There has been a lot of discussions lately around web analytics and how it ties in with UX. Whereas before the analytics team and the UX team were separate from each other, now the analytics team has joined in with product managers, business analysts, designers and developers in an ever growing team to complete shape a full UX project.
Web Analytics allows businesses to track and measure details about customer behavior, showing key areas of traffic on your website. This helps businesses compare specific dates in order to identify trends and analyze improvements or declination of certain areas of their websites. Below is an example of analytics to use that tie in with UX.
- Key pages users are visiting on your website
- What devices users are using to access website
- From which countries users are accessing your website
- How much time users are spending on your website
- Bounce rate – i.e. where users land on the website without interacting with any other page and exit your website.
I recently worked on a large-scale project within the financial sector where the client did not have the budget to do any user research upfront, but was open to do user testing after the wireframes were created. In my opinion this is a big mistake! I have always been an advocate of conducting research early on in a project even if it was on a low fidelity prototype. You then have the basis of usability issues that occurred during the user testing session. How do we come to the conclusion for the wireframes/mock-ups without users input?
In this project, we conducted stakeholder workshops with the Digital team to understand the business process, objectives and gathered background information from user research previously conducted. To gain more insights we decided to look at the web analytics. We analyzed key pages that users had visited in order to determine the key user journeys, for conducting a competitor analysis. This was done to investigate how easy it was to access the same journeys on competitor websites.
The problem with large-scale financial websites is they have many pages (in this case over 1000) and you are not going to wireframe every single page to start with. So analytics is a great way of showing you the key pages of the website. All websites are made up of a series of customer tasks; some will be the most important (top) tasks whereas others will be tiny tasks (this is known as the ‘The long neck vs. long tail process’). Unfortunately many clients spend more time concentrating on the tiny tasks rather than the top tasks
Recommended for YouWebcast: The Art of Growth Hacking: Gaining Early Traction by Doing Things that Don't Scale
For example, if the website had 1200 pages but only 30 pages are the most visited, (top tasks) these are the pages that need to be used to form a good Information Architecture. These pages need to be easily accessible or the route needs to give users a seamless experience. These pages are known as the ‘long neck’. The other 1170 pages on the website are known as the ‘long tail’ pages. These pages are still significant and do need to be kept in mind when looking at user journeys and content on the site. However if you have to prioritise then the ‘long neck’ pages need to come first.
How can analytics help with UX?
Analytics is quantitative data and would be looked at right at the beginning of the UCD/UX process along with the requirements gathering before you start the qualitative part of the process. It helps you build a picture of the ‘What’ before you investigate ‘Why’. You can’t know why things are happening if you don’t know what is happening.
Analytics also helps connect the UX findings and recommendations to business impact, which is primarily what stakeholders, are interested in.
For example, users are confused when attempting to locate the ‘create an account’ button. Your recommendation (based on qualitative feedback) is to make the button more prominent. This would most likely be a priority and get changed.
However if you combine this recommendation with analytic findings e.g. ‘Users are confused at locating the ‘Create and account’ button and 30% are abandoning the process, resulting in an estimated revenue loss of £120 000′.
Stakeholders would hear the statistics and would want improvement on the revenue loss. This recommendation would bear more weight. Stakeholders would be more open to conducting a series of usability tests to find out why users are abandoning the process. This would ultimately lead to more satisfied stakeholders while still balancing the user needs. Business needs are key analytics can help persuading stakeholders to invest in UX.
Web analytics are great at measuring and monitoring how well a website is performing at meeting the organisations goals, but it only gives you half the picture. Matching web analytics with qualitative customer insights will give you a much broader and deeper understanding of customer interaction.
Analytics will answer questions like ‘what’, ‘how’, ‘where’ and ‘when’. UX research and testing can tell you the ‘why’. Combine them and you will have the whole picture.