1 (888) 505-5689
Surveys are one of the most powerful content types available. If your brand can find fresh, new insights that peek into an industry or strategy, then it can go viral as people share the information and link back to it. While the data alone is powerful, it can easily be promoted through whitepapers, infographics, blog articles, and webinars. When done well, the data will be used and shared for months — even years — after it’s published.
While surveys seem easy to create at first, they’re actually hard to do well. There are plenty of surveys floating around the Internet that were created in a rush, are full of bias, and are only sent to people who will give the right answers. This corrupts the data and makes readers question it, along with the integrity of your brand.
If you’re eager to start creating content around survey results, keep reading. This guide will help you develop a survey that provides useful information that your brand can stand by.
Prevent Selection Bias From Clouding Your Results
Image via Flickr by lordsutch
The first obstacle that most marketers experience in survey creation is finding a large group of people to sample. CEOs might turn to their peer group or client list to get responses or quotes, making it seem like their opinions reflect the entire industry. While their opinions are certainly valid, they might not reflect the general consensus of the entire field.
The main type of selection bias that most surveys suffer from is undercoverage. This is when some groups of people are sampled more than others. For example, a survey sent out to Donald Trump’s donors or email list might find that the President has a 90 percent approval rating. However, the same survey sent to Democrats would find that he has a 10 percent approval rating. Neither is correct (no matter how badly each side wants to believe it) because the whole population wasn’t surveyed.
Closer to home, a survey at a content marketing conference might report that 85 percent of marketers believe content marketing is the future. Well, yes, otherwise they probably wouldn’t have paid to attend the conference. A truly unbiased survey would reach out to people to learn who doesn’t find content marketing to be effective and why.
Statisticians promote random sampling to prevent bias in survey results. In your organization, consider reaching out to third-party survey companies to sample data for you. They will move past your peer group and get answers from a diverse group of people who have unbiased opinions about your brand.
Make Sure You Account for Measurement Errors
Along with making sure the group of people who respond to your survey are unbiased, responsible content marketers also account for sampling errors.
Election polls can also be used to explain how sampling errors occur. If a pollster only samples a small percentage of the population, the statistics have high error levels. When news outlets report on polls, they often highlight the sampling error including in the results. They might report that one candidate is in the lead 52-48 percent, but if the sampling error is greater than four points, the other candidate very well could have the advantage.
Measurement errors are often found in surveys that segment results by demographic. For example, a survey might ask 500 people about their position in the company, and then ask executives specific questions. If only four people who take the survey are executives, then the sample size is too small to pull verifiable results from. It may be tempting to say that 75 percent of executives agree on X, but that really only means three out of four people you asked liked the idea.
In your surveys, pay attention to sample sizes. If your responses are narrowed to the point where only a few people are answering certain questions then you need to either throw those questions out or survey more people.
Even if their responses agree with your theories, responsible content marketers strive to report only the best results.
Proofread the Questionnaire for Clarity and Leading Questions
Once you have a clear idea of who you are going to survey, the next steps involve creating an unbiased and clear questionnaire. Clarity is one of the biggest issues when creating a survey, especially for complicated industry questions. While a lack of clarity might seem like a mistake on the part of the writer, it can also be used to manipulate answers based on the results.
Some survey developers specifically confuse readers (either with big words, convoluted phrases, or questions out of context) to get them to respond a certain way. This allows them to simplify the questions later and make it seem like audiences agree with them.
Along with clarity, many survey creators are guilty of creating leading questions. These questions encourage people to answer a certain way to generate favorable results. The questions below are great examples of leading questions:
All three of these questions are meant to make audiences react emotionally to choose the right response. According to the question, if you don’t support metal detectors, then you must not love your child.
Ask people to test the survey before you send it out and offer advice on clarity and leading questions. Leading questions are hard to remove. You don’t want the survey to produce results you disagree with. However, by removing them and remaining objective in your survey process, you can create a piece of content that is reliable and valuable.
Make Sure Your Respondents Can Provide the Best Answers
Even the best-made surveys can go awry if audiences can’t choose the best answers to the questions. People will respond to your survey in different ways, and you need to provide outlets for them to do so. For example, a few ways you can make your survey more honest include:
These steps make it easier for your respondents to answer fairly. By allowing them to choose multiple options, they don’t have to stress about which choice they lean toward. Furthermore, allowing people to skip or not respond means you won’t have clouded results based on people who don’t have a real answer to the question.
Panera actually has a strong customer response survey that takes these steps. It asks people about the food, drinks, and atmosphere. If a guest bought hot food, but not a bakery item, the survey doesn’t ask them about the selection and quality of bakery food. This keeps people engaged in the survey while ensuring the results are as clean as possible.
As you test your survey, ask people to elaborate on why they chose certain answers and if there are other choices that should be added. You can even A/B test your survey between a handful of people to see if one set generates better (higher-quality) results.
Include Qualitative Questions to Provide Context
One of the biggest mistakes that survey creators make is generating assumptions based on the data. They create a set of questions and, no matter how unbiased they are, generate biased conclusions from the information. For example:
If the majority of people answer no, then the survey creator could report that the majority of people are unsatisfied with airport security and the TSA needs to take measure to make searches more thorough. In reality, travelers might be unsatisfied because they feel like TSA employees should be trained better, the process could be easier, or they want to keep their shoes on.
While this question, in particular, could be solved with a follow-up question, it could also be clarified with qualitative, open-ended responses. By letting people explain their answers, marketers can understand why they respond a certain way in order to prevent bias.
As an added bonus, these responses can be used as quotes in your whitepaper or infographic.
Include optional qualitative answer boxes in your survey. Not everyone will use them, but they can provide important insight and ideas for your analysis. Plus, if you find people are answering a certain way, you can adjust your survey to generate cleaner results.
Let CopyPress Help With Your Next Survey
We have seen a great deal of poorly-created surveys that generated mediocre content marketing results. If your survey is weak or biased, then you will struggle through the whole marketing process. Let us help you. Contact CopyPress today and ask us about our content creation plans. We can develop a whole promotion strategy around your content that generates traffic, links, and leads.