How to Properly Do a Quality Survey

There are a lot of reasons for wanting to conduct a survey. It’s one of the best ways to understand public opinion, it’s quantitative, and it’s simple. Or at least that’s what people think. The first two are absolutely true, but the third is exactly what gets people in over their heads. Surveys can be simple, but they can also be incredibly complex. Here are some simple steps to ensure that a survey is successful.

The first step is having a very clear objective and plan. There should be a reason for the quality survey being conducted, not just a random assortment of questions. Once a reason is established a few key questions have to be answered. They are as follows. One, who is going to be taking the survey? Two, where are they going to take the survey? Three, what is going to be asked on the survey? And four, when is the survey going to be taken?

It’s only when these questions are answered that one has a strong foundation to build from. A strong plan makes the physical process of conducting the survey much easier. It also makes pivoting to new questions, locations, or times more easy to justify. Now with a plan established, the second step can begin.

The second step is all about the physical creation of the survey. A plan is the outline, but now the questions need to be specifically and consciously created. Even when one knows the types of questions to include, this is challenging. Asking the right sort of questions is a big issue a lot of surveys face. There are countless easy mistakes to make. For example, asking leading questions. 

A leading question influences the participant to answer a certain way. This can be as simple as the question “How much do you enjoy using this product”? The question is simple, but it assumes that the product was enjoyed at all. A more neutral question would simply ask the participant to rate the product. Other examples of common mistakes are loaded questions and questions requiring more than one answer.

Now if all mistakes are avoided in the second step, the third step is simply execution. There’s some wiggle room here depending on how serious a survey is. Surveys large in scope should consider testing the survey and seeing how it works in action. This helps to ensure the questions make sense and that things run smoothly.

Although for most execution will simply mean opening the survey and doing some early analysis. The way data is collected is important in conducting a survey, so each survey should tailor collection to its needs. And that leads us into the fourth and final step of survey creation, analysis and reporting.

There are two components to this step, each equally important. Analysis is where everything comes together within the survey. Now that the data is collected it is vital that it is interpreted correctly. To do this knowledge of statistics and general analysis is strongly recommended. For anyone lacking these things, there are companies aimed at helping explicitly with analysis. 

It’s important that every step in analysis is handled with care. Anything from exporting files, to doing the right tests, to reporting the right numbers is essential. Speaking of reporting, the second half of this step is where the survey becomes useful. Even if everything else went well, from planning to analysis, it’ll be useless with poor reporting.

Some easy tricks to ensure reporting is effective is keeping things simple, and reporting visually. There’s nothing more understandable than an effective chart. Ensuring that the people seeing the survey results understand it is essential to change. Of course not all surveys are aimed at change, but even for general informational surveys, make it accessible. 

Following all four of these steps there should be no problem creating a useful and effective survey. There will always be unforeseen issues and mistakes, but these steps help to minimize those. And for anyone who feels these are overwhelming, use a service which helps the process along. Companies like IncQuery will code the questions and handle analysis. This gives the creator creative control but outsources the more practical elements.

If even that sounds like too much responsibility, there are companies which will create the entire survey. For those all that’s required is some input, then they’ll create a finished product. Of course that doesn’t mean changes cannot be requested, it’s just a much slower moving and indirect process.

Regardless, for anyone creating a survey, even to those outsourcing, remember these steps. Surveys are a great tool for insight into any specific community, company, or population. They truly can be easy, but that doesn’t mean they’re effortless. Making sure they’re handled properly and are doing what’s intended is just a necessary part of their creation.

See the Full Infographic.