Forum Moderators: open
Background: When users register for an optional service on a web site, it is often desirable to collect some basic information about those users to help you refine your services. Question is, do make it a requirement that the user provide information about themselves in order to register, or do you make it optional? And if you make it optional, should you approach them with a "hard sell" or a "soft sell" to get them to fill out the optional personal survey?
On my nonprofit web site, visitors can register to receive quarterly e-mail newsletters on a range of subjects, and we decided to make the personal survey optional so as not to scare off potential subscribers. At first, we opted for a "soft sell" approach; after the user had registered, we confirmed their registration, told them when to expect the next newsletter(s), and asked if they would be willing to tell us about themselves. Their two options (on buttons) were "Yes" and "No thanks." Clicking "Yes" opened a new page with the survey. Clicking "No thanks" returned them to the page they were on when they decided to sign up for our newsletters. The survey was short and impersonal. All we had was their e-mail address and answers to six questions about the kind of work they did and what city, state, and country they were from.
We found that 64 percent of subscribers elected to fill in the survey.
After several months, we decided to test a more "hard sell" option to see if we could get a higher rate of responses: instead of asking if they'd complete the survey, we put the survey questions right on the subscription confirmation page, with buttons for "Submit survey" and "No thanks." After the text confirming their subscription was a short appeal asking them to complete the optional survey.
The original "soft sell" approach was chosen because we were basically walking on eggs and didn't want the "core" users who signed up for our newsletters to associate us with the 'untrustworthy' practices of some web sites. I theorized, however, that if they could see exactly what the survey asked, they would be more likely to complete it; in the "soft sell" approach we began with, I felt we didn't give them enough information on which to make an easy decision whether they wanted to fill out the survey or not. I suppose I should really call this the "more information" rather than "hard sell" approach.
Well, the results are in, and the "more information" approach wins -- 89 percent of subscribers elect to complete the survey with this design approach, which is a pretty good improvement over 64 percent. Combining the confirmation with the survey questions not only gave the user more information, it also saved them a step. I suspect those factors are what contributed to the higher capture rate.
I'd be interested in hearing other stories about how design affects users' willingness to provide information.
