Making Surveys Work for You
I’ve seen it go like this: “I’ve found some marketing budget left to spend before the end of the year, and I’ve talked to Analyst Firm ABC about doing a survey for us. I’ll arrange a meeting to put some questions together ASAP so they can get started.” From there, this often deteriorates into a waste of time and money. Yeah, the budget was spent in time, but the results were no more useful to the company than if I had spent it on shoes.
As with anything, knowing where you want to go before you start driving is critical. Here are a few things to think about before the first survey question is written.
First ask WHY!
Why do you want to generate a survey? Are you trying to get customer feedback on product direction, pricing, or messaging? Do you want gauge reactions to a design or a marketing piece? Or perhaps you want to craft a thought-leadership article on a hot topic that aligns with your strategic direction? Are you hoping to learn what technologies or products your potential customers are planning to purchase in the coming year? Or maybe you are hoping to develop information to support customer success stories. The better you know your objectives, the more likely the survey will help you achieve them.
Whom do you want to survey, and how will you identify them? This should rarely be a shotgun approach, so how will you determine the ideal survey respondents? If you are planning to survey existing customers, then you may have a lot of good demographic data to work from, plus details on what products they already own (not to mention their contact information). But if you’re going outside your customer base, do you have good contact lists (aka “mailing lists” back in the old days)? It’s likely you will need to rely on client lists from the firm who will be conducting the survey or—worst case—rent a list to use. Make sure you adhere to data privacy policies and regulations. Consider also whether one or more partners or sponsors might be interested in this effort and have contact lists to contribute. If you survey the wrong people, it’s unlikely that your survey results will be meaningful.
How many responses do you need? If you are looking for feedback for internal purposes, even a relatively small number can be useful. If, instead, you are planning an infographic or article for external distribution, you need enough responses to be compelling. I wouldn’t be impressed by an infographic saying that 4 out of 5 customers realized certain benefits if I found that only 5 customers were surveyed. If you plan to publish formal survey results (often done with partners or multiple sponsors), you need even more responses to ensure that the results are statistically significant. Be sure to send out enough surveys that even a modest response rate will do the trick—if not, you’ll probably need to send out a round 2.
How will you collect the survey data? Of course, you can put a bunch of questions in Survey Monkey and send out electronic surveys to your entire contact list, but this tends to produce a low response rate and perhaps questionable results. However, if you are looking for easily quantified data, this may be enough. More qualitative information is best left to telephone surveys. They provide more interaction, the opportunity to gather more color and details, and can produce better results, but this can take a lot of time and effort. You can even use a hybrid method. For example, present a simple survey in conjunction with a conference (live or virtual) to identify those who might fit your target profile—and get their permission for later contact. Then follow up with your electronic and/or telephone survey soon after the conference.
Who is going to do the work? If you plan to engage an outside analyst or other firm for survey distribution, interviews and even analysis, be sure they have background and credibility in your industry and have done similar projects. Discuss up front what you will expect them to do if the response isn’t adequate due to insufficient responses. If they will be conducting phone interviews for you, be sure those who will actually conduct the calls understand enough about your business, what type of data you are looking for, and how to drill into interesting details the interviewee might be able to provide. They not only need to understand the questions and answers, but also understand when one answer contradicts another so they can adjust accordingly. Over-communicate up front on your expectations so that neither party gets unwelcome surprises at the end of the project.
Plan your survey questions carefully
In general, the more questions you ask, the less likely you’ll get a complete survey response. There is a tendency to think, “As long as we are going to the work of sending a survey, we want to cover everyone’s questions to get the biggest bang for our buck.” Speaking for myself, if I receive a survey on a topic of interest and there is an indication that the survey should only take 5-10 minutes, I will probably try to answer it. If I find too many questions, on the other hand, I’m probably going to bail. As you write your questions, consider these tips.
Prioritize your questions. There should be a small number of questions that are of key importance to you. Know in advance how you plan to use the data. Perhaps you are gathering data for an infographic or a customer-facing presentation you are creating. In that case, asking several low-value questions—other than for basic qualification of respondents—is counter-productive.
Avoid asking “difficult” or cumbersome questions. If a question requires research before answering, you have decreased the likelihood of a response. If you must ask this type of question, it’s better to interact with a small, close group. When I was at SAP, for example, I posed detailed product usage questions to members of our customer advisory council, but I avoided sending a blanket survey to all customers of the product.
Aim your questions at one user. I’ve seen surveys that ask questions that require the knowledge of multiple users to fully respond. This usually results in a lot of “I don’t know” responses and frustration on the part of the respondent. It’s better to narrow your questions and issue separate surveys if needed. If you are doing phone interviews, though, you might be able to get contacts for other types of questions for future use.
Make your questions and possible answers clear. Beware of long questions, those using a lot of technical jargon, and any that might be easily misunderstood with a quick reading. Keep in mind that for an international audience not all respondents may be native speakers of your language, so avoid idioms and make response choices explicit. For example, if you provide a 5-point rating scale, a 5 might be great in some countries, whereas a 5 is awful in others—so be sure to specify what each rating means. Generally use multiple choice questions, and make sure your range of answers isn’t overly influenced by how you hope respondents will answer.
Organize your questions into logical order. If you start with general questions and proceed to more detailed questions, this can help respondents understand the context. Keep questions about similar topics together, as skipping around can cause confusion. That said, keep in mind that the questions least likely to be answered are the ones at the end, so make sure your key questions do not come last.
Test your surveys before sending to your list. It’s amazing how many times I’ve seen survey questions that seemed obvious to the creator but confusing to me! And I’ll even confess that I have crafted some questions that were later misinterpreted by respondents—sometimes because of their background, other times because I didn’t word the question carefully enough. So it’s a good idea to first try the questions on internal folks and see if their responses and reactions to the questions are in line with what you expect. Then, if feasible, send the survey to a limited number of respondents on your list. Review the responses and adjust your survey questions, if needed, before sending out to the wider list. Use a similar process for telephone interviews. This can eliminate MANY problems before they get out of control.
How to increase survey response rate
Most people are glad to express their opinions if it doesn’t take too much time or effort. Here are a few tips to help increase your survey’s response rate.
Dodge the SPAM filters. There are many articles available to delve into this, but a couple things stand out to me. If the sender of the email is already known to the recipient, it’s more likely to go through AND elicit a response. Avoid common SPAM trigger words—again, there are many lists online (things like “act now,” “don’t delete,” “urgent” and many more). Also avoid too many links or images, which can also trigger a SPAM filter.
- Create a meaningful subject line. Give this some thought, as you need to create interest while carefully avoiding SPAM trigger words. What would make you open the email if you were the recipient? Certainly not something with ALL CAPITAL LETTERS and a myriad of exclamation points!!!
- Set a context for your survey up front. Something like this: “We appreciate your purchase of our product and hope you will help us improve it by answering the following questions. Your responses will be aggregated and shared with our product management team.” In other words, give those surveyed a reason to respond. At the same time, indicate how the data will be used and address potential data privacy concerns.
Let them know how long it should take. If it should take 5-10 minutes, say so up front. Either put all questions on a single screen or page, or include a progress indicator so they can see when they are almost done.
What’s in it for the respondent? Depending upon the type of survey, you might want to include a participation reward of some sort. This might be a small gift (be sure to account for related costs), a discount, a later report of survey results, or a linked article not available to the public. While not necessary, it can be a nice touch.
Surveys can be a valuable tool, requiring planning and careful execution to be successful. Aventi has a variety of services that can help you plan and design your surveys or expand your existing skills. In addition, our content development and content marketing professionals can help you put your survey data to the best possible use. So if you have any questions or would like assistance, please contact us here. You can also learn more by reading our Content Marketing Practice Brief.