April 20

If web analytics are not being used on your UX project you are missing a piece of the puzzle

By: Chris Gray

For mine, at the heart of UX is making informed design decisions. By informed I mean decisions which reflect the actual needs of a particular audience. Web analytics can be a crucial input for making these informed and evidence based decisions. In fact I would argue that practitioners are remiss if they are not including analytics as part of their UX process. This post covers why all UX projects should utilise web analytics.

Web based projects are exceedingly fortunate in that they are typically an update to an existing site and most existing sites have web analytics data. As such we have a rich and full data source available to utilise. Industries such as retail spend significant sums of money on activities such as mystery shopper studies and other market research to understand how their customers interact with their business. In the online game we are exceedingly fortunate in that we are able to measure most of our customer’s online behaviour and engagement with our online business.

A key reason for utilising web analytics data is because it is relatively low cost and high value, particularly in comparison to other user engagement practices such as interviews, ethnographic studies, surveys, focus groups and user testing.  Furthermore, many of these UX research practices simulate the very behaviour that we can measure directly with analytics. For example in user testing we create scenarios or tasks to replicate the typical activities customers attempt on a website whereas with analytics we measure this actual behaviour. Tools such as the 4Q survey even allow analytics data to be analysed in conjunction with purpose of visit data. This is not to suggest using web analytics instead other UX research activities rather it should simply be used more often in UX projects to compliment these other methods.

Web analytics provides data to ensure that decisions are made based on evidence rather than design by committee or gut feel.  All UX practitioners can relate to the impact these decision making processes can have on a project. On a recent project we changed a label recommendation based on keywords data from within Google Analytics which suggested that a large proportion of the client’s audience were in fact searching for the technical term which we had previously assumed was too technical for their audience.  We were able to use the data to conclusively substantiate our final decision.

The main reason to use web analytics data within your next project is to design a better site. It provides an exceedingly valuable source of information which can assist in decision making throughout any project. Data on important content, user paths, search terms used to find the site and to search on the site, as well as common landing and exit pages paint a rich understanding of user behaviour to draw upon for design decisions.

The current maturity of the industry is a further reason to make web analytics a key input within the design process.  The industry appears to have reached a degree of maturity where accountability and the ability to justify key design decisions is paramount. This is a level of maturity where research with actual users is no longer considered indulgent but rather a crucial stage is designing an interface. This is also a time where senior management expect rigour behind any design. It is with this backdrop that it’s essential that we use the most appropriate information sources we can get our hands on and augment our qualitative findings with qualitative data.

In short web analytics helps to design better websites. As UX practitioners and website owners we need to better leverage this fantastic information source as part of our practice. This is not to suggest that it should be the only information source, in fact to make the most of web analytics it should be coupled with other UX research methods to better understand users. This is also not to suggest that analytics data is without its shortcomings. However these are topics for another day. Consider making better use of analytics on your next project.