Google's Chief Designer Quits

Wednesday 25th March 2009

Google's Chief Designer Quits

Google’s visual design leader Douglas Bowman, has left Google recently expressing frustration with Google's reliance on performance data to inform design decisions.

Google makes design decisions by analysing performance data and user behaviour. This means that often miniscule aspects of site design can involve an arduous testing process.

For example, Bowman tells that when Google had difficulty choosing between two shades of blue to be used in a recent design, the company eventually decided to test 41 different shade to see which one performed best.

The importance of statistical data

Google has an extremely analytical and scientific approach to every aspect of thier online busines, from design through to software functionality. Google takes decisions based upon detailed statistical data, including analysing user behaviour in detail. Although Google takes this approach to its logical extreme, the benefits of this approach are obvious.

Getting your message across

The key goal of the majority of commercial websites is to get the user to interact with the operators of the website, whether this be by filling in an enquiry form, or placing an online order. Without careful analysis of your statistics and comparitive testing, decisions about website design and functionality can only be based upon instinct and guesswork. The skills and experience of a great designer are hugely important (a good designer will know all about user interface design and how to commuciate through design). However, getting some cold hard stats can be even more important.

Web Design Example: The comparative test

One good exercise is to trial more than one version of your homepage or advertising campaign landing page.

Create two different page designs, and set your site up to randomly serve the different versions of the page to different users. This can be done in a variety of ways: Google even provides a framework doing this in their Adword advertising system.

Make sure you record which versions of the pages were delivered and track user sessions so that you can then analyse which version of your landing page generated the most enquiries or sales. From this, you can then begin to make some more informed decision about what works and what doesn't. If you think you know why A performed better than design B, the next logical thing is try another design, which incorporates the elments from design A which worked, but then refines or alters other elements of the design, then try running A, B and C side by side for a while to see if your instincts were right.

By doing this, we can build up a picture of what will work best for your online business. This practice can also be carried through every aspect of your site design and functionality. Here at Invent Partners have seen overnight improvements of a 100% increase in online business as a result of good design decisions. We have also seen business halved as a direct consequence of a bad design decision. The key here is to spot the trend using good statistical analysis.

Invent Partners have a wealth of experience in the area of user interface design and analysis. We can help you to report on your website performance in a more intelligent way. Drop us a line below:

{form enquiry_form}