10 Golden Rules For Creating Effective Surveys

As a business, it is good practice to check on your customers and employees through surveys. When using questions that discourage participation or an honest response, you won’t enjoy the benefits that a good survey offers.

In addition, if you are having trouble making sense of your survey’s responses. Then it is likely that you are not asking the right questions.

We compiled a list with the top 10 best practices you should follow to create great surveys. A great survey will help you make decisions faster and increase customer engagement and relationships as they feel understood.

1.   Keep it short & simple

Before you begin making your survey, define a clear and achievable goal or outcome for your survey.

 Once you have a target in mind, strive to make your survey short and simple. If your survey is too long, your respondents will give up halfway.

To slim down your survey, it should:

• Only ask goal-oriented questions

• Ask one thing per question

• Use simple and direct language

• Ask a question only once

If you must ask similar sounding questions, spread them apart so it’s not obvious. Besides, if your question doesn’t help achieve your survey’s key objective, remove it.

Longer surveys have a lower completion ratio unless they are incentivised. So a short survey is always better than a long one. No one wants to answer pages upon pages of a survey.

2.   Build upon your questions

Sometimes you may want to ask a big and complex question to your audience. Do not do it! Instead, break it down into several smaller questions and line them up.

Questions that share a similar context should be grouped and follow each other.

For instance:

• Do you drive?

• What vehicle do you drive?

• How often do you service your vehicle in a year?

Instead of:

• Do you drive?

• What is your date of birth?

• Are you married?

• Is the sun blue?

• How often do you service your vehicle in a year?

So, you should build upon your questions and keep the more personal or intimate questions towards the end. Gradually bring your respondent(s) to the right temperature, do not throw them straight into boiling water.

3.   Present limited options

You do not want your audience to be stuck by having too many options to analyse and choose from. You also do not want to ask complex questions that need the survey takers to do some deep thinking.

Your questions should have a limited answer set and therefore be easy to answer quickly.  

It is also recommended to often ask closed-ended questions. Closed-ended questions can be answered with a quick yes or no. For example, “Do you like our new product?” Closed-ended questions are effective early on in a survey.

4.   Filter your respondents

Every survey you make should have a target respondent. Publishing a survey without a target audience is a terrific way to get misleading data.

 Therefore, identify your target and peripheral respondents then devise measures to allow only them to answer your survey. This way, you capture trustworthy data from your ideal or target customer.

 You would not want to conduct a hardware store’s survey in a hair salon nor a luxury brand/car survey in a poverty-stricken neighbourhood. Sure, you may get responses, but they probably will not help your business.

 Also, avoid repeated survey responses from the same individual. You should limit one person to one survey session.

 Measure to filter your respondents include:

• Password protection

• IP or device address restrictions

• Enterprise logged in users only

• Set a cut-off Date & Time

• Email Invitation only restrictions

5.   Ask definite questions.

Always use closed-ended questions where possible. Do not overload the respondent with unnecessary thought or choices when a simple Yes or No would suffice.

 However, overusing closed-ended question can frustrate respondents because their desired answer is not a choice. Also, misunderstandings of questions can go unnoticed in Yes or No response.

 Therefore, you want to make sure you only use the closed-ended question in situations with binary responses (either yes or no).

Examples of closed-ended question are:

• Would you recommend our service?

• Did you enjoy our customer service?

• Did you find what you were looking for? 

6.   Ask neutral questions

Asking leading questions or questions with a biased set of answers defeats the initial purpose of using a survey.

 Therefore, you must only ask neutral questions that do not guide the audience to a specific answer. Your questions should be clear and easy to understand or answer. Avoid using jargon or uncommon words.

 Leading questions like:

• Do you like our excellent new product?

• You are happy with our new product, aren’t you?

Arm-twist your audience into giving certain responses. Instead, rephrase your questions using neutral wording as shown below:

• How would you rate our new product?

• How do you feel about our new product? 

Another thing to avoid is providing a biased set of answers.

For example:

• Good

• Very Good

• Extremely Good

This frustrates customers or employees with negative experiences leading them to leave your survey unanswered.

Ask for help from an uninvolved friend to see if they can spot any biases.

7.   Use a standard rating scale

Many surveys use rating scales with varying scales of judgement. For example, in one survey, you will find a rating scale that goes from 1-5 and another that is 1-3 or even 1-7. Even if this is pretty, do not fall into this trap!

Use a uniform rating scale throughout the survey to avoid confusing or overloading the survey takers. The rating scale should be appropriate and give the audience a sufficiently broad spectrum to express themselves. The scale should be balanced with:

• Positives (incl. extreme positive)

• Negatives (incl. extreme negative)

• Neutral

We recommend a 1-7 rating scale.

However, if you must use different rating scales, list them separately by providing details related to them. Moreover, make sure your questions are specific.

Below is a sample balanced 1-7 rating scale.

1.    Extremely bad

2.    Very bad

3.    Bad

4.    Neither good nor bad

5.    Good

6.    Very good

7.    Extremely good

8.   Incentivise Surveys

People are usually motivated by personal gain. Pen and paper surveys made it hard for people to refuse to answer surveys. However, on the internet, that problem does not exist.

So, if you are struggling to get responses on your surveys. It is time to consider adding benefits to your survey. 

The incentive does not even need to be financial, instead, it could be:

1.    Results that can be achieved with the survey's results

2.    Review of some unpopular item or rule

3.    Charity donations

4.    Sample product or service

5.    Honour Badge or T-shirt

6.    Coupon

7.    Raffle.

8.    Etc.

You can get creative with incentives if it is within your budget. The possible incentives are endless, but it is up to you to come with ones that fit your situation.

However, avoid bias brought by incentives to a survey. Profit motivated answers might change depending on the incentive.

9.   Avoid questions with plural interpretations

Avoid asking double-barrelled questions. Double-barrelled questions are the ones that can be interpreted in more than one way.

At first glance, it seems easy to not ask questions that can be interpreted in different ways. However, the evidence proves otherwise.

For instance, avoid using questions like:

• How would you rate the quality of our product and support?

• How satisfied are you with your salary and working conditions?

• How often and how much do you spend during each drive?

When you are coming up with a question, you already have an answer in your head so it’s easy to miss or have a bias or blind spot as to how other people might interpret the same question.

Thus, it’s good to conduct an internal survey that is not interested in the respondents’ (uninvolved) answers but how they understand the survey’s questions. In this case, you are mainly interested in the cognitive aspect of the survey.  

Let the survey takers rephrase the questions, in their own words, from their own understanding. The results may shock you!

10.   Preview before publishing

In any business, it is recommended to do your due diligence before you present something to an audience. Publishing a survey littered with errors and design fails just hurts your credibility.

Therefore, you must preview your surveys before releasing them to an audience.

Check if:

• All the questions are there

• The designs and branding are where they should be

• The order is correct

• If it works as intended

Basically, you much check if everything is on point.

You could get an extra set of eyes on your survey by asking a friend or colleague to check and answer it for you. If they find any errors or confusing statements, then they notify you.

In Summary

Follow these best practices when making your survey, and your survey performance and results should improve.

Bear in mind that, even with all these rules, you have barely scratched the surface of the science behind making high-performing surveys, after all, there are many books solely dedicated to the topic!     

Book a free coaching session today so we can create your first survey together. We can support you in this task and the next one will be easier.

FRANCOIS SOUYRI14 JUNE 2021