What's the real impact of usability testing on digital products?

When developing a digital product, designers and product managers usually raise many doubts about usability research costs and timings.

Let's dig into the real impact of usability testing on digital products: why they're important, learn the best practices, and break common myths.

What is a usability test?

It's a user experience research method that mixes qualitative and quantitative research methods. Also, it's the number one technique for determining how usable your product is.

“It means making sure that something works well: that a person of average ability and experience can use the thing — whether it’s a website, a toaster, or a revolving door — for its intended purpose without getting hopelessly frustrated.”

— Steve Krug, the author of “Don’t Make Me Think”

A researcher (called a “facilitator” or a “moderator”) asks participants to perform tasks, normally using one or more specific user interfaces, by observing them as they attempt to complete tasks on it.

Why a usability test?

There are many reasons why it's essential to do usability tests in your product, and if you're thinking:

"The main goal of a usability test is to discover usability problems"

It's more than that! The primary reason for conducting usability tests is to raise awareness among stakeholders, developers, and designers on how users interact with the product.

Sometimes the main concept of a feature might be so disruptive that it needs to be reviewed, even if it didn't show any critical usability issue.

This UX research technique will help you and your product team to:

To get the best from usability tests, let's enumerate some best practices.

Best practices in usability testing

1. Prepare a detailed plan

It'll help you to define your research with every detail: scope, purpose, timing and location, equipment needed, communications, participants profile, scenarios with the script, goals, and KPIs, etc.

Many designers or product managers ask how many users are necessary to do usability tests and when to do it. About these two subjects, let's break other myths:

"5 usability tests discover 85% of problems"

This isn't quite true, because usability tests should be run with 5 users per segment.

It's possible to test with only 5 users, but you won't cover 85% of all problems in your product. If your product has 2 user personas with different needs, you'll need to test it with both groups.

There's no magic number — it depends on the type of digital product, the competition, and the segment of users.

"We need to launch an MVP asap, we do testing later"

The product idea can be spectacular and game-changing in the industry, and the intention is to solve people’s need(s) about something.

But, if your users can't use it, there are two scenarios:

💡 Pro tip: If you're wondering when to run usability tests, probably it's right now.

Ideally, do at least 3 rounds of usability tests: during the wireframe phase, mockup phase with a low-fidelity prototype, and the last round with a staging version of the product, even if it is an MVP (Minimum Viable Product).

The cost of not running any usability testing before launch an MVP product is up to 2x more time from initial development to fix issues later and 2x more money spent with a higher chance of critical problems that can increase even more the costs.

2. Choose the best method for your research

There are 3 methods:

Consider the budget/timeframe available, cultural impact, and what type of insights you need to gather from the usability tests. The chosen option should be documented in the plan and why.

"I did a prototype in InVision/Figma, and I tested it with a friend or a coworker"

It's ok, but it isn't a usability test. If you need to see another person using your design, do it! Just be careful with any NDAs.

This isn't a usability test because:

The results from this "testing session" won't give you real data but biased and unrealistic insights, making your product deliver a solution that doesn't match your users.

3. Don't provide a walkthrough guide

When writing the tasks, avoid creating a walkthrough guide with the same words present in your user interface. If you give too many clues, the user will follow it without thinking, which's unreliable.

💡 Pro tip: Describe scenarios and let the user find them out. For example: “Find an article about football in our blog” versus “You want to know the last news about yesterday's football match."

4. Don't run away from the Pilot Test

Even if you're short on time, don't skip the pilot test. The test will show you how well prepared you are and solve potential technical issues.

In this stage, you can invite someone in the company to be your "user." 🙂

5. Together, we are stronger!

Include other team members in the testing session without the participants seeing them — not only the design team.

They may target something that you didn't notice and develop more knowledge about the users.

6. Run testing sessions up to 60min

We want to test everything but don’t want to overwhelm our users.

During the session, don’t interrupt the user with questions. If you are next to the participant don't stand in their vision or if you are remote, turn your camera off — they'll feel the need to ask you if they're doing well and ask questions.

If you want to ask some questions, leave them to the end of the tests.

💡 Pro tip: If you're an expressive person and have difficulty keeping your "neutral face," leave the room or train someone to moderate the tests for you. Otherwise, you may influence the participant to react to your body language.

7. Create a Findings Document

Build a report with all the data gathered, which summarizes the usability issues and a list of actionable points by priority. Use it to discuss it with your teammates.

"When users are quick to perform a task, it means the flow is good"

Speed itself isn't a good metric to define a task performance because it depends on the context and the product. Also, you need to have a comparative metric.

In usability tests, you want to see how well users performed a task, what their body language was, and their feedback in the end.

💡 Pro tip: You should use the data to validate your vision/ideas, but be careful not to “manipulate” its representation.

For example, “50% of users didn’t found the News section” versus “1 in 2 users didn’t find the News section” — looks the same, right? If the data sample is 20 or only 5 participants, it has an entirely different impact and communicates other priorities on the product.

Conclusion

Testing your product during the different design stages increases the product success and advocacy from users, resulting in high engagement and revenue to the business.

It also decreases the costs of releasing something that users can't use, hate or have so many problems that overwhelm your customer support channels.

So, what are you waiting for? Let's start testing!

Follow Pixelmatters on Twitter, Facebook, LinkedIn, and Instagram.

Catarina Borges
UX Designer