If you want to improve the conversion rate of your website it can be difficult to know where to start. If you have a big site then there are going to be lots of pages, or even sections, that are important to you and you might not have the time or budget to make changes to all of them. This post should give you some ideas of how to identify your key pages and paths which impact your conversions.
The first place to start is unsurprising analytics. Whether you use Google Analytics or another analytics package you really need to delve into your data to see what your users are doing on your site.
A simple starting point is to look at the bounce rate for your top pages. If you’re using Google Analytics then the Comparison chart can be useful here. This will tell you how each page compares to the site average bounce rate:
Here we can immediately see some pages that require a closer look. There may be a good reason why these pages have a high bounce rate but we need to identify them and then think why the bounce rate could be high.
Another useful metric to use here is Page Value, which gives some idea of the pages that currently help to generate income. If the page value is already high then we might want to look at getting more visits to this page, if it’s lower than expected though then we’ll want to look into why this might be.
Another option is to look at ‘leakage’. The term leakage was coined by Digital Agency FreshEgg and is used to work out the amount of revenue potentially lost by people dropping out of the conversion path at each step. Leakage is calculated by multiplying the unique page exits by the average page value. If used correctly, and carefully, this will help show where you might be missing out on potential sales.
To get a better idea of where to optimise though you really need to be looking at more than individual pages in isolation. You might want to try categorising sections of your site using Content Grouping to get a better idea of the areas of your site that are underperforming. As with all serious analysis you also need to make sure that you’re looking over longer time periods and taking seasonality into account.
To get beyond pages, or groups of pages you should start looking into user journeys through your website. You can use the Navigation option in Google Analytics to see previous and next pages that users are viewing. For a broader view you might also want to look at the Behaviour Flow and User Flow sections of GA.
One example of this is shown below. In this instance the website had a simple form on the webpage which when completed takes the user to a results page, shown here as the ‘form-handler’ page. The form gives the user options of locations and activities and the idea is that the user selects a location and/or activity and is then shown relevant results that they can then explore to learn more about each of the options. However, as the Behaviour Flow shows, what’s actually happening is that they are entering their search terms on the homepage, getting shown the results but then a large amount (just over 50%) of these users return to the homepage before moving onto the destination or activity pages.
This means that users are looping back on themselves rather than carrying on towards converting. This issue can be caused by a number of factors. It might be that users are getting shown poor search results, it could be that the search instructions are unclear or it might be that more details are required at the search stage to yield useful results.
Depending on what the issue is the solution might be fairly simple. It could be a case of simply adding faceted search or advanced search functionality on the results page.
Other Analytics Tools
There are lots of other tools offer functionality that Google Analytics does not. There are analytics packages that can show where people click on a page and how far they scroll. This will help you get a better idea of what areas of your page are working and which are not. Scroll rate often shows you what you already know; that not everyone will scroll down your page so you should position the most important information at the top. It can however be useful for working out exactly how many people are scrolling to a certain point. Not every elements on a page can be ‘above the fold’ for every user so it’s important to see how many of your users are seeing items that are further down the page.
Another method used to find out more about user behaviour is to track their mouse movements. Tools like ClickTale will show you where people are moving their mouse as well as where they are clicking. This information would normally be completely invisible to you so it gives you a new type of insight. Studies have shown that there is a correlation of over 80% between mouse and eye movements so by tracking how people move their mouse you may also be tracking where they are looking on your page and may offer a quicker, simpler alternative to eye tracking.
If you have identified an issue with the site, as shown in the examples above, then user testing will help you find out more detail about why this issue occurs and will begin to give you some ideas on how to resolve it with an updated design. User testing isn’t just useful for finding out more about why issues are occurring; it can be a really useful way to identify issues outside of your analytics data.
Observing users attempting to complete common tasks on your website is good way to identify where issues may be preventing them from converting. Ensure that the tasks you set don’t ‘lead’ the user into behaving in a certain way. Decide which tasks are the most important and let the users get on with them. These tests can be moderated or remote but whichever method you use you need to set aside plenty of time to analyse the results.
You might also want to experiment with quick basic online testing tools such as Five Second Test, though these tools are more useful for identifying issues when you have found problem pages on your site.
Getting customer feedback may seem a bit old fashioned when compared with web analytics and sophisticated user testing methods but it is a very useful way to uncover potential areas for improvement on your site. If you’re already collecting some form of feedback from your users then you can search through that to find common complaints. This could have been collected by a feedback widget or could even just be data collated from customer complaints. Sifting through this feedback to find common problems will help you focus on problem areas of your site.
Eye tracking is another form of user testing that may help you in your CRO process. I hesitate to include this here as it’s really used after you’ve identified problem pages. It can however be used to back up your ideas about what might be causing problems on these pages.
While eye tracking is not as useful as analytical analysis or task-based user testing it can give you insight that the two methods do not. Eye tracking is best used when you have at least some idea of areas of the site that you might want to improve. To get the best results you should approach eye tracking with a hypothesis in mind, e.g. “I don’t think users are noticing our sale link.” With this in mind you can run some eye tracking to try to prove, or disprove that theory. Be sure to follow up eye tracking with some questions for the users as the tracking alone will not give you a full picture of the user experience.
Knowing where to start optimising your site to increase conversions can be a daunting prospect. Rather than just focusing on what ‘feels right’ or the areas that you think are most important to your business you need to be looking to get some data to highlight true areas of importance. This should be quantitative data, such as analytics and also qualitative data like user testing.
You also need to ensure that you’re looking beyond the obvious ‘headline’ figures and that you remember to break down your audience using segmentation. It might be that a page looks like it’s performing well when in reality it’s performing very well on desktop but is underperforming on mobile. It could be that your conversion rates are great for one country, but are well below par for another.
The key here is not to rush into making changes or setting up a/b tests before first identifying the areas of your site that will have the maximum impact on the overall conversion rate.