Data-driven design

Even if you are a member of Mensa International or a professor in rocket science, I believe that I can confidently say that you are probably very bad at reading someone else's mind. If you want, you can give it a try. Look at the person closest to you and focus. Focus on what they might be thinking about right now and then check with them if you were correct or not. And? You were miles off. And now try and do the same for someone that you don't see or don't even know personally. And? Even worse. 

So if we're all so bad at reading someone else's mind, why do we often assume we know what people want, need & think when we make decisions about what products to build for these same people, what features they need and how they use the products we just sold them?

In most situations, the answer is probably that “we know that we don’t know but we have to make a decision so we just guess and hope for the best”. And in most situations, that means that you’re building your companies’ future on chance. And the chance that you are wrong is a whole lot bigger than that you are right.

At SDL, we don’t pretend that we can read minds but we definitely try and reduce the ‘not knowing’ part and thus reduce the chance that we’re wrong about what people need or how they’re using our products. We call this data-driven design. This means that we are always seeking data about what people need, how they work, and how well new designs and features match their needs and whether they can be used by them.

The SDL UX group uses a wide array of methods and tools to gather this type of data to help ourselves and the teams we work with to reduce the amount of guessing we do.

As people are very bad at talking about what it is they want or need, we prefer to go out and talk and observe them as they perform their regular work activities in their own context or during usability tests (also called qualitative research). This gives us the opportunity to learn from what they do as well as what they say. Personally, I always compare it to how hard it is for someone to explain to me how they ride their bike. Because they’ve automated most of their actions to the point where they don’t even have to think about them anymore, it is difficult for them to explain what they do but it is very easy for them get their bike and show me how they ride it.

We are also working with the R&D teams to implement technology that will help us understand how people use our products, which we have called Telemetry. This is something that is very common on regular websites but we don’t have this (yet) in most of SDL’s products. The quantitative data we get through this method will tells us WHAT people do and if we then combine it with qualitative data, gained through interviews and observations, about WHY they do them, we have a fairly complete set of data to help us inform what we should and could do in the future to improve our products and services.

So, in summary, we are constantly generating data through qualitative and quantitative methods to help us make the right decisions. To be able to do that, we are always interested in getting into contact with you, our users, partners, developers, implementers and general members of the community. If you’re interested in helping out, post a comment below and we’ll contact you.