How to Test Prototypes for Usability Issues

You’ve created an amazing website design and you’ve got your content in place, but is your site user-friendly? Often the only way to find out and improve your design is by paying for users to test your website. Those users provide feedback and then you have real-time data to improve your design. Then, you test again, get feedback. And again. And so on. Often when hiring outside companies to conduct these tests it can get expensive. But here’s a small guide on how to test prototypes for usability issues–within your budget.

Finish your prototype

The first line of the blog assumes that you’ve finished your prototype–fully finished it. The first step is getting as much of the design and content in place as possible. Before you start testing your website and usability issues, you need to ensure that everything is in place just as if you were going to hit the “launch” button today. That means delete any Lorem Ipsum placeholder text, add images instead of grey boxes, put your meta descriptions and tags in place behind the scenes, make sure you have your alt text on your images, set up your SEO metrics, and ensure that every page on your site works and has text on it (or if you’re doing a continuous scroll website, that all of the pieces are in place as users scroll down) since all of these elements of text, messaging, branding, and design work together to create the user experience.

Think about accessibility

Often designers and web developers come up with these fantastic designs, but they don’t always think about every end-user or as much as is reasonably possible. What we mean by that is that some users will be hearing impaired, have mobility issues, have sight issues, be working on a moving train, be using a phone or tablet, and so forth, which means that you’ll have to consider at least how you can get the best website design for these situations. We have a piece here on website accessibility if you’d like to find out more.

Tests to conduct before users get on board

Before you ask people to test your website for user experience, you want to make sure all of the nuts and bolts work. You’ll want to conduct the following tests before you get outside people involved as part of a panel.

  • Cross-browser testing – ensure your site works on different devices and in different browsers
  • Functionality testing – testing all internal and external links, test all links that jump on the same page, check for orphan pages, test all links that link to admin email accounts, check for broken links
  • Form checking – test all your landing pages and thank you pages and any pages with forms to ensure they work correctly and on different sized screens – there’s nothing worse than a form that doesn’t work for mobile! 
  • Cookie testing – check that your cookies are encrypted before writing to the user’s machine or ensure that they test the session cookies if you’re gathering that sort of data or user stats and check what happens if you delete cookies
  • Validate HTML/CSS – since everyone wants their website to be optimized for search, you’ll want to validate HTML/CSS – validate the site for HTML syntax errors and ensure the site is crawlable to different search engines
  • Database testing – check that all your data keeps its integrity when you delete, edit, and modify your site and ensure database queries execute correctly

So, once the above is in place, it’s time to start thinking about what you want your users to do when they test and what data you will gather.

What will users test? What should you ask and what are they looking for?

Users test to see how easy your site is to navigate and to identify any issues that need correcting. It’s a good idea to test your website at least four times and create new iterations every time but we’ll get into that later. Here’s what you should ask your users when completing their usability test.

  • How easy is it to learn on-site?
  • How easy is it to navigate?
  • Are menus easy to find?
  • Does the site work intuitively or do you struggle to find what you’re looking for?
  • Do you feel satisfied when navigating the site? Why or why not?
  • Can you find the pages you’re looking for easily?
  • Do the pages load quickly enough?
  • What is your overall satisfaction with your browsing experience?
  • What is your impression on the look and feel of the website?

User experience is paramount to good design. Here’s a piece we wrote about Nielson Norman’s advice on how user experience affects your bottom line.

The questions above are a roundabout way of getting users to test for the following.

  • Navigation – how a user surfs your pages, how they use buttons, how the user clicks on menus, links, or pages, how they move through your website
  • Ease of use – the website should be easy to use, instructions should be clear, the overall design should be consistent, and except on landing pages, all pages must have a menu
  • Content – content should be logical and easy to understand, check for spelling and grammatical issues (it’s a good idea to hire a writer and editor), make sure that the colors and fonts work together, anchor links should work, images should be properly placed and appropriately sized
  • Performance – web load time, web stress testing, connection speed, and page printing should all be tested

You may have other metrics to test for but these are the basics.

What kind of users will test your site? Consider your target audience and how many testers you need.

Before testing with a user, your website testing company would screen the users based on demographics such as age, gender, salary, position, geography, etc, depending on your test parameters. You will generally want to find twenty users that fit your target audience for the website to get the best data. That means you’ll need a good idea of who your business targets as its primary market. If you’re more of a general audience, then your company can find a more general pool of candidates to test.

Next, your users will execute tests with specific parameters. The questions above were more generic ideas, but in usability testing, users will be given prompts like ‘imagine you saw a PSA on social media and visited this website prototype, where would you go to find information about where to get involved in this campaign.’

You can discuss what you want to test with your usability testing company so that you can get the right data points for improvement.

Why twenty people? Find out why you need only twenty people to complete four rounds of testing.

So, in the point above, we said that you’ll need twenty people to test your website. Here’s why. You may think that it’s best to go in and have as many people as possible to test your website at once and give you all the data. Well, we’ll tell you that you won’t get any more data or not much more usable data with twenty people at once than you will with five. Let’s explain.

Round one

In the first test, you’ll want to test with five people. With the first person, you’ll discover all sorts of data that is new and exciting. You’ll find that the user may surprise you in how they go through the site versus how you intended people to navigate it. 

Then, you’ll move onto the second person. You won’t be surprised to find that person one and person two might have some overlap. People are all individuals, sure, but they will probably navigate the site in similar ways. You might find some new data points, which is great.

I’m sure you see where this point is going. By person three, there’ll be even more overlap between the first two users and you’ll see the same sort of testing for the third time around. Again, you might discover the odd new datapoint here and there.

The fourth person will have more overlap and some new information and the fifth person will have even more overlap and a little bit of new information. 

If you’d wasted your spend on testing all twenty of those people at once, you wouldn’t learn much more with any more subsequent information. 

Once you’ve collected the data points from round one, make improvements. Create a new version of your prototype. Take every suggestion into consideration. You may or may not agree with everything the user is saying but if they are saying something is wrong. It may not be what the user is pointing out, but it will generally mean something does need to be fixed and addressed. Once your next iteration is ready, test again.

Round two

Round two is like round one with your next set of five users. Listen to your users. Make changes and make another version of your website based on their feedback.

Round three

Now, take your next set of five users and learn from them. See what else can be improved and improve it. Then make a new prototype; hopefully, this prototype will be the closest version to your last one as possible.

Round four

Take your last set of five users and do a final usability test. You may do more rounds. Keep going for as long as you’ve budgeted. Improve and retest.

However, we suggest that you make the most of a smaller pool of users. You will see more improvement in your overall design and usability with several smaller iterations than all at once. In this method, you’ve had five chances of improvement. If you use your budget all at once, you miss out on that quality and quality.

On a side note, the only time you’ll need a larger pool of users is when you are conducting quantitative usability studies where the goal is to derive metrics by collecting results with statistical significance. For those tests, you need a larger sample size.

The takeaways

As Jakob Nielsen says, “You don’t get better insights with more users, just a better number.” When testing your prototype for usability issues, the goal is to drive design quality. You can do that with more iterations of your design, tested with smaller numbers. Fewer users testing means you can do more rounds of testing within your budget, which means your return on investment will be greater overall. Not just that, but the law of diminishing returns clocks in at 4-5 users to uncover a vast majority os usability issues. The good news for you is that Key Medium will conduct remote usability tests for as low as $1,000. Find out more.


User Research & Analysis Consultant in Philadelphia

Samaya is as a UX strategist with experience working at both startups and large corporations. She is currently completing her graduate degree in Integrated Product Design at the University of Pennsylvania; she received her B.S. in Mechanical Engineering from Carnegie Mellon University. She’s previously consulted for the likes of Facebook, J&J, Chase, Ford, Lincoln, Complex, NCR, AFS, and UPenn. When she’s not working, she can be found reading a good book, backpacking, going on road trips, painting, or having brunch with friends and family.