Improving Your Beta Program with Personas

Personas are a big part of product marketing, but they’re generally underutilized in beta testing. We have some clients who’ve carried personas into their beta programs, but they’re in the minority. On the other hand, they’re a very successful minority. So, in this blog post, we’re exploring ways that you can apply customer persona techniques to help improve your next beta test.

Before we can discuss how to use personas in your beta program, we should first talk about what they actually are. Personas are a set of fictional characters you create whose characteristics match specific identified segments of your target market. Thus, each one will have real attributes that represent the customers you’re trying to reach (e.g., age, location, product skills, pain points, etc.), as well as some that are merely biographical (e.g, a name, nickname, personal background, etc.) The idea is that once you associate these characteristics with people (even imaginary ones) it’s easier to plan your product and marketing efforts to accurately meet the underlying needs they represent.

Note: If you’re not sure how to develop or manage your personas, there’s a tool coming out soon called Usersbox that streamlines the process. They’ve been nice enough to offer us five beta accounts to give away to readers, so the first five readers to comment on this blog post will get early access to Usersbox.

Once you’ve developed your customer personas, here’s how you can apply them to your beta program:

Improving Your Recruitment Efforts

We always tell people that the candidates you recruit for your beta test need to be part of your target market. If you’re testing printers for consumers and only select programmers and IT professionals to beta test your product, you aren’t getting a complete picture of how the product will work in the hands of your actual customers. And relying on their feedback alone could lead your product astray.

When you incorporate personas into your beta program, you have a well-developed set of criteria for candidates that improve your recruiting. Instead of casting a wide net and hoping you get the right mix of testers, you have the opportunity to get specific. If Candidate A is a busy, stay-at-home mom, age 30, who is college educated and influences household buying decisions but isn’t entirely comfortable with computers, maybe she fits your “Sarah” persona and your plan includes recruiting at least 15 people like her for the printer beta.

Validating Product Assumptions

During product development, you have to make a lot of carefully reasoned assumptions about your customers and how they’ll use your product. One of the great things about beta testing in general is that you get to test how accurate those assumptions are in practice, rather than in theory, before you release to the general public.

Personas add an extra dimension to that validation process. If you designed certain features and work flows to match the needs of specific customer segments, you can assign tasks and surveys to testers who fit the corresponding persona and see if those parts of your product are working as expected. For Centercode software customers, this is as simple as creating teams within your beta project for each persona and assigning your testers accordingly. You can also use those persona teams to report on feedback, slicing and dicing your data to get a precise view of how each team is doing throughout the course of the test.

Prioritizing Issues and Feature Requests

The final idea we’ll discuss is how you can leverage personas to help prioritize feedback from your beta. There are many approaches beta managers use for triage. Some use how many times a bug or feature request came up as a signal of importance, others decide which bugs or features to tackle based on the difficulty of fixing them. When it comes to prioritizing incoming feedback, though, you can never have too many tools at your disposal.

What personas add is additional context to help you distinguish what requires immediate attention and what can wait for now. For example, if you see bug reports coming from testers whose persona is not very savvy, you know that it might not be a bug so much as a support or documentation issue. If you see bugs being reported across the entire spectrum of personas, including the very tech savvy, that’s a good indicator of severity. Finally, if you see a bug or feature request concentrated within a specific persona, and that persona makes up a significant portion of your target market, that’s something you’ll want to pay attention to.

This is, of course, only a brief introduction to how you can use personas in your beta tests. There are many other ways, and we’d love for you to share your ideas or experiences in the comments. Remember, the first five who do so will get beta access to Usersbox.

Get the most out of personas with Centercode

You Might Also Like

  • Saeed Khan

    I’ve written about beta programs here:

    http://onproductmanagement.net/2011/03/01/new-resource-building-a-better-beta-program/

    I’m don’t discuss beta personae in that document, but it’s an interesting concept. I’d like to hear more.

  • http://centercode.com Adam Wright

    Thanks for sharing that link, Saeed. We do have some other ideas for how personas can be used in beta management. One, in particular, came from a really interesting thread a week or two ago in the Software Testing and Quality group on LinkedIn. The idea is using a mixture of personas and session-based testing to run a product through internal beta readiness testing. The guy who first mentioned it (Curtis) has a forthcoming article on the topic, and once that is published, we’d like to look at the details in a future blog post.

  • http://testing.gershon.info Shmuel Gershon

    Reading the article got me thinking into another face of
    the Beta&Personas interaction… in the opposite direction:

    If you publish public calls for participants, analyzing
    the candidates that propose themselves as Beta testers can be used to study
    your choice of personas. I mean, if you considered elder retired doctors to be
    part of your user choice, but none of them answered to your Beta programs while
    thousands of young hippies did, then maybe there’s still more you can learn
    about the user base.

    The same goes for direct recruitment. If none of your
    Nancy Persona agreed to Beta test your product, maybe their relationship or
    interest in it is different than you first understood/expected.

  • http://centercode.com Adam Wright

    I’m really glad you mentioned this. I wanted to include similar ideas in the post, but was worried it would become too long. Just like you said, when you incorporate personas into beta tests, the recruitment phase can be very valuable for evaluating the quality and accuracy of your persona development and learning more about your true user base.

    The way I was thinking about it, if you find yourself discarding a lot of promising beta applicants because they don’t fit your personas, that might signal that your personas are either not inclusive enough (in the case of near misses) or that you missed some viable market segments (in the case of outliers).

    The one thing I would caution, though, is to make sure that unexpected results aren’t simply the result of recruiting in the wrong place for your target markets (e.g., if your beta recruitment strategy is an invite code giveaway with a blog like TechCrunch, that could easily explain no elderly doctor applicants but a lot of young technophiles).

    Thanks for the great comment!