One of the best ways to determine if your exhibit- and/or event-marketing strategy is having the intended effect on your audience is to directly poll your visitors. But to devise a good survey, whether it's executed in your booth, via a digital experience, or at an event, you need more than a bunch of cobbled-together questions. Customers and prospects are crazy busy in the best of times (which these clearly are not), so when you manage to slow them down long enough to provide feedback, your survey needs to quickly and effectively obtain the crucial information you need to assess and improve your program.

Granted, if you want highly detailed and statistically accurate results, you'll probably want to hire a professional survey company to handle the writing. But if your budget is tight or you're comfortable with obtaining information that is on target but not necessarily in the bull's-eye, then you may want to craft your own questionnaire. To help you do just that, we asked four survey experts for some of their most critical insights. They provided the following seven essential survey-writing recommendations along with a couple of common pitfalls to avoid.

Define Your Target
"Before writing the survey, figure out exactly what you need to learn from it," says Jeff Kidd, owner of technology-solutions provider BrightPipe. Did your product launch convey your widget's selling points? Does the new design of your booth help your company meet its marketing goals? Are staffers able to provide necessary information? Once you know what you want to achieve, you can start developing questions.
Along those same lines, our experts agree that you need to eliminate "nice to know" insights and focus on "need to know" questions. To do that, consider the data you're likely to obtain from each query. If you don't have a way to somehow implement this feedback, then don't ask the question, says Jack Semler, CEO of Readex Research Inc., a provider of survey services. For instance, if you want to get visitors' thoughts on your company's new logo but know hell would freeze over before management changes it, then there is no point in asking. The idea is to fill your survey with probes that provide insight you need to know in order to carry out an improvement.

Keep it Short
While there are no hard and fast rules about how long a survey should take, almost all experts agree that it's better to err on the shorter side. "Remember to always balance your needs with your respondents' experience," says Lynn Kirincich, marketing and promotions manager at market-research firm Signet Research Inc. "If you only ask for what you really need, your completion rate will be higher and your feedback better." The longer a survey takes to complete, the higher the dropout rate is going to be, leaving you with potentially unreliable or unusable data.

Kidd adds, "Research studies have shown that audiences begin to lose focus after seven questions." Joe Federbush, president of Evolio Marketing Inc., is in the same ballpark with his benchmarks. "Surveys from exhibits less than 200 square feet should take no longer than three minutes and have no more than 10 questions. Medium and large exhibits' surveys may take as long as 10 minutes with about 20 questions. And online surveys that take longer than 10 minutes and have 20 to 25 questions start to show substantial levels of respondent drop-off." In short, brevity is your ally.

Include Null and Neutral Answers
Avoid leaving respondents without an option that fits their experiences, as it can make them feel the survey isn't for them. One way to bypass this pitfall, according to Semler, is to include null options, i.e., answers such as "none of these" or "not applicable." Imagine your survey asks "What is your position in your company?" and none of the included options fit some respondents' roles. Instead of forcing them to choose an inaccurate title and skew your results, simply include a "none of these" or "other" option.

Similarly, if your questions ask for a numeric ranking, experts suggest using odd-numbered scales (e.g., a range from 1 to 5) so there's a midpoint or neutral rating. For example, if your query is "Rate your experience in our booth today with 1 being poor and 4 being excellent," those who feel the experience was neither positive nor negative won't have an option because selecting a 2 or 3 forces them to lean one way or the other. According to Federbush, it's better to have a scale from 1 to 5 so that neutral respondents can rate their experience a 3 and thereby provide more honest feedback.

Be Crystal Clear
Nothing kills a survey's reliability faster than a question respondents don't understand. "Be clear and specific," Kirincich says. "Avoid vague wording, industry jargon, complex terms, and acronyms." The goal of your questionnaire is to ensure that each respondent interprets questions the same way.

For instance, if you ask "How often do you like to attend trade shows and events?," visitors could get hung up on the word "like." Some might attend lots of trade shows but not necessarily enjoy them, which may cause them to answer differently than a respondent who really likes events but isn't able to attend many. Still others may simply assume you meant to ask how many they'd ideally attend each year. In this instance, you're better off dropping the "like," says Federbush, and asking something along the lines of "How many trade shows and events have you attended in the past 24 months?" Use Open-Ended Questions Sparingly "Open-ended questions can be useful to obtain unaided feedback," Semler says, "but they should typically be used sparingly. These questions require more work for survey participants, and too many may increase dropout rates." For example, if you begin your survey with, "Describe your interactions with booth staff, citing specific examples," you probably won't get many people to finish the first question, much less move on to the second. Rather, get attendees invested in the survey by opening with closed-ended queries and then sprinkle any necessary open-ended questions toward the end.

Another alternative is to simply give respondents the option of skipping a question if they have nothing to share. This tactic also ensures that only those with something important to communicate will do so. Having said that, don't be gun-shy about soliciting open-ended responses. "While they cannot be precisely measured, they can be examined for patterns and categorized to identify common themes," Kirincich says. And the patterns that emerge can be a goldmine.

Offer Incentives
"Incentives can provide between a 1- and 5-percent increase in your response rates," Semler says. And while that may not seem like a boon, every percentage point counts when it comes to establishing a pool large enough to be statistically significant. Given the potential benefit, you should consider offering respondents a little something in exchange for their time and participation.

Sources agree that survey incentives don't need to be extravagant – a $5 gift card to Starbucks or Amazon.com will suffice. Alternatively, you can opt to enter all respondents in a drawing for a single, more substantial prize. "Simple incentives can improve the quality of attendee's answers, as they often feel they're making a fair exchange of quality information for the incentive," Kidd says.

Don't Forget to Follow Up
If you're among the many face-to-face marketers who are now, thanks to COVID-19, limited to conducting surveys online, blasting out a single invitation isn't likely to snag you a sufficient number of responses. But on the other end, sending emails over and over to everyone in your database isn't the correct approach either – and will likely annoy people who've already taken the survey. So what frequency is appropriate? Follow-ups are crucial to your rate of return, according to Federbush, who notes that reminders often increase response rates by as much as 50 percent. The trick, then, is to maximize your response rate without spamming your audience to the point where they opt out of any future correspondence.

To maximize replies, Semler says Readex Research typically uses a three-wave emailing technique in which an invitation goes out to all sample members and the next two "waves" target nonresponders. "Invitations are usually sent five days or so apart, and we see the bulk of responses come in within 48 hours of the invite," Semler says. "This means it takes approximately two weeks of fieldwork to conduct the survey." E