In a recent 2016 study of higher ed marketing leaders, we found that approximately 50% of colleges and universities are currently implementing a marketing research initiative. The higher ed marketers in us were excited to see this number so high, since this critical step is sometimes overlooked due to time or budget constraints. However, the data geeks in us immediately started to wonder about the specifics: What survey software are they using? Are they surveying the right audiences? Are the data valid?
Regardless of whether you are implementing market research internally or hiring an external consultant, there are many things you should consider to ensure you get the most out of your research. Here’s a look at some of the key questions you should ask yourself, as well as our recommendations for designing a survey that is accurate AND gets you the critical insights you need to inform your strategy:
What kind of survey software should I use? There are many options out there for implementing online surveys. Some free versions are limited in what they can provide. We recommend that you pick a solution that offers you the following, at a minimum:
- Ability to send individual, authenticated links through a panel
- Ability to branch and/or create skip logic based on responses and embedded data
- Ability to rotate or randomize response options
How do I make sure I am reaching the right audience? It is important to ensure your survey is reaching the desired population, but how do you do that?
- Use Authenticated Links: Whenever possible, use your own distribution lists to send unique, authenticated survey links. This means that each respondent is assigned a survey link that is unique to them and can only be used once. This allows you to track each survey response back to the person you sent it to.
- Use Screener Questions: When dealing with external audiences—for example, prospective students, higher ed peers, guidance counselors, community leaders—screener questions are a vital tool for ensuring each respondent fits within the parameters you have set to define that audience. Screener questions are typically demographic questions that either validate demographic information you already have for the respondent or add new information that may not have been available in your file. Anyone who doesn’t match the established parameters can be omitted from your respondent pool.
Should I tailor the survey to the audience? You should always make sure that the questions you are asking each audience make sense for that audience. However, that doesn’t mean you need to program separate surveys for each audience. Branching and skip logic tools make it easy for you to choose which audiences see each question.
Wait, so I CAN’T use an open survey link? The question you should be asking is “are we ONLY reaching the desired audiences?” That is, are we possibly collecting responses from the wrong people? Can you authenticate that the survey was taken by the person you sent it to? That the desired respondent only completed the survey one time? Watch out for instances where you can forward a link to share with family, friends or colleagues, or even situations where a respondent can take the survey more than once. An open link, a URL where anyone can access the survey as many times as they want, can affect the validity and quality of your collected data and should only be used in rare cases as a last resort. If you have a list source and are driving survey traffic through email, there is no good reason to use an open link.
There are certainly many details to consider as you embark on a market research study. And we are here to help and love to talk research! If you have a specific question nagging you, feel free to reach out to us at hello@simpsonscarborough.com. We’ll be back in in 2017 to cover some additional questions you should consider as you get started on your next research initiative.
Also, stay tuned for the full results to the 2016 higher ed marketing study that will be published in December. Results from the inaugural 2014 can be downloaded here.
