The numbers’ game


Perhaps it’s a coincidence, perhaps a conspiracy. Perhaps there’s something in the west-coast water supply but the last week or so has seen a pretty sophisticated debate break out on the use of facts and intuition in design.

  • Douglas Bowman is leaving Google, having been heralded as one of the best things to happen to the search giant when he arrived in 2006. To make matters more interesting, Bowman has blogged about his reason for leaving – an excessive Google reliance on data-driven design (well, data-driven everything)
  • Google themselves are flogging a new product to help everyone else become as data obsessed as they are, bringing reasonably complex multi-variant testing to a car dealership near you. This is in a market which is rapidly waking up to the user of performance tuning and management
  • Everyone (and especially the Twitterati who might be expected to love it) hates the new Facebook design. Unlike v2 which was introduced in a vaguely consultative way, v3 has been foisted on all users rather suddenly, seems most to be ‘inspired’ by Twitter (which is frankly a very different beast), and jettisons many of the best features of the site. This has raised a debate about how much companies should listen to users in design.
  • Apparently Steve Ballmer has been off telling people that Microsoft has a challenger advantage in search over Google, who are too set in their ways and conservative
  • At the other end of the spectrum, Apple has released the frankly insane zero-button iPod shuffle. Bizarre: yes. Bold: certainly. Tasteful: perhaps.

So the question is: just how innovative can a business be by relying on data about current user behavior, rather than using creativity and instinct to come up with new things that people don’t even know they want yet. Bowman’s disquiet about Google is that the company would (for example) statistically evaluate the colour of borders or the size of buttons, making his role as head of visual design somewhat redundant but also making his aspirations – as someone who is looking to lead the market through visual design – impossible.

I think this says more about Bowman’s expectations (quote: ‘change the world a few million users at a time’) than anything that surprising about what Google does. The experiences Google are trying to enable are all about using great technology to build perfect mousetraps. The core ideas can be expressed in just a few words (“get the correct results from any search”). And I think it’s true that usability has always been much more important than visual design for the business. And, Google has always known that the user experience is as much governed by performance and quality as interface. With the exception of Chrome, all of their products have been almost deliberately ugle, liberating them to focus on function.

I’m sure too that there’s immense chortling at Google HQ about the idea that this focus on statistical research is stifling the multi-billion pound Google business, and equally that Microsoft is likely to fly right past them in terms of search. Google  is synonymous with search, it is the generic, and that has everything to do with what goes in and what comes out, and very little to do with how it goes in and how it comes out.

A similar reality is starting to exist in every other area of utility-orientated computing.

Facebook has done the opposite. Zuckerberg-centred design pervades with – I presume – the rest of the user experience team bowing not to statistics but to the great one and his views. Of course, time – and numbers – will be the judge, but it seems odd that Facebook has thrown out many of its best features and endorsed the micro-blogging format, just as that market become more competitive and challenging.

What Facebook appears to have lost, is its principles. Google’s area is straight-forward: ability to find, or ubiquitous access. Facebook’s used to be about enabling connections in groups. What are they now?

Whilst Microsoft’s ambitions for search seem unrealistic, there is a strong case to make that this unlikely candidate has the best overall approach to design.

Creating some of the most complex products in digital (Windows and Office), Microsoft has found a way to combine imagination, principles, many types of user research and engineering to produces fantastic products. I’ve posted the Jensen Harris Mix presentation before but it remains well worth a watch. An understanding of the core user behaviors gave rise to an overall framework for the application, which then used ethnography, user research and the amazingly detailed data from the customer experience programme to really find out what people do. Do we need a ‘save’ button if everyone uses CTRL-S? In fact we do because many actual users don’t use quick keys. Customization may seem like a neat solution to a complex problem, but we actually learn that only a tiny proportion of users ever turn it on.

Similarly, the Windows 7 team has done a great job (yes, an overdue one), of informing their design decisions through detailed understanding of customer behavior, but without just asking users to do the design for them. The team’s detailed analysis of problems (and again, understanding of the role of performance) is rigorous and inspiring.

Should data be at the heart of your design strategy? Yes, but it shouldn’t be the heart of your design strategy. The heart has to be the principles, and the team must believe that inspired thinking can change the game about sticking with those principles and achieving objectives. The fact that Google’s inspired thinking has almost all been in technology and architecture is besides the point. The fact that Facebook’s solution is wrong is not because they’ve ignored users, it’s because they’ve ignored users’ motivations.