The User is NOT You (User-Centered Part 2)

This is Part 2 of some thoughts about being user-centered. Part 1 was a Cautionary Tale that I wrote earlier.

We don't always apply our own rules

We all say that we are user-centered. We all talk about personas. A lot of us even use the names of the persona as we describe our designs. But when we form opinions about designs and design solutions, we still rely on that nebulous sense in the back of our heads that says, "Oh, yeah, that looks great!"

We like to say that this is the collective wisdom of our "experience," where we have internalized the many good or best practices of UX, that allows us to make these very quick judgments of quality. Ok ... maybe ...

But if that's true ... where is the user? Meaning, is that quick judgment also taking into account the perspective of the particular user in this interaction? When we say a design "looks good" using an aesthetic judgment that assumes the user is on a general retail ecommerce shopping cart flow ... do we use the same or different aesthetic judgments when we assess a design for a user who is a highly technical user working on an intranet system used for resource planning?

User needs are more diverse than our aesthetic judgments

In my own experience, my observation is that designs that people say "look good" are not that diverse. Most of us tend to favor designs that are clean, uncluttered, simple, etc. And these principles, as a whole, are not bad ones.

But what if that's not what the user, in this particular case, needs?

I made up these numbers, but it's representative of a study I worked on a long long long time ago. We conducted multiple trials on 3 designs and measured time to completion as well as errors.

It's my observation that the huge majority of UX design practices as well as User Research methods are biased to prototype 2. You see that most users figured out the design of prototype 2 pretty easily. The learning curve is flat. There's very little difference in performance through multiple trials.

I'd like, however, to draw your attention to prototype 1. It's obviously very complicated. It took users by far the most time to complete in early trials and they committed by far the most errors. But look at the 10th trial.

New Users vs. Expert Users

It is my observation that that huge majority of our aesthetic judgments are highly biased toward designs that are likely to perform like prototype 2. They are intuitively obvious, and a first-time user can complete the task just about as easily as an expert user.

These kinds of designs are probably appropriate for public-facing experiences and probably for the majority of retail experiences. You don't want to have to train your user how to use your app or website. That would be crazy.

But what if our product is designed for expert users? What if the typical user is someone who uses this application for hours every single day? In that case, the data for Trial 1 is practically irrelevant. Trial 10 is what we should care about. And if you look at Trial 10, the prototype that initially took almost 3 times as long to figure out at the beginning took almost half as much time by the end.

Gaps in Methods, Experiences, and Best Practices

It is my observation that the majority of UX practices, Design Systems, User Research methods, etc. have gaps in assessing the best interaction design for expert users.

For example, I have yet to talk to a User Researcher who can recall conducting 10 or more trials on a prototype with the same user -- not 10 sessions with 10 different participants. I'm talking about asking that same participant to do that same boring task 10 or more times in a row. Does anyone do this regularly? If not, then how do we assess the learning curve of a design and accurately evaluate the usability of a design for expert users?

What's worse, is that just about all of our intuitive judgments about what a "good" design is are biased toward prototype 2. Most of us and our experiences would probably judge prototype 1 as terrible, complicated, cluttered, confusing, etc. That's why it would take a new user so much time and cause that user to make so many errors.

But, sometimes, the aspects of a design that make it complicated for a new user are what make it efficient for an expert user. It probably has no instructional text, because that's a complete waste of space for experts; it probably has abbreviated labels, which are easier for an expert user to scan quickly; it probably has multiple selectors, multiple states, etc. It probably allows an expert user to see a large amount of data in a single view and to conduct multiple transactions in a single batch.

I'm not saying that all intranet application should be complicated ... In a lot of cases we want intranet applications to be simple too. There are plenty of enterprise applications I use 1 time a year, so I'm a new user pretty much every time. But there are others that I use every day. And some of those things that we do to make it easy to use the first time around, make it annoying the 100th time around.

Conclusion

I used a fairly extreme example of a system that users would use every day -- where they are likely to build up a lot of expertise to make the case clearer.

My point is not that we need to focus on the "expert user" persona more. Neither is my point that I have identified this or that characteristic of "expert users." Rather, my point is that we have internal biases that affect our aesthetic judgments that sometimes prevent us from seeing the actual user needs.

I believe that there are many cases where we substitute our own likes and dislikes based on our own perspective that, in some cases, can result in non-optimal experience for the actual users of the system -- because we are not those users. And, in many cases, this tendency is more prevalent in more experienced designers -- because of their confidence in their own experience.

We need to actively refrain from relying on our snap judgments about what looks good. Coleridge coined the term "suspension of disbelief" in describing what good readers of fiction must do. As designers, we must suspend our own aesthetic judgments in order to arrive at the right solution for users. There are cases when our own experiences get in our way and prevent us from discovering the most effective design FOR THE ACTUAL USER -- WHO IS NOT US.