How lean usability testing can improve the user experience
Every day we build great products for our clients at Squiz. Part of building these products is understanding how people use them. One way to achieve this is through lean usability testing.
Some of the benefits of this style of testing;
- You test your riskiest assumptions and mitigate risk through an evidence based approach
- You increase confidence with all your stakeholders
- There are no research labs or lengthy reports
- Just quality actionable insights that improve the metrics that matter to business
Where you can you apply lean usability testing
Lean usability testing can be applied to all customer facing products. Some of the common scenarios you can apply usability testing to include:
- A person trying to self serve on government website
- A student setting up their account and applying for a course online
- A person signing up and depositing money into their online bank account
What is usability?
Steve Krug says, “a thing is usable if a person of average — or even below average ability and experience — can figure out how to use the thing for its intended purpose, without it being more trouble than it’s worth”.
Usability problems are not hard to find. Have you ever encountered one of these doors? Are you meant to push or pull?
The Norman Door
If you’re confused you’re not the only one. Every day people have trouble getting through these doors and for this reason they have been given a name — the Norman Door.
What is lean usability testing?
At its core lean means creating more value for customers with fewer resources. Lean usability testing helps identify usability problems and create value for customers, using limited resources. It’s not to be confused with a focus group where you ask people about their opinions.
Lean Usability Testing is the art and science of testing the designs you have created with your target audience and then turning your insights from observation into actions.
Taking action on your insights
Following an evidence based approach gives you confidence to take action on the insights you find. This is because an evidence based approach ensures your findings are valid and reliable. There are many things that need to be considered with such an approach however we will focus on two here. These include;
- How many people do you need to speak to?
- What are the most common cognitive biases?
How many people do you need to speak to?
For each primary user group, five people can help you find 85% of the usability problems. However speaking to three people will generally help you find more problems than you can fix.
The most striking truth about the graph is that speaking to zero users gives you zero insights.
As soon as you collect data from a single user, your insights shoot up and you have already learned almost a third of all there is to know about the usability of design.
As you add more and more users, you learn less and less because you will keep seeing the same things again and again. After the fifth user you are wasting your time observing the same findings over and over.
What are the most common cognitive biases?
There are many cognitive biases that you can fall prey to with this style of testing. We will cover three here;
- The observer effect — when the researcher subconsciously influences the way people respond.
- The framing effect — when the way in which the questions are asked impact the way people respond.
- The helping effect — when the researcher helps the user and they become reliant on their help. Helping users can lead to all sorts of problems.
How do we apply this?
A client from the Government sector approached Squiz to build a site that allows people to find and download spatial data. At the beginning of the engagement we had a lot of assumptions about how the final product would work and a limited amount of data to support us. Before building the live version we built a clickable prototype and tested 15 high priority tasks with 5 users.
Speaking to five people allowed us to find more usability problems than we could fix. It also disproved some of our riskiest assumptions. The findings indicated that 77% of all tasks completed were successful while 23% of the tasks completed were unsuccessful. We focused on fixing the tasks that caused the most problems first and then moved ahead confidently with the build and release.
This approach highlights how following an evidence-based approach can yield great outcomes. We were able to reduce risk and increase confidence with our client in under a week. Had we not conducted the testing we would not have had any data to support our decisions. Not having data to support product based decisions is risky business.
The next step to compliment the lean usability testing would be to setup analytics on the live site. The purpose of this would be to gain a deeper understanding of how people are using the site to then make additional refinements.
- Lean means creating more value for customers with fewer resources. Lean Usability Testing is a cost effective way to achieve this.
- Five people can help you find 85% of the usability problems. However speaking to three people will generally help you find more problems than you can fix.
- Following an evidence based approach gives you confidence to take action on the insights you find. This is because an evidence based approach ensures your findings are valid and reliable.
- Why you only need to speak to five users by Nielson Norman Group
- Do-It-Yourself Usability Testing with Steve Krug
- The Scrollwheel
- Usability Testing of Fruit
About the author
David Di Sipio
Technology is all about people. David helps create great products by focusing on what makes people tick. He is a registered Psychologist and UX Consultant at Squiz. David’s approach is grounded in academic research, big-data, best-practice and ethical practices. The work he does leads to improvements in metrics that matter to business. If you have any questions feel free to reach out to him on LinkedIn.