Why Should I Segment My Customers?

When asked about the ideal number of contributors to test with during studies, we answer that there isn’t a single number of contributors that works across all types of tests.

Fortunately, there are ways to estimate how many users you should test with.

At UserTesting, we often run qualitative remote usability studies. Qualitative studies help direct you to isolate general usability problems and trends by watching contributors use an interface. The goal is to understand why the contributors are experiencing usability problems so that you can make improvements to your design, website, or app.

Before you begin running studies, we first suggest grouping and organizing the target audiences into your customer segments. This segmenting is important because different customers have different needs—one group’s behaviors, likes, or needs may be different than that of another group.

So what’s a customer segment, and how do we create them?

Customer segments—they may sometimes be referred to as "user groups," "personas," or "user profiles"—are groups of people who share certain characteristics. Segmenting your customers means focusing in on the group that buys your goods or services. Doing so helps you to better understand...

• What’s important to the customers.
• Their typical web behaviors.
• How to design for them.
• What makes them convert.
• How to customize their experience.

Customer segments can be broken down in many ways. For example, you could segment your customers by...

• Industry
• Demographics
• Level of experience
• Buying behavior
• Lifestyle
• Customer type
• Income
• Profession
• Technical expertise
• Domain knowledge
• Location
• Web usage

How many customer segments do you need to test?

You may argue that you just want to test your "target audience." And while that’s all well and good, you’ll typically have more than one customer segment within that target audience.

For example, let’s say you are a business-to-business (B2B) company that sells a SaaS marketing solution. You know that your target audience is marketers, but you’re not sure how to break down that group further.

But there are ways! You could segment contributors by:

• Their knowledge about your product/industry: No Knowledge versus Some Knowledge versus Expert Knowledge
• Their role: The "Information-gatherer" versus the "Decision-maker" versus the "Buyer"
• Their buying phase: the research phase versus the purchase phase

How many customer segments do you need to test? That’s specific to your company and what you are trying to learn, but it’s worth taking the time to think about this segment question and consider ways to organize your audience into smaller groups. By doing so, you’ll be able to better target and design for the needs of the specific people visiting your site or app.

Well, now I have customer segments. But how many contributors do I need in each segment?

When running qualitative studies, five contributors is often the accepted number. While there are many schools of thought, one expert claims that five users will find approximately 85% of problems in an interface (given that the probability is 31% of a user encountering a problem).

However, while five is the generally accepted number, there are some instances in which you could include more (or even fewer) users.

When do you need more than five users?

There are some situations when you should test with more than the recommended five users. For example, if you’re running a quantitative test and want to do the following:

• Compare designs and prove one is better
• Gather user opinion and preference instead of user behavior
• Develop statistical proof that X% of users will show the same behavior

When could you use fewer than five users?

Fewer than five users may be acceptable in cases where you expect contributors to run into some of the same usability issues. This could happen when...

• You test with users that have overlapping demographics within different segments.
• Each user is completing the same tasks within all segments.

Do I need contributors for every customer segment? We suggest including 4-6 contributors in each group. By testing with multiple contributors in each segment, you’ll find out how to give your customers a better experience, which will lead to better results.

Please provide any feedback you have on this article. Your feedback will be used to improve the article and should take no more than 5 minutes to complete. Article evaluations will remain completely confidential unless you request a follow-up.