Design smarter surveys for better business decisions

You’ve identified a research need for your business. You know what you want to learn and what you’ll do with information when you get it. Now you’re ready to go out and start collecting survey data – right? 

Not so fast. 

Effective marketing research—including surveys—can provide business leaders with powerful insights for smart decision-making. But a poorly designed survey risks delivering misleading data that can steer decisions off course. Launching a survey without first confirming it’s well designed, could result in wasted time and money as well as missed opportunities that better information could have uncovered. 

Here’s how to design surveys that drive business results. 

Define the research objective 

Before survey questions are even drafted, you need to define the research objective. What specifically are you trying to learn, whom are you trying to learn it from, and how will you measure it?  

For example, if you are planning a customer satisfaction survey, your research objective might look something like this: 

“Conduct a survey among customers who have made a purchase within the past year to learn about their satisfaction with our company’s products and services as measured by customer sentiment and our company’s overall Net Promoter Score.” 

Well-designed surveys capture the information you need to meet your research objective. 

Ask the right people 

Next you’ll need to figure out who and how many people you need to participate in your survey. Sampling – selecting a subset of individuals from a target population – is a key part of survey success. Due to budget, time and logistical constraints, it’s simply not possible to survey everyone. That’s why it’s so important to make sure your sample group is representative of the larger group you’re targeting.   

For an effective sample, you’ll need to clearly define your target population and generate an appropriate sample size.  

• Target population: Think about who would make the most sense to study based on your research goal. Depending on what you’re trying to learn, you may want survey participants to meet certain demographic, geographic, psychographic or behavioral criteria. The more specific you can be about who you want to survey, the more likely the sample group will be representative of the larger group you’re interested in learning about. 

• Sample size: Sample size can significantly impact the cost and reliability of a survey. Survey too many people, and the cost will be too high; survey too few and the reliability will be too low. Sampling shouldn’t be based on a gut feeling or a guess; it’s a science. A market researcher can determine the optimal sample size for a particular survey through statistical calculations.  

Ask the right questions 

If sampling is more science than art, crafting survey questions is a blend of both. Survey questions need to be measurable, precise, aligned with the research objective, and easily understandable.  

As surveys are a type of quantitative research, they aim to gather measurable data. Unlike the open-ended questions used in qualitative interviews, most survey questions should be close-ended to simplify result analysis. 

Common pitfalls in survey question design include: asking the wrong type of question (e.g., asking for a rating through an open-ended text field instead of through a rating scale); difficult to answer questions (e.g., double-barreled question that ask about two topics but only allow for one answer); and leading questions (e.g., questions that prompt the participant to respond in a certain way). One way to spot any potential issues with the survey questions early is to run a pretest where you send the survey to a small group of people and ask for their feedback. 

Consider the following when designing or reviewing a survey draft: 

• Is the survey measuring what it’s intended to measure? (In other words, is it aligned with the research objective?) 

• Are the questions specific, clear, and easy for participants to understand? 

• Is the survey designed in a way that will result in objective feedback and reduce as much bias as possible? 

Begin with the end in mind 

Remember, bad data can be worse than no data, so take steps to confirm that your survey is well designed before launching it.  

Identifying and sticking to your research objective, targeting the right sample size and group of people, and crafting clear and unbiased questions will make your survey results more insightful and more valuable for your business.