Selecting the Beta Test Group

Option 1: Recruit users from your existing website.

This is a good option if you're expecting your first wave users to be those who have already developed some sort of relationship with other websites you've built. As a matter of practice, this is the option we most frequently use since it's both convenient and appropriate to get the opinions of some of your core customers before you open the doors to the cold, cruel world. Solicit these users by means of a pop-up JavaScript invitation, set to sample a small subset of the visitors accessing the front door (or other appropriate crossroads) on your site.

Option 2: Recruit users who have participated in earlier research.

Perhaps you've heard the gospel and have conducted earlier audience or product development research. If you've had the foresight to realize that these willing guinea pigs might be useful again in the future, you might have included a question asking permission to contact these users again. If so, you're in luck! Simply pull together a list of e-mail addresses along with pertinent background information you'll use to profile the testers. We can already hear you asking the question: "What, exactly, is pertinent information?" We'll get to that in a moment. In the meantime, here are some other options for beta-test recruitment that you're bound to use sooner or later.

Option 3: Recruit users from a panel.

If you're starting from scratch, or plan to market this new product to an entirely different audience than you're currently reaching, you'll need to find testers who represent your target market. Online panels can be a real time-saver here. Panels are groups of Web users who have agreed to be contacted and participate in a variety of marketing research studies and who have typically provided the panel company with a host of background information on their characteristics as consumers and Web users.

What this means to you is that you can custom-order a group of testers who meet your criteria. In fact, you can gather several groups, each having special meaning to your business goals. Let's say you're testing a site designed to provide product information and technical support for a line of networking equipment. You expect to have two general categories of users:

• Purchase decision-makers (the ones who authorize the purchase or sign the checks).
• IS/IT professionals who actually install and run the network.

Option 4: Recruit from convenience samples.

A "convenience sample" is a research term used to denote groups of respondents (or testers) who were not gathered in any statistically representative manner. It's research-speak for "a wee bit shoddy." Convenience samples are advantageous because they're (as the name suggests) convenient to get, and can be more flexible in use. They can be cheaper, too, since you're not covering the overhead and profit margins of a large research house to access their panel. Convenience samples can come from almost everywhere: that list of e-mail addresses from the fishbowl at the last internet conference you had a booth at, the people who downloaded a demo of your last software release and, most commonly, lists of e-mail addresses who've opted to receive e-mail about your product category.

There are also "opt-in" lists, sold by so-called list brokers who typically sell to direct-marketing types. They generally price lists by the name, and offer you broad interest categories such as "investing" or "webmasters." Individuals on these lists are used to receiving lots of marketing e-mail, so you'll need to put some time into making your solicitation stand out from the pack. Emphasize the importance of their feedback in driving the site's development. And offer them some incentive, to make sure that you're not just getting the odd user who's got nothing better to do with their day than give you grief about your new site.