optimisation and product evolution, but don’t confuse optimisation with innovation. Innovation is something else. Innovation challenges the analytics; it challenges the market, it challenges expectations and it challenges users themselves.

Do you innovate or optimise? To be remarkable, I think you need to do both.