Some aspects of a website’s design hinge on assumptions. A designer might choose colours and typefaces he or she feels will best resonate with the audience. Similarly, the information architect might believe his or her organization of content most logical. While there are ways to validate said assumptions, real-life data ultimately indicates how a website performs. Data also facilitates better decision-making as a site evolves.
In particular, we focus on gathering three sets of data after launch: qualitative, quantitative and observational. Without all three, the perspective is partial and without context. As you will learn, the more ways you can get to know your visitors, the more informed your marketing becomes.
Qualitative data: Generating user feedback and helping visitors in real-time
Data can take weeks to act on; you need enough to uncover trends and patterns. That said, there are ways to capture feedback in real-time that can quickly give you actionable ideas on how to improve your website.
Prompting users on your website to answer a question can help you define:
- User satisfaction ratings;
- User demographics / psychographics;
- User first impressions;
- User hurdles.
Any question that you ask should be concise, carefully worded and worthwhile. Use these four examples from UX for the Masses as a starting point:
- What do you find most frustrating about this website?
- Overall, how easy to use do you find this website?
- If you could change one thing about this website, what would it be?
- How can we improve? Send us your suggestions.
The example above comes from Hotjar—a software we rave about later in this article.
Rather than asking a question, live chat lets visitors start the dialogue. Over time, it can reveal common behaviour that you can address on the website. For example, frequent questions should get worked into the content. Similarly, you should assess navigation whenever users complain they are lost. In such ways, live chat lets you retain visitors who might have otherwise left frustrated.
The screenshot above is of Zopim, a popular software for websites experimenting with live chat. The basic package costs nothing per month, with paid subscriptions from $11.
Quantitative data: Bench-marking and measuring website performance
If you want statistics on how users interact, then turn towards quantitative data:
- How many visitors enter/exit the website;
- How many convert;
- How long do visitors spend;
- Where do visitors come from;
- How fast did the page load;
- And so much more.
Quantitative data lets you crunch numbers and plot trends. For instance, if you care about conversions, you can track week-over-week the percentage of visitors who become leads. For meeting quotas, this is a reliable metric. However, numbers never say why a customer converts (more on this in a second).
Quantitative data also allows us to create cohort reports—performance comparisons. Whether this means month-over-month or year-over-year, we can measure progress and compare results. Cohorts are particularly useful when applying changes to a website and measuring the effect.
By far the most popular quantitative analytics suite is Google. We set it up for every new project so that we can start collecting visitor data from launch. Unfortunately, without substantial traffic, it’s tough to gauge the significance of some numbers. Software like MOZ falls into this category; however, it is not something we setup by default. MOZ aligns more with search engine marketing than usability. Nevertheless, if you care about your Google ranking, then MOZ lets you easily monitor and improve it. To learn more, click here.
Observational research: Spying on your visitors as they navigate and interact with your website
Insights gleaned through observation reveal how visitors behave: they tell a fuller story. When something doesn’t add up, use observational data for context.
Heat, click and scroll maps
These images visualize “hotspots” on your website—where users linger, hover, click and scroll. Such data can explain why certain elements aren’t clicked or why some visitors miss information.
Like the maps, form analytics shows drop-off points on a given form. For example, if a form contains five fields, you can see at which number visitors abandon most. This refers to form friction—required pieces of information that reduce the success rate. You might discover that asking for a phone number heightens abandonment.
Ideally, you want visitors to progress along a sequence of pages. Using a conversion funnel, you can see the percentage of visitors who successfully follow this path. You can also see at which point people drop-off or divert.
For observational data, we rely on Hotjar, an analytics suite we use and install on all client websites. Under the basic subscription, the software is free—but there are paid services available.