Setting up a market test is by no means a difficult task, but it does require foresight and patience. In the direct or database marketing worlds, we might test offer, copy, medium (e.g. postcard vs. letter), subject lines, response rates and so forth. In online marketing research, several options are available for testing. Let’s take a look at a few…
Releasing an online survey is no different than any other direct marketing campaign. We are concerned with response rate, percent completing the survey, and the quality of the responses. We also have a vested interest in managing the cost of our marketing research projects. To this end we can test the times and dates that we release our surveys. There are several existing benchmark studies that report on the best times for releasing emails to ensure higher open-rates. Use these as a starting point. Nothing can beat the results you achieve with your own data.
Response rates are a function of release time and date, but also subject line, awareness of the ‘from’ address, and incentive (or absence thereof). All of these can be tested. For example, subject lines of varying intensities and character length are suited for head to head comparison. Use of an incentive in the subject line such as:
An 8GB flash drive to the first 50 respondents…
Or should the incentive be mentioned in the invitation copy? Different types of incentives are known to appeal to different groups. In B2B marketing research it is common to offer a copy of the results to respondents. Researchforgood.com offers respondents the option of donating an incentive to a charity of their choice.
When considering survey design we can incorporate differing question formats. The format becomes the test variable. On the back-end you would compare the data for the test and control groups for differences. Surveys designed to promote differing levels of engagement, perhaps incorporating visual, audio or graphic template, should include a question that asks about the experience. The ‘new and improved’ survey (the challenger) would be compared to a standard in the search for a new champion.
Testing keeps not only the data collection process fresh, but also the researcher.