How to psychologically interpret website analytics

The massive volume of analytics data can confuse and point you in the wrong direction. Here’s how to understand what your website visitors are really doing...

Web analytics can be extremely useful, but it the data can also be a curse. Indeed, online we now live in a world swimming with data, so much so we have to call it “big data”. The problem with big data is it needs big analysis. The more analytics you collect, the more time it takes to wade through the reports and try to work out what is happening with your website.

Google Analytics, for instance, has 103 standard reports you can look at. That’s before you start adding your own specific requirements and comparisons. In addition, there is a problem with all of these reports; none of them tell you “why” a visitor did what they did. All that analytics data can show you is an historical picture of what happened. This is a real problem in marketing because the crucial piece of information you need to know is “why” someone does whatever it is they do on your website. Only by knowing “why”, instead of “what”, can you improve your website to match user intentions, emotions and desires. (See Analytics only provides a partial answer for marketers for more information.)

So, this article will help you use your analytics data to find out the psychological factors involved with your website, to help you get closer to your visitors’ true requirements. You will discover:

• How to zoom in on key psychological issues revealed in analytics
• What data to concentrate on for real behavioural factors
• Ways in which you can use the data to improve visitor experience

Analytics don’t always tell you what you think they tell you

Recently I sat in a meeting with the directors of a large manufacturing firm. They proudly spread out pages of analytics date which confirmed that they were well-known and that people “loved” the website. I inquired as to which parts of the data demonstrated that. They pointed to the keyword data for organic search which showed that almost 90% of all searches were for the company name. Then they showed me the data for the time spent on the site, showing that people were spending four or five minutes at a time.

This is a familiar scenario, the “headline” data can mislead.

In drilling down into the detail, you could see, for instance, that even though people were spending around four minutes on the site, they were doing so by visiting dozens of different pages, all for a few seconds at a time. The “user flow” data showed that people were often going back to pages they had already visited. Meanwhile the “content drilldown” report showed that people were going through several different pages, with very low amounts of time actually spent on the final page prior to exit.

The directors were focused too much on the headline data, whereas the detailed data showed a picture of confused visitors, not knowing where to go, or not finding what they wanted. The fact that almost all of them had arrived on the site as a result of searching for the company name suggests they knew the company and its products and wanted more information about something, but were struggling to find it.

Analyse using narratives

Data analysis can be confusing. In busy offices it is often easy to focus solely on headline data, but this can lead you to false assumptions about a website.

Drilling down into the data helps, of course, but you have to balance that with the time it takes.

An easy way to sort through the data is to create a narrative. Instead of focusing on the detailed data itself, consider instead a human being and write yourself a story about their behaviour, based upon the data you see. Your narrative might be about a person and what they are looking for. Start with the keyword data to see what someone typed in. Then, when you have chosen a phrase you know that someone was sat in front of their computer and typed in those words. Take a look at the page they arrived at having typed that phrase into the search engine and use the data on the keyword to see how long they sat looking at their screen and how many pages they clicked on. In other words, instead of focusing on data and graphs, turn what you see into a story about a user and what they were doing and looking at when they sat facing your website. This makes data analysis easier to interpret because you can visualise what is going on with your website. Once you have the narrative and visual in mind you can more easily decide what to do with your website to improve the situation.

Focus on the right data

There are plenty of different analytics reports to consider, but in terms of human activity on your website the important ones are the length of time people spend on each individual page and the route they took to get there in the first place.

Online engagement is very low indeed – most reports from Google Analytics shown in the “engagement” report suggest people have spent 10 seconds or less on each page.

That means they are really only engaging with relatively few pages on each website. Attention spans online are low – we expect “instant answers” and we move on quickly if we do not find them. The bulk of analytics engagement data suggests that most pages we visit we do not pay much attention to.

When you look at the time people spend on each page you can see which pages people are engaging with. But you then need to check those individual pages – people may be taking a long time on them because they are the subject they want, but the page is poorly written and so takes longer to read. Once again, seemingly positive analytics data could be negative, unless you take it in context. The combination of time spent on each page, together with an analysis of those pages and what factors contribute to the increased length of time spent is a good means of understanding what your visitors are doing on your website – and why.

Another analytics report to concentrate on to understand behaviour is the originating source of the visit. Has the visitor typed in a keyword to a search engine, have they clicked on an advertisement or have they come from social media? Each of these kinds of visits suggests different motives and so you then need to look at the pages which these visits led to so that you can see if the page landed on actually connects with those motives. For instance, people coming from a social site are doing so on the basis of a recommendation from a friend. That means they are already more positively framed for the visit, so you have the opportunity for this kind of visitor to be more amenable to calls to action. But a visitor coming to the same page from a keyword search is more likely to be looking for information – you will be able to see this based on their pathway after the landing page. You will be able to detect subtle differences in behaviour for different kinds of visitors coming from alternative sources. You can use those difference to create better-focused landing pages for those individuals.

Are you analysing data for data’s sake?

The vast array of reports in analytics software is fascinating and potentially time-consuming. Much of it can help you prepare better web pages, create better pathways and lead to more clicks where you want them. However, all that work with 100 reports or more can be more effort than necessary. Hours of analysis might only lead to one or two additional sales, for instance.

However, your visitors are generally after one thing when they land on your website. It might be information, it might be to make a purchase, or it might be to carry out a specific action.

Generally, though, you can find out what those things are by looking at a handful of analytics reports, analysing the pages that are involved and turning what you discover into a narrative.

Once you have this clear picture in mind, you can then work out what you need to do in order to improve the situation. This could be, for instance, by rewriting the copy. Or it could be by creating different pathways for different motivations. Or it could be using alternative landing pages more frequently. Ultimately, from a psychological perspective, your website needs to engage people emotionally if visitors are to take the actions you want them to take and buy from you. Analytics does not tell you the emotional triggers of visitors. But by carefully selecting the data you analyse you can get closer to understanding what makes your visitors tick. Just concentrate on how long they are spending on each page and how they got to that page. That will tell you the beginning of why they are doing what they are doing.

See also: Web Analytics Strategy