Web analytics – Use in improving usability

Spotless  | 

Summary

Web analytics gives a unique insight into users’ real world behaviour. Important issues include: value scores for all KPIs; early warning systems; A/B testing alternative designs; setting targets based on past performance.

Web analytics – Introduction

Web analytics is the collection of data from the real-world use of a website – generally, where people arrive from and what they do once they are on the site. Web analytics offers a unique source of information to usability experts in helping to identify and diagnose usability issues. This is because web analytics gathers data from real-world users as they are using a site, for their own personal reasons.

Value of web analytics to usability experts

Web analytics’ value comes from its ability to actually show what people are really doing on a website within the ‘real world’. No other method of getting users’ feedback on a site can really make the same claim. Even the most fervent supporters of usability testing must admit that placing participants within an artificial testing environment, with the full knowledge that they are being observed/recorded, poses significant challenges when seeking to gain an insight into their actual motivations and preferences.

This is not to say that web analytics will ever fully replace usability testing. Usability testing will always be valuable for the in-depth investigation of particular usability issues. Web analytics is, however, playing an increasing role within the field of usability because of its ability to provide an insight into genuinely ‘real world’ behaviour.

Popular metrics in web analytics

The field of web analytics is continually evolving; however, certain metrics have proven perennially popular. These may not always be the most useful metrics within the field of web analytics, but we should certainly be aware of them:

  • Conversions: The most important web analytics on a site are often the number of forms completed, brochures downloaded and/or products sold. Finding ways to increase these numbers often drives much of the web analytics and usability work on a site.
  • ‘Pages per visit’ & ‘Time per visit’: This metric is amongst the most difficult pieces of web analytics data to interpret – a short visit could be due to the user quickly achieving their goal, or their quickly abandoning it. Equally, a longer visit could result from an engaged user finding lots of interesting content, or a user becoming hopelessly lost and confused. In truth, a high number of very short visits are probably a bad sign for most sites.
  • Bounce Rate: The ‘bounce rate’ shows what percentage of people leave a site immediately after having arrived. Most web analytics tools allow you to determine the ‘bounce rate’ across the whole site and also at a ‘per page’-level. In general, a high bounce rate would be considered a warning sign of poor usability and/or poorly-qualified traffic.
  • Exit Pages: Identifying where users are choosing to leave your site may help you to understand how to encourage them to stay. A key recommendation for ‘exit pages’ would be to use your web analytics tool to track the ‘exit rate’ for pages where your organisation absolutely does not want users to be leaving the site (such as form or checkout pages, for example).

Although the above data is interesting in isolation, the real value of web analytics can only be delivered by carefully considering its data in relation to a site’s overall goals and the business value they deliver. It is also the case that web analytics can supply a very large amount of data on almost any topic one could wish, so finding a way to focus one’s attention is very important.

KPI scoring in web analytics

When deciding upon how to use web analytics in order to improve a site’s usability, it is essential to fully define and prioritise a site’s KPIs. For example, the web analytics KPIs for a site selling electrical equipment might include:

  • Product page viewed
  • Product video viewed
  • Product image enlarged
  • Add to basket
  • Completed Checkout

The important factor is that all valuable and desired customer behaviour (whether or not it involves filling out a form and actually ‘converting’) is captured, defined and prioritised within the web analytics KPIs. In addition, this web analytics approach should assign each KPI a relative ‘importance’ score in order to quantify its business value, for example:

  • Product page viewed = 10 points
  • Product video viewed = 15 points
  • Product image enlarged = 15 points
  • Add to basket = 50 points
  • Completed Checkout = 100 points

Assigning relative scores to a site’s web analytics KPIs will allow everyone in an organisation to clearly understand their relative business worth. This is vital in allowing web analytics and usability professionals to direct and prioritise their efforts. It will also enable more effective measurement of the value of individual traffic sources.

Web analytics and usability

Web analytics can be used to support usability improvements to a site in a number of ways. These are some of the ways in which web analytics can be used to help usability professionals:

  • Early warning signals: Web analytics can be used to identify possible areas of concern which can then be explored through other usability methods (such as usability testing, for example.) If a site is seen to have lots of ‘short visits’, for example, the web analytics might be used to identify where on the site these visits typically occur. Usability testing might then be used to investigate the reasons behind this user behaviour.
  • A/B testing: Many web analytics packages allow for a site to test two alternative versions of a page against each other. A well-designed test will be able to test different usability designs and ideas against each other (in terms of their performance on a number of pre-determined KPIs). Such an approach allows web analytics to directly support usability and design decisions.
  • Targets: Measuring past site (and/or digital media campaign) performance through web analytics allows benchmarks to be set. A major benefit of web analytics is not only that is can report on current performance, but also compare performance over time. As such, setting targets based on past web analytics data can help provide context as well as an impetus for improvement.

There are, of course, many other considerations involved the use of web analytics in supporting site improvement. However, we believe that those mentioned above will allow an organisation to gain a great deal of benefit from their web analytics.

Summary

Web analytics’ value to usability professionals comes from its ability to actually show what people are really doing on a website within the ‘real world’. In this regard, web analytics has an advantage over almost any other usability method.

For an organisation to get the most benefit from web analytics, it is essential that it fully captures, defines and prioritises its web analytics KPIs. These web analytics KPIs can involve anything from viewing a video to completing a purchase. All web analytics KPIs should have a relative score that allows their value to be directly compared.

Some of the main ways that web analytics can benefit usability professionals is through: providing early warning signals that can be further investigated (such as lots of short visits to a site); supporting and reporting on the A/B testing of alternative designs; setting targets based on past performance.

Got a project in mind?

Danny, our Client Services Director, is on hand to answer your questions.

Related insights.