UI, UX & Web Analytics
1 March 2013
Gabi Ben- zion
When planning a new system and designing a user interface, much thought is invested in properties that will contribute to the success of the interface such as planning its structure, defining the architecture of the information and a navigation method, considering technological constraints, prioritizing the information relevant for actualizing the business goals etc. Also, different examination tools are utilized such as need identification surveys, usage checks etc. Yet since much has already been invested in planning the interface, an informed decision has already been made and a suitable interface has been chosen, we occasionally find ourselves resting on our laurels due to misconception that since we have chosen well we should leverage other aspects of the system before it airs.
It is important to remember that although good planning can predict user behavior fairly well, even the best planning cannot simulate actual user behavior in practice. Therefore, even after a certain interface has been assimilated we must utilize the resources in hand for reevaluation, planning, and assimilation in accordance with the system's needs.
Website managing can be analogized to vehicle maintenance, since in order to ensure optimal functionality we must monitor the state of the vehicle's oils, ensure wheel integrity and perform periodic treatments. When dealing with vehicles, examination tools are available such as an oil gauge and a dashboard. When dealing with different websites examination tools are available as well, in the form of statistics for monitoring and examining user traffic in the website.
In many projects, the main focus wasn't on establishing the system and positioning it as quickly and optimally for the organization and sometimes during the establishment stage the whole theme of analytics is neglected. The lack of a statistical database is usually realized during the assimilation and evaluation stage. That said, most platforms have a built-in statistic ability and in many cases external statistic tools can be added (e.g. Google Analytics or Cardiology added to SharePoint etc).
Through a simple analysis of the amount of visits and views in the site's different areas (Visits/Views/Unique users) we can find the "hot" areas in contrast to the areas less attractive to the user. We can also change the location and content in light of organizational considerations (what I want to highlight) or functional considerations (if we're searching for a certain subject, maybe we should project it in order to assist the user). This information can be used to locate "Winner Applications" and utilize them to increase usage and assimilate the site. There are advanced examination tools with which we can follow the user's moves on the screen and as a result identify the "hot spots" and leverage them for the organizational needs.
[An elaboration on the difference between visits, views and unique users is presented at the end of this article.]
Another noteworthy feature is Average Time on Site-from which we can learn if whoever reaches the site does so since he/she must (if the site is defined as a homepage) or accidentally (the content "sounds" relevant but actually the redirecting link was misleading). If Average Time on Site-does not exceed several seconds it is highly probable that the site was only a 'transit station', i.e. not the user's real destination. If it exceeds a minute, the user probably did read the site's content and used it.
Sometimes, Average Time on Site immediately following the site's launch is high and months later declines substantially. This can attributed to a decline in usage yet it is also possible that the users have already learned to know the site and so the Time to Target has substantially decreased.
Occasionally, we'll monitor the Bounce Rate or Exit Rate who share purpose with the Average Time on Site and allow us to notice 'fictional users that merely pass through the site. On the other hand, a page with a high Exit percentage can also mean that it is the user's target page and so after visiting it the user saw no need to wander throughout the site. This tool enables identifying what contents are important to the site's user.
The Exit Rate enables us to understand what existing content from the site is important for the user, yet it is common knowledge the great challenge in knowledge Management is identifying subjects we don't know about-including those we don't know that we don't know about, i.e. what the user's are searching for that site doesn't offer. Therefore, an excellent statistical tool for this need is analyzing the search words and search results. It is possible to identify how many successful/unsuccessful searches were conducted in order to identify if the failure to find the information is due to a lack of information or a lack of accessibility and correct taxonomy. When we can identify the users' expectations from the site, we can improve the interface and UX by fulfilling these expectations.
Other statistics that can shed light on UX and possible improvements of the system:
Clicks to Target: how many pages did the user pass through before reaching his destination?
Input Devices: which devices/operational systems/browsers were used to access the site? This data enables identification as well as adapting the interface to target audiences (such as a specific cellular device or a prominent browser).
Growth Rate: does the site serve as a format for Knowledge Reflection or Knowledge Sharing? Is the site growing or does it remain outdated?
It is important to remember that statistics are not used exclusively for assimilation and usage reassurance; rather an informed statistical analysis can be utilized in order to identify user behavior and draw operational conclusions regarding user experience and required improvements in the user interface accordingly. If when planning the system we proactively approached the user with a question such as "what would you like to see in the site"-when launching the site we can answer the question ourselves based on examining the behavior of the masses and not the answer of individuals. This is a surer recipe for a successful UX and user interface.
*The difference between visits, views and unique users:
Views: any refreshment and reloading of a page/item counts as one view
Visits: a visit is the user passing throughout the site (number of pages) till exiting/closing the browser (visit= 1 session). One visit will usually include several views.
Unique users: each user is only counted once, regardless of the amount of visits and views he/she performed.