How to Use Surveys to Create Better Content
Engagement is the secret sauce of digital marketing; every marketer wants more of it. Whether it’s on social media, on the company website, in email messages or anywhere else, we all strive to get more engagement for our content.
There are a lot of ways to get traction for your content through external forces. Timing is one; screaming link-bait headlines is another. Outrageous images can work. You can also try content promotion, influencer marketing, and making your content easier to share.
But the very best way to boost engagement is to give your audience exactly the kind of content they wanted in the first place.
How to create content your audience loves
There are a number of tried-and-true techniques to create great content. They include:
- Using analytics to determine what your audience’s behavior reveals about their favorite content
- Doing competitive analysis to see which pieces of your competitors’ content have captured the most engagement
- Being a member of your audience, then serving up content you yourself would like
- Asking your audience what content they want
- Asking your audience what their aspirations, fears and challenges are, then creating content that addresses those issues
Faster horses or an automobile?
Let’s focus on the last two options: ask people what they think they want and ask people what motivates them. You can absolutely ask people what specific content they want more of. But you risk missing the opportunity to create content that is better than what your audience can imagine. As Henry Ford put it, ““If I had asked people what they wanted, they would have said faster horses.”
Creating the content version of the new automobile sounds exciting. It is also risky. Fortunately, most of us don’t need to aim for the stars and create content that borders on art. We just want our content to work to support our business goals.
Whether you ask people exactly what kind of content they want or what their motivations are is up to you. Whether one or the other is better depends on how evolved your content marketing is, and what you want to do with the information you gather. But the two approaches are very different. Once you see that difference, you’ll view surveys differently, too.
You might be asking yourself, do you really need a survey if you’ve got analytics set up? Isn’t what people say different than what they do? It’s true – people say one thing and do another. And that’s okay. That doesn’t mean surveys can’t be extremely useful.
This issue actually brings up an excellent way to filter out potential survey questions: Don’t ask your subscribers anything your analytics data can answer.
Here are a few sample questions a survey could answer that analytics data can’t:
- Would you like more beginner, intermediate, or advanced content?
- Is our content easy to understand?
- What topics would you like us to cover that we haven’t?
- Are you aware we offer the following content (or services)?
- How important are these issues to you (please rank in order of importance)?
The survey as a content marketing tool
Though we are specifically going to focus on using emailed surveys to learn what your email subscribers want, this is not the only way to ask people what content they crave. You could also:
- Use a survey on your website
- Use a poll on your site
- Run a Facebook contest that gets people to give you their most pressing question about your topic
Of all these methods, though, I like emailed surveys best. Here’s why:
- Most email service providers (Act-On included) offer a way to create and send surveys to your subscribers easily.
- Emailed surveys get results fast. You’ll get most of your results within 72 hours of sending the survey.
- If you’ve been good to your list, your subscribers are already trained to open your emails. That means you’ll get a drastically higher response than if, say, you sent a survey to your Facebook.
- You’ll be surveying people who are your core audience. These are your subscribers, buyers, clients – they’re into your stuff. Better to find out what they want than to try to figure out what a general audience wants, or even what somebody else’s audience wants. And if you want to segment out a portion of your email list to send the survey to, that’s easy.
How many subscribers do you need to run a survey?
While emailed surveys are great, they do have one stringent requirement: You gotta have subscribers. Quite a few email subscribers. Unfortunately, while you can run a survey with only 1,000 email subscribers, you’d probably only get about 15 responses. That’s not enough to get statistically relevant answers. It’s not even enough answers to guess at a trend.
Hopefully, you’ve got more than 1,000 subscribers. Let’s say you’ve got 3,000. Let’s also say your clickthrough rate is around 3%. That means you’ll get about 90 people to click through to your survey. If half of them finish the survey, you’ll have about 45 completed surveys.
Is that enough to get decent results? Yes. But it could definitely be better. Ideally, you want at least 200 completed surveys to see clear trends and have statistically valid results. And more completed surveys would be even better.
With only 45 responses, you’ll need to see big differences in the survey answers to draw any conclusions. But 45 is still enough to make the exercise worthwhile. For more details on how many subscribers you’ll need, see Survey Monkey’s sample size calculator and their explanation of how to calculate sample sizes.
What to do if your survey sample size is low
Need to boost the number of finished surveys you get? Here are some excellent ways to improve survey conversions (which means the number of completed surveys). These tips will help even if you’ve got a huge list.
- Keep the survey invite email super-short. Aim for just a few lines of text, a reading level of about 5th grade, and fewer than 300 words. Something like this would be good:
- Keep your survey questions to a minimum. I like surveys with no more than five questions, max. But conversion expert Peep Laja says it’s okay to go longer: “My experience is that the sweet spot is around seven to 10 questions. More than 10, and the number of people who take the survey goes down; less than seven, and you might not capture as much information as you could.”
Either way, consider each survey question as carefully as if it was a wish granted by a genie. Here’s a super-short survey that still manages to get highly actionable information:
- Give survey respondents a gift. A really nice gift. Maybe that’s a discount or a copy of the finished survey. Whatever it is, make it good. You can also add a scarcity factor to your gift, like Mint did with this survey invite. Maybe the scarcity factor let them get away with offering only a $10 gift.
- Tell people how long the survey should take to complete. This can improve your survey’s conversion rates a lot. Don’t expect many people to give more than five minutes of their time.
You can also skip the time measurement and just tell them how many questions the survey has, like this:
- Make it clear you will protect your respondents’ privacy.
- Resend your survey invite to people who didn’t click through to the survey the first time. King Arthur Flour sent me this handsome survey invite on March 25. I didn’t click through, so they sent it again on April 2:
Two types of survey questions to avoid
1. “Do you plan on buying our product in the future?”
Why it’s bad: It’s never a good idea to ask people what they’re going to do. Things change, and you’re wasting a survey question on an issue that behavioral data is better suited to answer.
Also avoid asking people about whether or why they’re going to buy. This immediately punches a hole in your premise that you’re trying to help them. If you’re asking whether they’ll buy or not, you’re obviously asking this for your own benefit.
2. “Do you like our blog/newsletter/videos?” or “How could we improve?”
Avoid open questions. They often give no actionable data. It’s also really easy to take one answer from a question like this, from someone with a strong opinion, and let it completely skew your survey results.
Notice how the survey examples above structure the answers to their questions. This can make a huge difference in how well you can analyze your data after the survey has been run. Generally, it’s good to use radio buttons, but in some cases you’ll want to give people the option of making more than one selection. Text box answers can be interesting, but they’re also really difficult to analyze. The whole point of a survey is to have actionable data, so it’s critical to make sure your questions and answers all serve that goal.
Back to you
What’s your experience with surveys? Have you run any for your content marketing? Did they help? Tell us about it in the comments.