Have you ever said something at a party that resulted in immediate and deafening silence? The comment need not be inappropriate. I could simply be that no one knew how to respond. And the more effort you put into the idea or question, the more embarrassed you are. Questions arise in your mind; “what did I say?” Maybe I’m no good at parties.” This is essentially how it can feel to put a survey out into the world only to meet with low response rates.
What did I do? Are the survey questions bad? Maybe I’m no good at this….
If you have ever posted a question on social media and virtually (pun intended) no one responded, then you have conducted a survey with low response rates. But with your business, finances as well as your reputation are on the line. If you really need to know about your customers or why sales are slumping, responses to your survey becomes critical.
First, let’s alleviate some of your anxiety if you have found yourself in this situation, even on a small scale. Even as an experienced researcher, if I get a 10 percent response rate on a survey, I am happy. If your target population is 1,000 people and you receive only 100 responses back, then count yourself lucky because your research methods succeeded. That may not seem like many people but rates like are fairly standard in survey research. National presidential election polls in the U.S. for example are usually assessed with less than 10,000 respondents, in a country with over 330 million (so only a 0.003% response rate). Some product analysis is started and ended with a focus group of only 12 people. Millions of dollars of reshoots might be done on a movie based on the critiques of a 100 people in a screen test. Numbers of respondents always matter the quality of the data is what counts.
10-percent or better is a good rule of thumb in survey response. You’ll also want to make sure your respondents reflect the population you are examining. If you are studying a general population, but 90% of your respondents are women, then your responses will be skewed. So, knowing who responded is also key.
But let us say you know all of this and the responses are still low. Like if only 10 of those 1,000 responded. What’s to be done? Below are a few ideas for anyone in this situation.
1. Keep It Open Longer.
If you have low response rates and the survey has been out for a week, keep it out for another week. If the second week has even less responses, maybe keep it out longer. Responses decrease like radioactive element half-lives, you’ll get a lot in the beginning, but it will rapidly decrease over time, which is why I don’t recommend really long survey durations (like 6 months to a year). However, it is possible you were anxious to get a response and you didn’t allow for enough time. You may intend a month-long survey time but when the responses are low it’s time to improvise.
2. Know thy Target Audience.
If your responses are low, its always useful to know who you are surveying in the first place. Remember that low responses can be an issue within the response group as well as in the population (i.e. like our skewed ratio of respondents by gender). Knowing the target group in detail can also shine light on why the responses were so low. Survey timing is a common issue. If you are surveying a group of sportsmen and it’s the start of deer season, you should have a good idea why they didn’t respond. If your spring survey is about preparing for winter driving season, you may get lower responses because that topic is not on peoples’ minds at that time of year.
3. Consider Distribution Methods.
When we say a survey is “in the field” we mean it is now published, sent, and people can respond. But that proverbial field has many paths through it. Surveys can be conducted by mail, phone call, email, social media, and the good old fashioned in-person variety. Using common sense and knowing how the audience would be most likely to respond is key. Sending a customer satisfaction survey to people who live in a retirement home by email won’t be as effective as mail or in-person.
And don’t be afraid of in-person distribution. If you can sell in person, you can ask for feedback in person too. In-person methods can often achieve higher response rates since
4. Offer an Incentive.
Complete this survey for a chance to win a new PS5! Stores and companies offer these kinds of rewards to encourage responses. They do this because the benefits of the information they receive far outweigh the cost of the reward. If you run a million-dollar corporation, who cares about a new car or a few gaming consoles. Even if you run a hundred-thousand-dollar restaurant, how about a pair of football tickets or a free meal at your restaurant. Incentives are extremely useful, but they can’t be so big that you see most people filling out the survey just for a shot at the prize. That will result in a biased survey.
Again, know your audience to gauge the right size of gift and appropriateness of said gift.
5. Remind Them.
It’s entirely possible many people set aside your survey with the intention to fill it out later and then forgot. A gently reminder can be incredible for increasing response. A general rule of thumb would be to remind the population once per week for every week the survey is in the field. Your survey is important to you but not necessarily to them, which is why they need the reminder. Also consider reminding them via different distributions methods.
Finally, remember that you are not bad at research if you have a failed survey. This happens to even experienced researchers from time to time. However, it is bad if you have a failed survey and use the data anyway. That may seem harsh, but a failed hiring process resulting in a flawed candidate getting the job can be a huge problem. The same is true here. Low response rates are a problem when they paint a skewed picture of how people think about your questions. Acting on this bad data can lead you to make the wrong decisions for your business. Bad data helps no one, which is why getting a sufficient response is so important.