User Testing for Discovery and Accuracy

two modes, one great human-centered goal

As you make a plan to create or improve a product user testing is critical. Early on you will need to make a case and make plans to test. This article will help you communicate the value and utility of testing to stakeholders early on so that they too can understand how testing is essential to creating a product that is both useful and desirable for your users!

There are two approaches to user testing: Discovery testing and Accuracy testing. I encourage you to consider both types of testing in your work.

Using a combination of discovery and accuracy testing not only clarify the needs your users have, but it will let you point towards the discovery of problems your users are facing , and then clearly show how your solutions helped to remove roadblocks that were creating a poor user experience.

Let’s take a closer look at Discovery testing and Accuracy testing and explore the utility of each.

Discovery is very much what it sounds like. You need to find the fundamental issues your users are facing within your website or application, and you need to hear about the big gaps or major pain points in your offering.

Discovery testing is about finding your bearings: you want to know which direction to head in and the general destination you’re aiming for. For example, you could do discovery testing with an existing product, knowing that your users seem to investigate something like a product page, but then they rarely checkout. Your testing might be about investigating why this happens, and hearing what they are thinking as they explore products and attempt to make a purchase.

Discovery testing can be more qualitative than quantitative. You don’t need to know the exact time it takes someone, down to the milli-second, to find a button. You just need to notice they are not finding a button easily, and ask questions about where they expected the button might be. Your are discovering what your users think, and why they think it.

Ask Questions about what’s bothering users. How do they react to a problem they face? Do they give up or find a way to work around the problem? (image credit:

During discovery testing try starting with a small set of questions that follow critical paths (maybe a purchase flow or a sign up flow) and approach testing sessions with an open mind. Get to know your users- ask open ended questions, learn what other sites and software they are using that is shaping their expectations and learn what’s difficult or interesting to them about your product.

Pay careful attention to moments when users express strong emotion or pause in confusion…ask them to “think aloud”, narrating their actions and choices within your site or application, and allow them time to answer questions at length.

While it is less helpful at this phase to record something like a satisfaction score for a website or application you already know you are going to design, it is helpful to ask users how challenging they find tasks (very easy, somewhat easy, challenging, very difficult), or if they find tools or navigation relevant to their needs (very relevant, somewhat relevant, not relevant). These more general statements about ease of user and relevance can help you focus your priorities and may also help you let go of some of your preconceptions about what users hate the most or need the most!

Now let’s talk about testing for Accuracy. This is a phase of testing that you undertake when you have already created a first draft of your product. You’ll be testing the accuracy of your approach: your “theory: about what a good solution will be. Testing can be done using a prototype, an early live code build. You will be testing to understand how right (or how wrong!) your assumptions have been.

While some of your test questions may be qualitative, quantitative testing approaches play a key role in this phase. You feel confident you’ve taken a solid approach in your designs- but what is the degree of accuracy in your design? What results have your changes had in the ease of completing a task and in over all customer satisfaction or conversion rates?

The accuracy phase of user testing can be an extended one because it examines your work with a more concise and detailed lens. In some cases there may be multiple solutions to a problem: you may find that you need to make a number of adjustments to find an approach that works. It may be necessary to adjust and retest your work more than once to be sure you have effectively solved the challenge users were facing.

image credit:

To learn more about user testing for discovery and accuracy and to see some nifty info-graphics about where to apply user testing to your process check out: Measure once, cut twice or thrice: User Testing modes for Lean and Waterfall