BLOG

on Conversion Rate Optimization

Structuring your Customer Surveys: Asking Questions for Actionable Answers


Customer on-page surveys are a great way to gather feedback for your business. These surveys help you in understanding your customers and address issues that they face.

In a perfect world, surveys would work magically. We would create and deploy an on-page survey and the answers would just start trickling in. But, in the real world, it’s quite the opposite.

Why?

The average customer attention span is getting shorter as you speak. If you don’t ask the right questions with the help of the right platform, you may never get an answer.

That’s why thoughtful survey designs and questions are a must. These help you get better and actionable results, as well as a good ROI.

Here is a list of things that makes on-site surveys the de facto tool to gather actionable feedback:

  • Uncover flaws, which could affect various aspects of your business-like conversions, retention rate, churn rate, and so on, in web flow or user experience.
  • Evaluate your users and segment these for enhanced targeting.
  • Identify and evaluate the needs of your customers. For example, identify if it is time to get serious about a feature your customers are longing for.
  • Understand how your customers find your offerings to be different from those of your competitors’.
  • Evaluate new features or versions by asking for user feedback.

Hence, using app-based or on-page surveys is a great way to let the users know you genuinely care about what they think.

So, without much further ado, let’s begin.

Surveys: How to Begin

You may consider planning a survey to be straightforward. Start with building a list of questions individually or with a team, and brainstorm to pick the right set from the lot. When done, use a survey app and deploy it.

Consider the points below:

  • Why are you doing an on-page survey, and who is your target audience?
  • What channels and mediums are you going to target for this survey?
  • What do you expect to figure out?
  • What should you do to maximize data collection from this survey?
  • What actions are you going to take based on the survey results, which might be either positive or negative?

This process seems intuitive and productive. Yet, if you just turn the tables, you’ll be able to frame better questions.

Instead, start by brainstorming the answers that you want. When you point out the answers that you are looking for, the questions will automatically come to you.

Now, let’s revamp the above points into answers and see how these might help us come up with the questions:

  • You are doing a survey to increase the upsell rate. So your target audience would be your current set of users who are regularly using your product or service.
  • You want to target mobile users and get their feedback. Hence, the channel you pick is mobile and the medium would be push notifications, in-app surveys, and maybe
    email-based surveys.
  • You would be able to figure out the segment of users to whom you can upsell.
  • Incentivize the program by offering a free trial or a free beta period to maximize data collection for your survey.
  • Depending on the feedback you gathered, assign a team of account managers to the users who are open to upgrading. And, let’s say, in an unfortunate turn of events, your survey results in a backlash, it’s time to fix the system.

Now you know what your survey expectations are and the answers you are looking for. It’s time to create questions based on the above information.

Types of Surveys

As you frame questions based on the expected answers, you also need to think about the structure of the feedback.

You can gather answers to your survey questions in a variety of ways. The type of survey you select will affect your campaign results. Here are the most common survey types that you can choose from.

1. Yes/No surveys

This is the simplest form of an in-app or on-page survey that is running on various websites, as we speak. A question is followed by only 2 options, Yes and No.

 

Yes/No Survey

 

2. Multiple option-based surveys

Need more details than just Yes or No? Then multiple option-based surveys are exactly what you should try. This survey-type allows a marketer to provide multiple choices as possible answers. You can allow participants to select as many options (checkbox-based survey) as they deem fit or just one (radio button-based survey).

 

Multiple Option Surveys

 3. Drop-down menu-based surveys

Drop-down menu-based surveys work the same way as multiple option-based surveys. The catch is that at most times, drop-down options are sorted in order.

 

 

Drop Down Survey

Example:

What is the traffic of your website per month?

  • 0–10,000 users
  • 10,001–20,000 users
  • 20,001–50,000 users
  • 50,001 or more

Note that the options are in ascending order.

Drop-down surveys also help you run surveys where the responses could be from a vast set of data. For example, while trying to figure out the organizational function of the customer, there can be more than 10 options. Hence, a drop-down option would accommodate these options, without breaking your web design.

4. Textbox-based surveys

At times, you just need direct feedback from your customers or some contact information from the surveyee. In these circumstances, textual surveys come to our aid. These textbox-based surveys can have a small or large box.

 

Textbox Survey

When Should you Time The Survey?

As you can target a definite set of users, note the following points before deploying your survey:

  1. Does the user qualify for the survey? (Does the user fit your sales funnel, is the customer a valuable lead, and so on?)
  2. When should you ask the question? For instance, you should not ask a user at an early stage of the funnel, “what went wrong with your order?”
User Targeting
Target visitors based on their activity. For example, here the survey will only activate for returning visitors after 30s have elapsed.

Analyze your website dynamics and metrics. Also, watch your customer’s web flow and use the gathered data to pinpoint the right stage when your survey can connect with the reader.

How to Get People to Respond?

Asking the right questions is one part of the problem. The other important part is to get responses from your audience.

To do so, you need to frame questions that are easy to understand and easy to answer as well.

Here are a few ways marketers approach surveys:

  • Ask simple questions, concentrate on questions with well-defined options, a simple Y/N question or a multiple-choice survey. When answered, ask for an explanation.
  • Ask generic questions, which can have multiple answers.

We conducted a study on both types, and you can see how option-based questions performed better than open-ended questions.

 

VWO Survey Report -2016
The survey that had options resulted in 493 replies.

 

VWO Survey Report - 2
The question that had options to select from. One can clearly see that we got a massive 493 reply here.

 

Source

We learned that questions with concrete options performed better than questions that are open-ended. People tend to respond to surveys that require less time to complete. The former survey had answers to choose from and hence users were at ease while answering the survey.

We saw that surveys with predefined answers had a better scope of being answered. Moreover, our customers preferred to follow up with an explanation for their answers.

Now comes the part with best practices—what all you should keep in mind when you want to run surveys.

Do’s of Web Surveys

1. Ask questions based on a scale

 A simple way in which you can extract a rich dataset from your survey is by introducing a scale for the answers. Consider changing your Yes/No questions to more figurative questions.

For example, consider this question:

To what extent are you frustrated with the traffic on the 31st interstate?

2. Avoid asking leading questions

Some words and phrases interjected in your survey questions can subtly encourage readers to answer in a certain manner, thus compromising the purpose of the question. For example, “Due to a high cost of web hosting, do you think we should stop refreshing data every 2 hours?” is a question that may bias readers to choose an answer. Instead, you can frame your question on these lines, “Should we stop refreshing data every 2 hours?”

To evaluate if your question is leading readers, run a test with your colleagues or friends and see if they are biasing toward a similar option. If the percentage is high, then you need to reassess that question.

3. Offer incentives for long surveys

If, for some reason, you are unable to keep your survey short, offer an incentive to complete the survey. The incentive does not need to be something huge. For example, you can offer an entry to a giveaway contest for a product when a user participates and completes a long survey.

4. Keep a survey short and simple

If your surveys are brief, there is a greater chance of your visitors or customers completing the survey. Target a completion period of 10–15 seconds for best response rate.

Use of checkbox-based or radio-button-based surveys is recommended over textual surveys. Both these survey types have the highest response rate in our market study.

Also, do not delve into framing questions that are unclear to your readers. Be specific. For example, instead of asking, “Do you commute regularly using a cab?”, you may ask, “How many days in a work-week do you take a cab to work?” Such questions get you short and precise answers.

5. Make sure that every survey question is answered

 You are running a survey to gather actionable insights for your business. Weigh-in on every question, analyze what answers you are looking for, and decide if the question has to go to the survey.

6. Avoid questions with “and” and “or”

If your questions contain these connecting words, it may confuse the people taking the survey.

 For example, “Do you find your MacBook reliable and fast?”

Here, the customer can get confused, as ways to assess reliability and speed depend on individual requirement and vary across different users. Such a question would discourage the user from giving a concrete answer.

Instead, break the question into 2 parts. First, check the reliability and later, check the performance speed of the MacBook.

  • Extremely frustrated
  • Moderately frustrated
  • Slightly frustrated
  • Not affected

As you can note, this question could have an alternative yes/no variation; but this version provides you with a better context on the frustration level of the visitors. Hence, you can make better decisions on how to proceed toward a solution.

7. Place your Surveys at the right position on your website

According to our report, we noticed that a higher number of people completed on-page surveys when they were on the right side of the page, with a higher contrast ratio.

Don’ts of Web Surveys

1. Avoid open-ended questions

Per the above discussion, it is clear that open-ended questions in a survey have a higher chance of getting ignored than being answered. Always delve into option-based surveys, as discussed earlier.

2. Avoid a survey bias

Survey bias is a phenomenon where the person or the team putting the survey together develops questions in a specific manner that leads to the participant answering as desired by the surveyor.

For example, to confirm a fact about your product, you may frame a question in a manner where your customer answers precisely what you want.

Here is a hypothetical situation, which demonstrates the intent of bias toward a specific answer. If you ask your customers, “Due to the increase in accidents on campus, do you think motor-based vehicles should be banned in the college,” the answer is going to be mostly “Yes.”

Why?

Because surveyee attention may shift toward the issue of higher accidents in the campus and they may answer the question keeping safety in their mind.

But when you frame a question like “Do you think motor-based vehicles should be banned from colleges?” students may have a more neutral answer or outlook.

3. Not testing your survey

Don’t ever forget to test your Survey. Here are two ways of testing your survey.

The first test would be testing its compatibility with the on-page survey. Things like device compatibility, responsive nature, and survey triggers should be tested before going live with the survey.

The second test would be doing an internal run of the survey. Try to at least run the survey across 5–10 members of your organization and see if you can get unbiased and honest responses. If something looks wrong, revisit the questions.

Conclusion

On-page or On-site surveys are a crucial part of conversion optimization. Whether you are looking to decipher your low conversion rates or improve your already decent figures, on-page surveys can play a significant role.

Surveys are going to help you in a lot of ways. A survey allows you to understand your visitors and existing customers, and learn about their expectations from your product or service. It’s also a tool that allows you to segment and target your user base. Visitors with advanced knowledge even use surveys to garner their NPS.

This post is written to help you conduct surveys that bring you actionable data. So the next time you begin your survey:

  1. Understand the need for a survey. The answers you are looking for will help you create your questions.
  2. When you have the questions, decide which format of on-page surveys suits the feedback mechanism.
  3. Learn to earn a response from your customers or visitors. Experiment with best practices for surveys to maximize the response rate.
  4. Stay away from open-ended questions. Avoid a survey bias, and always test before you go live.
  5. After you have actionable results, start delivering to your users.

 

Leave Comment

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes : <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

Contact Us / Login

Product
Resources Home