Surveys are not a new idea. We’ve seen many different survey formats and many unique mediums. The first surveys, as we know them, started showing up as long ago as the early 1900’s. By 1930, the US government was using surveys to document economic and social conditions in the country.
There’s hardly anything you can do these days that doesn’t end with getting a survey of some kind. Whether you interact with support staff, make an online purchase, or any number of other things, companies are interested in hearing your opinion.
As surveys have become more common-place, potential respondents have become more and more difficult to entice into responding. The National Science Foundation (NSF) theorizes that the sheer quantity of requests, and growing distrust in those sending surveys are likely culprits for the decline.
So, you may be wondering how can your survey be successful? The answer is thoughtful survey-design. Survey design is about making deliberate choices to increase the success of your survey.
In this article, we give some tips on how to best design your survey to fit your audience and, hopefully, make your survey more successful.
The basics of choosing a survey format
Getting the exact right survey format for your audience will depend on a number of things. First, who are you sending the survey to? What’s the content of the survey? Those and many other questions will guide you toward some more specific choices when designing your survey.
That said, there are some fundamentals of survey design to be aware of. In this section, we cover three of those: question types, survey length, and when to send your survey.
Questions types
Multiple choice, dropdown, short-text, long-text; these are all different question types you may use in a survey. All have their place and different uses, but all fall into one of two categories: open-ended or close-ended.
Open-ended: Open-ended questions are any questions where the respondent isn’t given options to choose from. In most cases, there will be a prompt of some sort, and maybe a couple of other details, but the respondent has free-reign over their answer.
If you want to learn about a respondent’s feelings on a subject, their attitude toward something, or their understanding of a certain topic, an open-ended question may be the best option.
Eg: Where do you see yourself in five years?
Close-ended: Close-ended questions may take the form of a multiple-choice, yes/no, or opinion scale question but they all have the same thing in common; the respondent is choosing from a list you presented them. They don’t have agency over the options in the list. As mentioned above, yes/no questions and multiple-choice questions are very common examples of close-ended questions.
In cases where you need quantifiable data, you’ll want to use a close-ended question. They’re also useful for gathering demographic information.
Eg: How likely are you to repurchase The Original Company toilet paper?
Not very likely
Somewhat likely
Very likely
Pew Research suggests doing a pre-test with open-ended questions to see how your audience interprets a question. It can also be a means to gather common responses to a certain question that informs how you turn it into a close-ended question later on. Close-ended questions can result in cleaner data that is easier to analyze.
Survey length
When considering survey length there are really two things you’re concerned with: time to complete the survey and number of questions. For surveys that you’re sending to customers, you want the time to complete the survey to be under 7-8 minutes. Also, research has found that exceeding 12 questions total in a survey caused a 17% decrease in response rate.
When to send
We all know the cliche, “timing is everything.” With surveys that cliche rings true, kind of. There isn’t one specific time that’s best to send a survey, but more so a couple of times that seem to perform better than others.
Though it may seem counterintuitive, you should consider sending out your survey after or before work hours. Research found that people are less likely to respond to survey during work hours. In some ways, that makes sense. So, try sending at 5am or 8pm so it’s what they see first-thing in the morning. You might be surprised by the result.
One other thing to consider is whether you’re sending an internal survey or external survey. For internal surveys, Mondays are the best day performing about 13% better than other days of the week. The same study looked at customer surveys and also found Monday to be the best day, with Friday performing the worst.
Survey channels
We all know the phrase, “it’s not what you said, it’s how you said it.” That phrase is really a reminder to pay attention to the details when you’re communicating. So, even if you have all the right questions but don’t distribute your survey in the right place you might miss your audience.
Realistically, the best survey format for a particular audience is to stick to the channel they use to interact with you. Whether they subscribe to your e-newsletter, follow you on social media, or read your blog, the place they already are will be the most natural channel to contact them.
With that in mind, we’re going to cover a few popular survey channels to consider:
Email is a very popular channel for surveys. They’re relatively unobtrusive for respondents and are easy to distribute widely, which means the cost is quite low. The average response rate is hard to pinpoint exactly, but it seems to be around 26%.
If you’re looking to cast a wide-net for responses, email may be a good option because of its lower cost. You do need to be careful to not send email surveys too frequently, too. Since it’s a more direct and personal channel, it’s easier for the requests to start feeling spammy.
Text (SMS)
Surveys sent through text messaging aren’t incredibly common. The main reason for that is there are a number of additional barriers to sending a survey through text, such as gaining consent from a potential respondent, along with their phone number.
Along with the legal barriers, there’s also a character limit to SMS surveys (160 characters). So, you’d, potentially, be very limited in what questions you can ask. Also, research done by Gallup found SMS surveys had lower response rates than phone surveys.
Additional research by Pew Research Center found that sending a notification to complete a survey by both phone and text resulted in a larger initial response for surveys, but didn’t increase the total number of responses in a significant way.
If you have a very short survey and previous consent to contact through a mobile number, you may consider using SMS as your delivery method. Otherwise, there will probably be better options.
Phone
Even though more and more people have mobile phones - over 60% of the world’s population, in fact - phone surveys have been in decline. The response rates were relatively stable for about a decade, but have seen a decline in recent years - 6% was the average in 2018.
Along with lower response rates, phone surveys are also quite costly. One study estimates that phone surveys cost about $45 per response. Other research also found that phone respondents are more likely to give positive answers to engagement and attitudinal questions when compared to results from web respondents.
All that said, doing a phone survey does allow certain flexibility not offered by other survey methods. The best example being the ability for an interviewer to ask for more detail about a certain response. So, if you’re after qualitative data and have a larger budget, phone may be a good option.
Web
Web-survey is a catchall term for any survey that lives on a page. It could either be embedded, or function as a pop-up, but you’re not sending it directly to anyone. With that being the case getting statistics for these types of surveys is quite difficult.
The other difficulty is that you’re very limited in targeting an audience. Realistically, these would best be used for contact forms. If you’re looking to gather more meaningful data, it will probably be best to use one of the other channels mentioned above.
Baseline metric
There are many different types of survey formats that exist and each has its own nuances. Some are better at measuring customer sentiment. Others are more useful for collecting quantitative data. With that being the case, you need to be deliberate when choosing. In this section, we cover three common survey formats and the audiences they may work well for.
CSAT
Customer Satisfaction Score, or CSAT, is a very common survey type. Its goal is to measure how satisfied a customer is with a specific interaction with your business. They’re usually sent very soon after the interaction so it’s still fresh in the customers’ mind.
Most of the time you’d ask the customer to rate the interaction on a scale of 1-5 or 1-10 and then leave an optional comment section for the customer to elaborate further on their experience if they so choose.
Pro Tip: When writing your CSAT question make sure you’re abundantly clear that what the customer is rating is the interaction they had with the agent, not the company as a whole or how satisfied they were with the answer given. Being clear will help you get the most useful data.
NPS
NPS stands for Net Promoter Score. These surveys are very simple and only have two questions. The first question asks on a scale of 0-10 how likely the respondent is to recommend your company to others. The second question is a text question about why they gave the score they did.
It’s very simple and that simplicity has made these types of surveys very popular. If you’re wanting to get a general read of customer sentiment, NPS may be the way to go. Since it’s not tied to a specific interaction, you can send them at any point.
Pro Tip: Do be cautious that you’re sending an NPS too soon, though. If they haven’t been a customer for long or haven’t had much experience with your brand they may not be the right audience. Longer-term customers may be able to provide deeper insights.
CES
CES stands for Customer Effort Score. Its aim is to find out how easy, or difficult, a given task was for a customer. These, like CSAT, are generally tied to a specific event. However, it’s different in the sense that they may not be rating an agent, but more the entire experience.
For example, they could’ve had a great interaction with an agent, but found it very difficult to find the contact information in the first place. That would most likely have the result of a low CES score. It takes the whole experience into account, not just one isolated part.
Pro Tip: CES works best when it’s a consistent metric you track over time. Since it’s more about overall experience improvement and not one isolated case, there will always be the possibility to gain insight.
Choose the best survey format to improve responses
There’s a lot to consider when making a survey. First, you should start with the basics. Make sure your questions are clear, the form isn’t too long, and that you’re sending it out at an optimal time for your audience.
Next, you should consider the channel you’re using to send the form. Email, SMS, phone, and web surveys are different options to consider. Remember, all have their pros and cons, so be sure you consider all the factors before settling on your choice.
Last, know what metric you’re measuring. CSAT, CES, and NPS are very popular survey formats. You may end up having a combination of a few, they’re a great starting point for any project. Building a survey may seem daunting, but we know you’ve got this.
Learn how GetFeedback can help you exceed customers’ expectations—start your free trial today.