4 comments

  • Mike BulajewskiMike Bulajewski, almost 3 years ago

    I'm guessing that UserBrain doesn't offer demographic targeting.

    There are many good reasons to test with representative users. Age and experience with technology are two factors that are important to match with your user group. You probably don't want to test an Android app with iPhone users or vice versa.

    Domain knowledge is often important. If you test an analytics tool with general users, you'll get a lot of false positives and waste time identifying "issues" that your users don't experience.

    0 points
    • Stefan RösslerStefan Rössler, almost 3 years ago (edited almost 3 years ago )

      Hi Mike,

      you're right, Userbrain doesn't offer demographic targeting. There are 2 reasons for that:

      1) We don't have enough qualified testers yet

      2) It's not necessary for most websites (Userbrain is only for websites) to test with representative users

      E-commerce sites are an obvious example, but there are many web-based services out there, that should really try work for everyone (from a usability perspective) which is why testing them with random web users just works.

      You're probably right about the problems of testing an Android app with iPhone users or vice versa. I wasn't aware of platform differences when I decided to add the word "app" throughout the article. Thanks for pointing this out :)

      One last word about the danger of false positives: I think the worst that could happen, is that you're getting no (or inconclusive) results if you're testing e.g. an analytics tool with someone who has no idea about it. Identifying non-existent problems is not really a problem in my opinion.

      Thank you for reading the article and taking your time to write a comment. I really appreciate this :)

      0 points
      • Mike BulajewskiMike Bulajewski, almost 3 years ago

        One last word about the danger of false positives: I think the worst that could happen, is that you're getting no (or inconclusive) results if you're testing e.g. an analytics tool with someone who has no idea about it. Identifying non-existent problems is not really a problem in my opinion.

        I disagree. How do you know if a false positive is false? If testers can throw out any results they don't agree with, it defeats the purpose of testing.

        There are cases where you don't need an exact match for your target audience, and it would be interesting to read an article that explained where it makes sense.

        0 points
        • Stefan RösslerStefan Rössler, almost 3 years ago

          If testers can throw out any results they don't agree with, it defeats the purpose of testing.

          The purpose of testing (how it's described in the article) is not to understand people's opinions and what they agree or disagree with. It's to observe them and to draw your own conclusions based on other tests, your experience, professional judgement, and common sense. I think the danger of false positives is low, because usability tests are usually not your only source of knowledge (nor should they be – even if you're testing with your exact target audience).

          Thanks for the idea about writing an article on when you'll need a target audience and when it's save to ignore it. I'll definitely start thinking about this in the future :)

          0 points