Best practice tips for survey-writing

Patrick Phelan | 5th May 2017

“It’s not what you say, it’s how you say it…”

You may have a well-thought out survey that covers every aspect of your business – but that isn’t enough to create an engaging programme that provides you with the right data. The way you ask the questions is equally as important.

The layout of a question can change the recipients answers and affect response rates, employee buy-in and the validity of your results. Syntax, emphasis and grammar has an incredible amount of influence on your recipient’s feedback. A good survey should be clear, simple and unbiased.

Below are some common faux pas you should avoid if you want to create an effective survey that provides reliable and relevant data:

Don’t over-complicate it

Simplicity is the key to success. If your questions are too complicated or lengthy then it can significantly reduce your response rates. To avoid people being intimidated by your questions, keep them short, simple and relevant. Your surveys should include 3-5 questions, to ensure they barely impact the recipients’ day.  This will boost completion rates and ideally improve the validity of your data. Surveys of this length, that are sent regularly are known as pulse surveys.

By reducing unnecessary phrasing, you will shorten your questions without affecting the meaning. Let’s look at an example for clarity:

[icon name=”times-rectangle-o” class=”fa-2x”] On a scale of 1-10 how likely are you to recommend us to a friend or business partner?”

[icon name=”check-square-o” class=”fa-2x”] On a scale of 1-10 how likely are you to recommend us?

Knowing your audience will allow you to write the questions in a way that aligns with their tone and language. For example, if you are surveying clients of a law firm, you should avoid using legal jargon. However, if you are surveying the lawyers themselves, then it is advisable to use their unique jargon and specialised tone.

If you need to include industry terms or acronyms, then make sure you provide an example or a definition:

[icon name=”times-rectangle-o” class=”fa-2x”] On a scale of 1-10 how do you rate the leadership team?

This question is quite vague, as some of your people may be unaware who is in the leadership team.

[icon name=”check-square-o” class=”fa-2x”] On a scale of 1-10 how do you rate the leadership team (HR Director, CEO, Managing partner)?

This will ensure everyone fully understands your survey and can therefore provide reliable feedback.

Don’t ask two questions at the same time

These are known as double-barrelled questions and are often very confusing for the recipient as they are unsure what question they are answering. Consequently, these questions rarely generate reliable data.

[icon name=”times-rectangle-o” class=”fa-2x”] On a scale of 1-10 are you satisfied with your salary and bonus package?

Some recipients will answer this question with their salary in mind, while others will answer with their bonus package in mind. This will ensure your data is unreliable, as you won’t know what question their score relates to.

Instead, you should separate the questions. This will ensure that only one subject is being measured for each question and the recipient fully understands what they are being asked.

[icon name=”check-square-o” class=”fa-2x”] On a scale of 1-10 are you satisfied with your salary?

[icon name=”check-square-o” class=”fa-2x”] On a scale of 1-10 are you satisfied with your bonus package?

Avoid leading questions

Leading questions will sway recipients and lead them towards a specific viewpoint. When your people are faced with a leading question they may lose faith in the surveying programme – as it can look like business leaders are trying to push them towards a certain answer. Most businesses do not want to send out leading questions, but some are still unknowingly doing it.

The first step is to check if your surveys contain any biased or one-sided language.

[icon name=”times-rectangle-o” class=”fa-2x”] On a scale of 1-10 how much fun was the recent project?

By using the word “fun,” you are clearly suggesting to the recipient what your desired answer is. This can massively impact how your people answer the question and the validity of their feedback.

[icon name=”check-square-o” class=”fa-2x”] On a scale of 1-10 did you enjoy the recent project?

With neutral, unbiased wording, this question is more balanced. This will help you to gather reliable and honest feedback.

Avoid loaded questions

At this point you may be wondering what the difference between a loaded question and a leading question is? A leading question will direct the recipient towards a certain viewpoint, whereas a loaded question will force them to answer in a way that may not align with their opinion – or even their reality.

[icon name=”times-rectangle-o” class=”fa-2x”] On a scale of 1-10 what do you like best about working here?

This question forces the respondent to discuss why they like working here, which is very presumptuous. They may not enjoy working here at all – let alone be able to pinpoint the specific thing they like best.

[icon name=”check-square-o” class=”fa-2x”] On a scale of 1-10 do you enjoy working here?

This is much more balanced and will not come across as self-serving, as it provides the recipient with the option to answer freely.

If you want to learn from your people to help you improve the culture and make smarter business decisions, then it is key to provide them with the opportunity to share all their ideas, innovations and concerns. Often you can do more with negative feedback, as is it is more actionable and therefore presents better opportunities to make a real difference.

Now we’ve established how to write your surveys, let’s examine how you can select the best questions:

Choosing the best questions for your survey

To determine what questions to ask, you should think about what you want to get out of the survey and what sort of data you want to collect. Consider who the data is being gathered for and whether it is relevant to them. This is always the best starting point for any successful programme.

Think about the strategic objectives of your business and what questions will add impetus to each strategy. For clarity, split up complex areas into smaller and more manageable sections.

If your surveys are designed to be reported back at boardroom level, then it is important to consider boardroom discussions and identify core areas of focus to help draw up your question set. Think to yourself “What will make the board sit up and take notice?”

Construct questions that align with your business goals

Each question should have a clearly-defined purpose and be useful for both parties. Consider how you will use the data – if it is too difficult to devise strategies then you should consider asking different questions. Be strict with your selection process and ensure that every question will help to improve your organisation and ultimately help you generate more profit.

Sanity check your survey

Always consider the following questions before launching your survey:

  • Will this provide me with relevant data?
  • What am I expecting the feedback to look like from this question?
  • Is my meaning clear?
  • Could any of my questions be considered too vague or broad?
  • Am I leading the respondent?
  • Are my questions biased?
  • Am I asking any double-barrelled questions?

Before you send out your pulse survey – ask someone who is not directly involved in this project to proof-read your questions. What makes sense to you, may not necessarily make sense to everyone else – especially when taken out of context. A fresh pair of eyes is always beneficial.

By following these best practice tips, you will create surveys to help you effectively track and measure your people. You can use this data to improve working conditions, build on successes and ultimately create a culture of engagement and productivity.