8 User Testing methods that work

why innovation! uses cookies to provide you with an optimal browsing experience on our website. Please check our Privacy Policy & Terms of Use to learn more about how we use cookies. By clicking on “I agree”, you consent to the use of cookies.
Sign up   |   Sign in

8 User Testing methods that work

Ellen Soedirman 7 min 166 5
8 User Testing methods that work banner image
8 User Testing methods that work

In our previous article on the 4 User Testing approaches you need to know, we covered the fundamental difference between User Testing and Usability Testing, what User Testing actually is and why it is extremely important for companies to perform thorough user testing. Then we explored in detail the four main approaches you will use during your testing. 

In this article, we dive into the specific techniques and methods to put your testing goals and assumptions into action! There are different testing methods out there, each with their own pros and cons for your to choose the most suitable one for your team and your product.

1. Observation

Perhaps one of the oldest user testing techniques, field observations are used to understand how personas interact with the set activities. Like all approaches, it has its merits and setbacks. 

Here are some of them:


  • No specific interaction, but an Observation Guide might be useful in some cases
  • Recording material (with consent from your users)


  • A lot of genuine insights (including body language, facial expressions)


  • Both the observation procedure and post-test analysis can be time consuming

2. Card Sorting

A user experience design technique where a group of users are asked to structure information based on their understanding & preference. Other than being quite fun for the users, card sorting is useful to understand the market’s common mental models. 


  • Physical / digital cards (face-to-face/remote)
  • Interviewees recruitment


  • Easy & cheap way to structure information
  • Direct interaction


  • Remote session will require participants to know how to use virtual whiteboard

3. Surveys

A method of gathering specific information and data from a group of users, usually in a form of questions. Surveys may be held over the phone, in person, or via the internet (e-mails / websites).


  • Surveys
  • Interviewees recruitment (for qualitative surveys as quantitative is usually blasted out to general public)


  • Large number of data
  • Paperless & automated results (online)


  • Quantitative surveys can lack in accuracy
  • Potentially time consuming

4. Interviews

Perhaps one of my favourite testing methods, interviews (when done properly) offers one of the most in-depth insights to your topic at hand. 

  • Interview Guide/Questions
  • Testing Scenarios (if doing prototype testing)
  • Interviewees recruitment
  • At least one moderator & one note-taker (highly recommended)
  • Reliable & detailed insights
  • Large amounts of qualitative data
  • Potentially time consuming & expensive

5. Sessions Recording

In sessions recording, we perform the one thing many people, to this day, insist is a conspiracy theory more suited to a Hollywood film about online privacy; observe our user’s online movements, such as mouse clicks, movement, scrolling, etc. to better understand their behaviours, needs, and frustrations. This is done through the power of internet cookies, Google Analytics and a variety of easily available tools in order to measure online metrics and consumer behaviour. 

Having those tools, however, are just one part of this. Being able to analyse properly how your user interacts with your site or product is key. A sessions recording can tell you if people are interested in you what you have to offer. Be warned however that much of the data sometimes requires calculated assumptions and the entire analysis can take a long time. 

Here's what you need:


  • Your product/prototype (website, landing pages etc.)
  • Recording tool/software
  • Fast & relatively inexpensive results
  • Eliminate guesses by viewing user’s actual behaviour
  • Result analysis is often time consuming – expect hours of footage, often repetitive in context

6. 5-second Rule

Just as the name suggests, the actual testing lasts only five seconds. A user is presented with an image of your prototype (home page, checkout page, etc.) for five seconds and are asked a series of follow up questions. The aim of this test is to gather the user’s first impressions and reactions to your product. 


  • Prototype/screenshot of your product
  • Target question(s) – e.g.; who do you think the target audience is? What do you think the product or service is about?
  • Cheap & large amount of answers
  • Limited scope of insights
  • The 5-sec test is usually used holistically with other testing techniques as the test itself does not cover a lot

7. Landing Page experiment

Landing pages are somewhat like a mini website that has only one focused call-to-action (CTA), unlike a website that may have multiple CTAs to encourages exploration. And because of the focused CTA, a landing page experiment is a good technique to present a product that may not exist yet to test market traction. 

A similar concept to the popular Kickstarter platform, the main objective of the landing page is to pitch your product idea to users and receive their feedback on their interest towards the product, usually in the form of pre-orders. Moreover, landing pages can also be used to collect funds, especially useful if your project does not have a high budget.

8. Wizard of Oz

Perhaps one of the most interesting testing techniques there is, Wizard of Oz is an experiment simulating a digital product/service, but manually carried out by a human being in the backend. 

What is the Wizard of Oz technique?

Imagine an ATM machine with an actual accountant sitting inside, manually extracting and delivering requested information and money to you. Or a drink vending machine with someone inside mixing and dispensing the drinks you choose, all while you were under the impression that a machine was carrying out the tasks. 

This is the concept behind Wizard of Oz. 

What is the point of this technique? Wizard of Oz allows you to fully test out an idea without investing in technology. Coming back to the vending machine example, imagine a world without vending machines and you are the first person to come up with the idea of a fully automated self-serve drinks service. 

The riskiest assumption is that people are more open to the idea of receiving their drinks from an automated machine, and not a person. In order to prove or disprove this assumption, a prototype of said vending machine is required for testing. Creating this automated machine will cost a lot of money and it carries the risk of being a failed investment if your assumption is wrong. 

This is where the Wizard of Oz comes in. Rather than spend time to design said machine and building it, all you need to do is build the outer shell of the machine, and secretly place your human wizard inside. This way, the user is still presented with the accurate product idea and experience. Even if your assumption is disproved and you need to stop or pivot, your investment was minimal.

One interesting case study is Amazon's Alexa. During testing, they utilised real people to simulate Alexa in response to what their users asked. This was done completely unknowingly to the user and the data was useful in identifying how people would talk to Alexa and what kind of responses they'd expect. 

With the final, Wizard of Oz technique explained, it’s now back to Kansas, or more accurately, reality. These are some of the most popular user testing approaches and techniques you can try for your next project. 

In closing, whatever technique you choose to implement, remember these five important things:

  • As much as possible, incentivise your users. They are dedicating time and energy to help you develop your product, and should receive a token of appreciation at the very least. Plus, recruiting users is a task in itself, and incentives will definitely make it faster and easier. This can be anything, from being a paid testing, to offering food or some vouchers. If you’re a big, well-known company, things like your brand memorabilia might be nice too. 

  • Have a clear goal for each testing or experiment. It is easy to get carried away in experiments as they might be very interesting, but without a clear goal, your experiment might not produce any useful results and waste a lot of time and energy instead.

  • Get into the mindset of gathering as much feedback as you can, as early as possible. Early feedback often saves you from any potential surprise obstacles.

  • Don’t be married to your idea. It is normal to get attached to the product you worked hard on, but keep in mind that testing is to make sure your product grows in the best way possible. Sometimes that means making sacrifices.  (Editor’s note: Personally, for me, I have learned to appreciate surprising/bad testing results, because for every disproved hypothesis a bullet is dodged!)

  • Last but not least, have fun! Testing is basically a science experiment, your own version of product mythbusters if you may. It’s a real life Forrest Gump gamble with the Box of Chocolates; you never know what you’ll get, which is part of the charm in the fascinating world of product design.

If you'd like to watch our recording of the webinar that inspired this series, click here!

Don’t be married to your idea. It is normal to get attached to the product you worked hard on, but keep in mind that testing is to make sure your product grows in the best way possible. Sometimes that means making sacrifices.
Don’t be married to your idea. It is normal to get attached to the product you worked hard on, but keep in mind that testing is to make sure your product grows in the best way possible. Sometimes that means making sacrifices.

More Blogs