For assessing performance we had google analytics and CrazyEgg tracking installed on the website. Since we were interested in how people behave in between application dates we selected an appropriate range to take the data from. Any data outside the application dates could cause users to behave differently.
Firstly I look into audience information such as location, device type and new vs returning users. This data is used to show the client the type of traffic that is passing through the site. From looking at locations we can see where the most applicants are applying from. With device data we can see how many users are on laptops and mobile phones etc.
Acquisition data tells me where the traffic sources are coming to the website from. On this particular project I could see how social media referrals were very low. As this site is for a young audience we suggested that social media could be leveraged to raise awareness of the opportunities available.
Overview data of each page tells me how long users spend consuming information and where they navigate next. The usual bounce rates and exit pages percentages reveal less popular pages. I take note of popular flows through the site and highlight areas that could be used better.
Lots of the Google analytics data is very useful information but when it comes to pinpointing why users do certain things we have to dig deeper, looking where they click and how far they scroll. I do this in the next steps with CrazyEgg and user testing.
Crazy Egg is like a pair of x-ray glasses that lets you see exactly what people are doing on your website.
I use crazy egg to see what google analytics can’t tell me in detail. Users may be spending time on our pages but we don’t know exactly what they are doing. How much of the page do they scroll down? What do they spend most of the time looking at?
The data that CrazyEgg provides allow give me a deeper insight into how the pages are being used. Individual clicks are recorded and show me exactly where users click. The thales site contains many elements that can be interacted with. I started to see a pattern in which users were clicking objects that weren’t interactive. One problem was some of the visuals looked like buttons used around the site. The scroll map showed me how after a certain call to action was passed on the page users did not scroll any further.
This data was very helpful. in the last stage of user testing. I started to identify where potential problems were and began to write tasks, for the user testing session.
User testing sessions
One big help I had when planning and undertaking these sessions was a book called Rocket surgery made easy by Simon Krug. The book is described as ‘The do-it-yourself guide to finding and fixing usability problems.’ It covers how to approach user testing from a pretty lean point of view and non nonsense approach.
I already knew the things I wanted to test from the data collected in the first two steps, plus a few more obvious things. I ran 4 sessions on a range of users with screen and face recording during the tasks.The results from the user sessions helped me back up statistics from the other data collected and reveal new usability problems. I could now start suggesting solutions to any of the problems or goals for the next phase.
I threw together a report of my findings and presented these to the client. We we’re then able to discuss the quick wins and more detailed changes. These types of usability reports give solid evidence to clients as to how their sites are performing and what can be done to increase engagement.