A Hands-On Guide To Data-Driven Design
In world where analytics rules, design is becoming evermore data-driven. Is it something worth paying attention to and are the bonuses of it worth the effort?
Getting clear with Data
Before getting straight into Data-Driven Design, we have to make sure we understand what data is in this context. Unfortunately, web-related content does not follow any strict design rules. Designers are operating on their own terms and knowledge and that’s why misunderstandings are often inevitable. But let’s talk less about the frustrating things, and more about how to solve them.
The wide range of analytics tools now available provide designers with somewhere to turn to in order to make more informed decisions. However, despite the precision of analytics, we only see the outcome. We neglect to see the “why”. You gain numbers, statistics, charts and diagrams but what are they really saying? Where do you head after knowing this or that?
So what are the types of data we will be using as a tool for achieving our goals? There are two of them:
- Quantitative data, or the data of numbers that shows you who is using your site, when they are visiting it, what they are doing on your page, etc., generally consists of charts, numbers and diagrams.
- Qualitative data, the less appreciated yet highly important part of your data, tells you how your users are achieving any particular goals and why they’ve chosen that very path.
There are many analytics tools, but most of them focus on the Quantitative aspect of data, as it is arguably the easiest the measure. Using these tools you’ll be aware of what your users are doing on your page, how they actually got there or who they are.
Yet, what will remain unclear is WHY a particular user (or a set of users) bounced straight off this or that particular page of your website. There is a limit to what purely numerical data can tell you. Believe it or not, there is still a huge part of data-driven design where the ‘Why’ has to be answered.
The “Why”: the hardest data to measure
The ‘Why’ aspect is of highest importance, it opens the mystery of user logic. For example, let’s imagine you have several groups of visitors on your web page. You have divided them by some actions they are taking, doing what seems like a reasonable step. You’ve got a group of people leaving your page after they’ve just landed on it, and another group of people leaving at a certain place or post. Why are they doing so?
Qualitative data is that type of information that may provide you with these necessary answers. thus, qualitative data provides you with more perspective.
Knowing this ‘Why’ may dramatically help you in various occasions. Imagine a blog you have (or any blog you have ever visited), for example, our corporate IT blog I always use as a visualization tool when I am hypothesizing on the mentioned topic. Keeping visitors as entertained and amused in a way you were planning them to be may proves quite a challenge for everybody.
Let’s take a look at some parameters you might wish to adjust in order to make some improvements:
So, where should you start? The Google Analytics account is my weapon of choice. My second step would be locating some data about the bounce as well as exit rates.
- The exit rate is the amount of times any user leaves your domain – divided by the total amount of the page’s views.
- The bounce rate is the total amount of visitors entering a domain on one page, but leaving before even viewing the other pages a domain may offer – divided by the general amount of the page’s views.
Suppose the information you have gathered points you to the fact that you have three pages rated much higher than your blog’s average bounce and exit rates. This is a swell example of quantitative data at work. The information is useful, yet meaningless without any extra data on why users are choosing to leave your blog.
Thus the next steps are important for you to take in order to achieve some qualitative data:
- You know that there is something fishy going on with those three pages as they were rated high in exits and bounces. You are checking those pages out and are noticing that two of them have some links to your website. That means they are doing great and are sending visitors from your blog to your website as you have originally planned them to. So nothing to worry about here.
- But still there is the third link with a great, cared, long post you have been working a week or two on. You obviously don’t see any reason why users are leaving as the post is magnificent (at least to you).
So what may be causing such an activity and what are the things you are capable of in such a situation?
There are two options you opt for here:
- Beg your user to stay (just as below)
- Or you can opt for the more serious route and manage the issue using qualitative data. You should begin observing. Are you experiencing something fishy when visiting this particular page? Perhaps there is something that irritates you. What is your colleague’s opinion? What do your friends have to say about that particular page? What may be forcing them to leave?
User Testing is what will give you the answer to the ‘why’ part of the equation. It allows you to be way more efficient as you have already narrowed the search to one particular page. Keeping your blog user-friendly is something you are not to neglect.
How do we apply this?
Now that we have the appropriate data, how do we improve the design with it?
- Let’s begin with our analytics. Ensure you are collecting the data relevant your needs. Data relevant to the context.
- The same needs to be done with our analytics tools. Ensure you are fully aware of what the data your tool has revealed means. Many people might find qualitative data rather fictional as it is difficult to measure with common numbers. It always helps to compliment it with quantitative findings.
- Use only empirical data (the data you have gathered via observing and experimenting). This has proven to work, and is relevant to the situation.
- Ensure you use questions specific for your needs. User real user data to answer them.
There is no ‘perfect solution’ in Data-Driven Design. Every effort should be individualised to individual pages. There are dozens of thousands of different users. The ones that are returning to your page might have different goals and needs from those who are visiting your site or blog for the first time. Whereas the first are certain of what they are desiring from you (and you are to know that and provide it), the second ones have probably landed on your blog straight from a Google search and thus are to be encouraged to stay there with some appropriate content. So knowing that data can really drive your design to the stars.