“Would you like Walmart to be less cluttered
In 2009, this question lead Walmart to remove 15 per cent of their inventory, shorten shelves and clear aisles. The decision cost them $1.85bn.
Walmart interpreted the response to this poorly phrased question as support of their existing plan – to compete with Target by providing a neat and organised shopping experience. Sadly for them, the answer was meaningless. No customer, even those completely happy with Walmart, would ever say they like clutter.
A well-phrased question can bring out insights into your business, which can be critical for growth, directing investment or avoiding mistakes. But for this to be useful, the data must be meaningful, which means asking the right questions. Here are some tips on how to write survey questions to get the most from your market research.
Set a goal
Before you start writing your questions, you need to work out your objectives. Do you want to understand how happy your customers are, how your brand compares to competitors, or the appetitive for a new product or service
Agree what business objective your research is setting out to inform. Then make a list of goals that you would like to achieve from your research. If you can’t think of three aims then put the pen down theres no point bothering your respondents if you dont know what insights you are looking for.
Once you know what you want to achieve, you can start writing the questions to gain the insight you need.
Avoid leading questions
Ice cream is the best thing to eat at Wimbledon : Strongly agree – Agree – Disagree – Strongly disagree
Make sure that you’re not writing questions solely to elicit the answers you want. This will only skew your data and lead to misguided business strategy.
Marketers choose words very effectively to elicit a particular response. Market researchers must do the opposite by creating neutral wording which draws out the respondents genuine opinion.
Be sure not to guide your respondents answer by offering your opinion. Asking open questions with a series of responses will provide you with more accurate information than asking respondents to what extent they agree with something. Your goal is to understand your market in order to shape your plans, not to create evidence to support your existing plans.
“When at Wimbledon, do you prefer to eat strawberries or ice cream
Dont forget to include all possible responses, such as I dont know” or None of the above and Not applicable . Asking questions that presuppose certain actions from you respondents (e.g. attending Wimbledon) will skew your data by forcing them to pick a choice that does not reflect their view.