8 questions to resolve before you undertake a survey research project

Professional research organizations often invest tens of thousands of dollars in their survey apparatus, maintaining expensive assets like research panels (groups recruited to answer multiple surveys) and teams of analytical professionals. But you don’t have to go big to create something useful.

Anybody with a SurveyMonkey subscription can conduct a research survey quickly and cheaply these days. However, if you don’t prepare properly, your results will be crap.

The key to creating and publicizing a useful survey without breaking the bank is planning.

Ask yourself these 8 questions first

Here’s what you’re going to have to figure out in advance if your results are going to be useful.

1. What’s your research question? A well-designed survey attempts to discover patterns that were previously unknown. That’s primary research. Your survey design should be set up to test a specific set of hypotheses, not to just poke around and collect random information.
2. Who’s paying for it? Good research and writing up the resulting conclusions demand the work of data professionals, and they don’t come cheap. If you want a quality job, expect to invest at least $8,000.
3. Where is the sample coming from? A well-designed survey should be aimed at a specific target group, such as freelance workers, diabetes patients, or middle managers at financial companies. You need a strategy to reach your target group. Working with associations, media, or events that attract people with a specific profile is a great place to start.
4. How will you motivate them? You’ll need a way to reach your target group, such as with online ads or email. And you’ll have to provide incentives, such as the chance to win prizes. Often, it takes multiple rounds of prodding to accumulate enough responses to make the results significant.
5. What’s the schedule? A typical project takes at least three months to plan, design surveys, field the surveys, analyze the results, and write up conclusions. Don’t start imagining research in February and expect to publish it in March.
6. Who’s doing the analysis? Tools like SurveyMonkey include some analytical capabilities; more sophisticated users use software like SPSS. But the challenge is getting the analytical talent lined up. You’ll need someone with a background, not just in statistics, but in analyzing survey data. Such analysis tends to take several weeks to complete.
7. How will you write up the results? You can compile your conclusions simply in a PowerPoint presentation, or write them up more attractively in a well-designed document (often known as a white paper). Either way, you need someone to assemble the writeup and if there’s a final report, design it to make it attractive and effective.
8. How will you publicize the results? Some projects are designed for use within an organization; if so, you can just email or present the results to the internal team that needs to see them. But if your objective was to generate interest or publicity, you’ll need a PR team to promote the results on social media and to attract media coverage.

Research is fun

When I joined a major research organization, I was delighted to find that my statistics background would come in handy. We created a set of surveys called Technographics in 1996, and it’s still running.

Since then, the tools to find samples, conduct surveys online, and analyze the results have become accessible to nearly everyone. I was delighted to find that I could complete a whole survey on my own. Both of my recent books included data from surveys I conducted myself.

The thrill of watching people answer surveys and discovering new information about the world can’t be beat. I love doing that, and will always be up for more.

It’s not always simple or easy. But it’s always rewarding.