A common sense crash-course for improving your user experience.
Improving user experience is as easy as cleaning your room. (PHOTO CREDIT: Jonny Scholes, Flickr.com; IMAGE CREDIT: Bret Mavrich)
My kids routinely do what all children do: they don’t clean their room, they leave the bathmat wet on the floor, they leave the cereal box out. The typical excuse is “oh, I forgot…” to do whatever. Lately, that thought process has struck me. Their expectation seems to be that following the rules is about memorizing, and complying with, a long checklist of appropriate behaviors.
I remind them instead that I don’t remember anything, either. I actually have a sort of terrible memory for processes and steps and procedures. So what I do is turn around and look over my shoulder. Just like you look before crossing the street, you should make sure the bathroom looks neat before you leave it after taking a shower.
We as humans actually become pretty good at doing this; when we take a moment to step out of our normal, rushed lives we notice what is wrong, broken, or out of place. Yes, I’m talking about this in our professional lives also.
I am a user experience designer. I spend a lot of time reviewing products, or product ideas, and pointing out what is going to be confusing, difficult to use, unexpected, inefficient, or generally could be better.
But better yet is when I can lead the people who hired me to recognize these problems and opportunities themselves. I work with a few organizations who believe that considering the user is so important they wouldn’t consider a new project or startup without a user experience (UX) designer as the first, or maybe third hire. But I know many of you aren’t there yet, and instead what you are looking for are easy-to-follow methods to integrate into your everyday work.
Try these out periodically through the process to make sure you are planning and building not just something quick and reliable, but are spending your time making the right thing for your users, so it’s actually successful.
Know Your Customers
Okay, I lied a bit. The first thing you need to do is know what you are doing, and why. You’ll get laughed out of any bank or VC meeting if you have terrible plans, or huge gaps in them. Just add to those some key user-centric insights. What is your audience? What are their goals? What are they using now to solve this need?
Stop and Look Over Your Shoulder
Add this to your work plan. Maybe a bit like unit test even, so after each feature is completed, or every few days as you would do an integration test, walk through the whole product like an end user would. Try to make no assumptions based on what you know about the system.
- Does the information architecture (the visible structure of the site or app, and arrangement of items on the page) make sense?
- What about errors? Can the user get out of them? Can you eliminate errors from the process?
- Throw away the happy path. What if, instead of the home page, users pop into a random page from a Google search or shared link? Does it even work? Does it make sense?
- Are you lying? Does the structure or language obscure the true organization, data, or structure of the system or process?
An App is Not a Mobile Strategy
Remember when I asked what your audience uses now to solve the problem your product will fix? Well, that might not be a competing website, but a book, catalog, pair of scissors, or Post-It note.
Think about all the systems we use every day, maybe by stopping and just observing how much you yourself email, IM, SMS and just walk over and talk to people. A website is not a digital product, and an app is not a mobile strategy. You have to consider the whole ecosystem, and design the password reset emails, make it easy for customers to send links to their friends and family, and make sure monthly billing statements use the same terminology as the checkout process.
UX is not UI
User experience (UX) is not the same thing as user interface (UI). Recognize that good user experiences are about being intelligent, seamless, friendly, informative, trustworthy, and easy to use. There’s nothing in there about color and font size. Good UI is important, but it cannot fix a fundamentally bad product.
Good UX cannot be bolted on later or saved for release two, but extends through your whole product. If a problem emerges because you stumble across it, or people complain, a better error message or new icons will not fix it. You need to be mentally prepared to redesign the whole process, to eliminate or add sections, or to redesign the database from the ground up.
Show it Off
I am starting to see lots of small, fast, nimble startups show their ideas and alpha products to actual people. That’s great and if you aren’t going to make a UX hire your next task, you should at least do this. But, there’s very little guidance on how to do it. And therefore I see too many testing with their developer friends, confusing usability tests with market research, and otherwise getting results that aren’t to be trusted.
Basic usability testing doesn’t need a masters degree, a lab and a staff, but try to follow a few basic guidelines:
- Context is more important than working code. You can get good results by sketching ideas on a bit of paper and letting users interact with the “system” (have another developer be “the computer”). But try to do it in their office, or at a desk like they would use. If mobile, tape it to a bit of wood, or an old phone and have them carry it while you follow them around.
- Find your audience. Find people with jobs similar to your real audience, and especially avoid anyone involved in the project. I went to my neighborhood auto repair shop recently to test out an auto parts finder on actual mechanics. Try to remove yourself a step and use friends of friends, if you have to recruit like that.
- One on one. Do not do focus groups. Just trust me, it’s the wrong thing to do. Get one user, and one moderator at a time and run then through the process. Make sure they talk about what they expect, and what they think they are looking at. Reassure them you are testing the product, not them; say “there are no wrong answers” in the intro. And lie if you need to, that you are just testing, but didn’t build it. They will be more honest if they don’t worry about hurting your feelings.
- Ask them, don’t tell them. Don’t lead users. Be careful what you say when writing a test plan, and let them fail for a bit before you assume they won’t get it. Don’t tell people to click on a link, but give them goals, then listen and watch how they try to get there. Only correct them after recording their actions, and if they have to go down a particular path to complete the test.
- Your memory is terrible. Record, or write everything down. Really, you will remember only what reinforces your pre-conceptions of how the tested interface will work. Ideally, bring someone else along who can just take notes, or film the test on their phone.
- Make good decisions. You will find a lot of complaints and failures. Look closely at the notes. No matter how badly one person does, it’s just one person and you can ignore it. Even without real statistical analysis, you can look at the raw numbers and find that only a relatively few problems are shared by most participants.
- Don’t trust users. Go back through the steps above, and think about ecosystems, structures and bigger solutions to fix your problems before changing anything. Separate users’ preference from their performance: generally you can ignore what people think the system should be like (suggestions to move things and change labels), and find better answers to the deeper issue instead.
Ask For Help
Lastly, there are more UX consultants than you might think. If you can’t hire one full time, try to get one in to help for a bit at those key junctures, and as early as possible.
Not sure, or really, really hard up for cash? I am even giving free trials right now. If you are a startup in the Kansas City area, in the Village or not, I’ll be holding office hours at the KCSV most Tuesdays. Come talk about how UX can improve your product, how to hire the right people, or bring me a specific problem and we’ll work through it for a little while. Find out more and sign up for your time