Selling is about more than moving product. It’s about moving people. Touching them. Connecting with them.
The relationships you build with your customers top the list of your company’s most valuable assets. So how do you know if you’re connecting with them the right way?
The customer survey, when done right, is one of the most effective business research tools for developing better marketing strategies.
But it’s not just about asking questions. It’s about asking the right questions that will yield truthful responses and allow comparing your customers’ awareness, satisfaction and attitude with other relevant brands, organizations and events.
These are some of the best practices for building and distributing successful customer surveys that will achieve useful and actionable results:
1. Clearly define your research objectives and address them specifically – don’t try covering everything in one shot
Cut unnecessary questions from your surveys. Every question you include should have a well-defined purpose and a good reason for being asked. Do you really need to know things like customer name, age or income?
Adding questions you thought are “nice to know” can make survey takers abandon your survey without finishing it.
If you think such questions are important, but not critical, add them to the end of your survey and make them optional.
2. Make your survey short to keep respondents engaged
Find the shortest way to ask a question without losing its intent. It’s not just about reducing the word count. Ask questions that are straight forward and simple to understand.
Your survey length is critical for keeping abandon rates low. Our research indicates that if you keep your survey under 25 questions you will achieve high response rates.
3. Avoid leading and loaded questions
Questions that lead respondents toward a certain answer due bias in their phrasing are harmful for your surveys.
Try avoiding loaded questions in your surveys by eliminating emotionally charged language that hints at preferences or assumed facts.
Here are several “real life” examples of survey questions:
In the first example, a researcher wanted to better understand consumer awareness of Prebiotics:
Q1. Do you know the main usage of Prebiotics?
- No sure
Q2. To the best of your knowledge, which of the following statements about Prebiotics usage is correct (please don’t guess)?
- Prebiotics are used to treat high cholesterol
- Prebiotics are used to restore healthy bacteria
- Prebiotics are used to feed healthy bacteria
- Prebiotics are used to kill harmful bacteria
- Don’t know
While over 50 percent of the respondents responded “Yes” to Question 1, only 13 percent were able to answer correctly Question 2 (“Prebiotics are used to feed healthy bacteria”).
Take a look at another example. Subscribers of an internet magazine were asked to respond to this question:
Q. Why do you like our magazine?
- Because it’s informative
- Because it’s available on line
- Because it’s free
- Because it has great ads
- Other (please specify)
Well, if the survey is conducted to better understand the magazine readers, wouldn’t it be more meaningful to ask what is missing in the magazine and what can be improved?
Finally, take a look at this question. The question was asked following a new product presentation:
Q. How much would you pay for the product?
I bet you know the answers received by the researcher.
4. Add smart open-ended questions and create connection between quantitative and qualitative (open-ended) questions
While some of your most valuable and insightful feedback may come from open-ended questions, nothing may intimidate survey takers more than a huge text box.
Ask a brief quantitative (single/multiple choice, “scales”, etc.) question first to create a sense of progress, and then follow up with a targeted an open-ended question such as, “Why were you dissatisfied with the service?”. This approach will also make the answers you are looking for more specific.
5. Keep rating scales consistent, but randomize question topics
Commonly used survey scales can become confusing when the context changes.
If you start using scale where you ask survey takers to choose between 1-5, where 1 = “Strongly Disagree” and 5 = “Strongly Agree” keep the same pattern for all scale type questions.
Do not assign 1 to “Most Important” and 5 to “Least Important” if you had been using 5 as the agreeable answer (“Strongly Agree”) to previous questions.
If you do this, it will not only be confusing to respondents but many of them will miss the change and give inaccurate answers.
If is very beneficial however, for accuracy and quality of your survey, to randomize or mix your question topics (not scales).
6. Consider using interactive survey tools, pictures and movies, but only when necessary
In some cases respondents’ engagement and response rate can be further improved by using interactive survey tools such as Virtual Shelf, Hot Text, Heat Map or other Interactive ranking questions.
When used smart, interactive survey tools can provide better quality and accuracy of your survey results.
Do not overwhelm, however, survey takers with such tools as they also can destruct respondents from the actual question.
7. Make sure your sample statistically represents the targeted population
Sampling is the foundation of all research. Reliable sampling helps you make business decisions with confidence.
A small, representative sample will reflect opinions and behavior of the group from which it was drawn.
The large the sample size the smaller the chance for an error, but the sheer size of a sample does not guarantee its ability to accurately represent a target population. Large unrepresentative samples can lead to wrong conclusions the same way as small ones.
8. Guarantee anonymity and confidentiality, and provide feel of neutrality
If possible, use the Respondent Anonymity Assurance (RAA) approach. This technology allows tracking who has and has not completed the survey and following up with individuals who have not completed the survey.
In addition it allows identifying each respondent and linking his/her specific responses to any additional, customer/ employee specific information such as level in the corporate structure, demographic information, tenure, etc.
In RAA enabled surveys, computer generated identification numbers for individuals are generated. The researcher, in this case, does not have access to both the respondent’s personal information as well as the response data at the same time.
Surveys conducted by an independent research company using RAA usually deliver more honest responses as compared to surveys conducted by employer or service provider themselves.
9. Choose the right timing to send your survey
Our studies found the highest survey open and click-through rates occur on Tuesday, Wednesday and Sunday.
Since there was no significant difference between the response quality gathered on weekdays or weekends, send out surveys first thing during a new week or on the weekend.
10. Reward the respondents
Entice customers to take your survey. Our research shows that incentives can increase survey response rates by up to 30 percent.
If possible, use discounts for your products or services. Alternatively, use easy to distribute and traceable electronic gift cards. Cash electronic cards from amazon.com, for example, will be relevant and enticing for a wide variety of survey respondents.
There is an opinion that freebies can reduce the quality of responses, but our studies show that this isn’t likely to be the case.
11. Use analytical tools and approaches for advance analysis of survey results
Research is about more than just getting answers. It’s about gaining confidence. We believe advanced analytics must be part of every company’s business intelligence strategy, regardless of its size.
If possible, consider using advanced analytical approaches such as correlation, factor and conjoint analyses, quadrant analysis, etc. to arrive at the type of conclusions that’ll drive more precise, meaningful results and, ultimately, better business decisions.
For more information about advanced analytics: http://www.macorr.com/marketing-analytics.htm
The MaCorr Team