Now that you have learned to research users, set goals, plan information architecture, direct the users’ attention, make good wireframes, and understand the mind of a user, it’s time to launch! And launching means we have something to measure, so we need to know:
What is Data?
(If you’re just starting the UX Crash Course: Start Here.)
Data is Objective
In one of the early lessons we learned about user research.
Data is different.
Data measures user behaviour. What they do. How many times they did it. How long it took. And so on.
It is collected by a computer, so it can’t influence the user. It has well-defined measurements, so there is a very low margin of error. It can measure millions of people, with no effort from you. And it can tell you things about your users like which browser they use or which country they are in.
And data never lies. This is science, bitch!
But it also doesn’t tell you anything about context, so be careful. Unfortunately us designers have to interpret the data, and that is where mistakes can happen.
Data is Made of People
You will be tempted to treat data as “just numbers” that mean whatever you want them to mean. Remember that those numbers represent the actions of real people with complicated lives.
Do not reduce millions of people into a single number and expect it to be reliable in every situation.
You may also be tempted to look for numbers that “prove” you were right. DON’T. And say no to anyone who asks you for that.
More data is better data
If you measure the clicks of 5 people, they might all be drunk, and you have no way to know. If you measure the clicks of 5 Million people, it’s pretty unlikely that they are all drunk, unless you only test people in Cancun during spring break.
The bigger the decision you are trying to make with data, the more data you need before you decide. But once the data speaks, the data has spoken!
A Few Ways to Collect Objective User Data
There are just as many ways to get objective data as there are ways to get subjective user research:
Analytics — Google and many other companies offer cheap or free ways to track what your users do, anonymously. Basically, every time they load a page or click something, you will know. And you can design custom measurements, so the sky is the limit!
A/B Tests — Design two versions of the same thing and launch both! You will know which one works better, because you will test it with real people, in real time. The software also lets you know when to stop, because at a certain point, more people isn’t going to change much.
Eye-Tracking — This is done in a lab, but the user can’t control it, so I consider it objective. Special software and equipment are used to measure where the users look as they use your design, so you know where you have guided them well, or not.
ClickTale — This is one example of using heatmaps for clicks and scrolling and flows, but there are others. ClickTale’s software allows you to anonymously record the interface as real users use it. Their input is hidden, but you see where they click, where the mouse goes, how far they scroll, and which pages they see as they move through your design. Super useful.
Search Logs — Many people don’t realize that a search field on your site can save every word that is typed into it. If people are searching for it, it means they can’t find it, so those logs are super valuable for improving your information architecture and layouts!
That was just the beginning. Tomorrow we will look at some basic numbers that are relevant to any website: Summary Stats.