How to use website data to prioritise improvements

How to use website data to prioritise improvements

Posted by Aoife Porter
Picture of Aoife Porter
on 26 March 2020
Websites

Use data in conjunction with experienced thinking - it will shed light on issues that aren't visible to the naked eye, and provide the support you need to take action. In the wise words of Peter Sondergaard:

 

‘Information is the oil of the 21st Century, analytics is the combustion engine.’

 

But, you shouldn’t depend on data to do your job for you. When it comes to prioritising website improvements, value is in the insight you can draw from the data, not the data itself. You need expertise to turn the information you gather into action. (Def: ‘Expertise’ - a fancy word for ‘opinions with a PhD’.)

Here are three important data points to look at:

 

1. Exit points

Exit points allow you to think about how you can improve the user journey. The aim is that the potential customer stays on your website and gets knowledge and value from it. At Articulate, we frequently delve into this data when making improvements. 

For example, we found that our contact page was getting a lot of traffic but was also where users were dropping off. Recently, with data we collected and some innovative brainstorming, we’ve created more value and increased conversions by optimising the page design.

What’s next? Why did the marketer get off the trampoline? He was worried about his bounce rate. 

 

2. Bounce rate

Let’s talk about bounce rate. Is it really the big bad villain that it’s made out to be? Or, does it just look bad because your visitors’ interactions with your website aren’t being measured properly? Often, the latter is the case and this can result in an unnecessary panic.

For example, let’s say the site visitor enters your homepage, goes to the bottom of the page and clicks on a video. They view the entire thing. They’re interested in your website and products/services. But, you haven’t set up tracking for any of these interactions. So when they leave the site without visiting any additional pages, that gets counted as a bounce. 

Bounce rate can be a valuable measurement when optimising your website, as long as you’re tracking it properly. 

New call-to-action

3. Impact pages 

When looking at data you’ve collected, find the pages that have the highest traffic and longest session time. Those data points show that these pages are providing value and keeping the user engaged. Focus your energy here! Add more CTAs, A/B test the copy or add subscription forms. These are pages that get a lot of views so they create a great opportunity to increase lead capture. 

High impact pages are already performing well. By providing insight into what is already working, you have the data to back up more website testing and optimisation. 

If a landing page is converting well once a visitor gets to it, there is no need to A/B test the whole look and layout of these pages. However, if you aren’t getting much traffic to the landing page in the first place, you should look at testing CTA design, placement and buttons. See which of these work best to increase the number of visitors, which should, in turn, help conversions. 

 

Heatmaps 

Heatmaps allow you to track what your page visitors read and do when they’re on your impact pages. 

They provide insightful data by showing you how much of your content is being read, and the points of the page that get the most attention and clicks. For example, If you look at a heatmap of a page and notice that the CTAs sit in areas that are highly read, but don't convert, you can start A/B testing them, or rethink their position.

For example, HotJar and LuckyOrange have tools which allow you to dig into this user interaction data.

Heatmap from LuckyOrange showing Articulate marketing website homepage

 

Using data to back up your intuition

When trying to make sense of your website data  the following steps are essential: 

 

  1. Structure a scenario you want to prove
  2. Organise the available and relevant data 
  3. Ensure you have selected comparable data sets
  4. Analyse and see if the results fit with your intuition

 

Knowledge behind the data is essential! For example, a drop in traffic initially looks bad, but there are many underlying factors that drive traffic and just the traffic data alone can't tell you which one caused the drop. Making sure you let the data support or disprove your thinking is important.

 

Failure is not fatal

Experiments are experiments for a reason. Even when using mounds of website data, not all of your testing will be successful, but you will always learn something and be able to use that data for further analysis. 

‘Failure is simply the opportunity to begin again, this time more intelligently.’ - Henry Ford

New call-to-action

See also: website optimisation

Related service: Websites