Helping an agency succeed their campaign to increase visitors at the hand of trend spotting online quantitative research.
Tip No. 1 – Know the headlines you’re looking for
Having a set of desired headlines that you’d like to get from the survey as a base makes writing the questionnaire much easier. Do you have a hypothesis to validate? Do you have an impact or maturity model to create? If so, then make sure you create questions that lend themselves to that specific headline or model. The questions should support the story of the headline and it will be easier to get the chosen results. But don’t make them too self-serving!
Tip No. 2 – Use closed questions
Use closed questions that have a list of ready-made responses (i.e. code frames) that respondents can tick. By doing this you remove the possibility of having an almost infinite list of free-text responses, making analysis much quicker and easier. Understanding and researching issues related to the topic/market prior to code-frame creation is key to ensure a comprehensive list. Keep in mind that the longer the list of options, the more likely you’ll get disparate results. Consolidating similar codes into one can help, while still maintaining detail.
Tip No. 3 – ‘Other please specify’
If you’re in doubt of knowing all the potential ready-made options, use an ‘other’ category, particularly if you are looking to understand the impact of a problem or list of information sources.
Tip No. 4 – Decide between ‘select one’ or ‘select all that apply’
Use select one when you want to know exactly which is the ‘most important’ (or equivalent), or when it’s appropriate (e.g. on an
agreement scale). Bear in mind that if you ask respondents to select one from a long answer list it can not only be tricky to answer, but it can result in very split results with no clear ‘winner’. In many situations using ‘select all that apply’ or ‘select up to three’ is the best option. This allows you to get higher percentages, whilst also providing a rank order of answers (i.e. you can still see which comes out as first, second, third from the list).
Tip No. 5 – Make it as easy as possible to answer
To make it as easy as possible, try to avoid ranking questions (e.g. what’s your 1 st , 2 nd , 3 rd choice), as they are more difficult than they seem for respondents to answer (and also more difficult to analyse!). Instead, use rating question or Likert scales (e.g. marks out of 5 or 10, strongly agree, disagree etc.). This provides better quality responses, but can still be reported as 1 st or most preferred when doing PR research for example. However, we also recommend limiting the number of grid questions (i.e. to what extent do you agree or disagree with the following statements) – see below!
Tip No. 6 – Make sure that the format is interesting
Mix it up with short and long questions. Include images or video. Put some simple to answer questions among a set that require more thought. Also, make sure to keep the formatting of the questions clear and interesting, don’t just ask lots of grid questions (i.e. rating questions) as respondents easily get bored of these.
Tip No. 7 – Watch the length
Try not to ask more than 30 questions (roughly a 10 minute survey) as this can mean people lose interest, impacting the quality of
considered responses as the survey progresses. Be concise, avoid ‘nice to haves’. Respondents prefer the shorter surveys and they cost less to do.
Tip No. 8 – Be mindful of the audience sensitivities
Try not to be offensive, even if you want to create a thought provoking or controversial headline. Ask yourself how you would feel if you were in the shoes of the respondent and asked that question?
Tip No. 9 – Make sure the questions are asked in an order that flows
Asking questions in the right order can be the difference between good and bad data. Ensuring questions which are related to each other are asked consecutively, and placing questions into sections means that respondents will be ‘in the zone’ for that type of question before moving onto the next topic. Also, asking demographics (if they aren’t being used as screeners or quotas) at the end means respondents answer the questions requiring more thought first, before they start suffering from survey fatigue (which they shouldn’t be anyway if you keep it under 30 questions!)
Tip No. 10 – ways of maintaining quality responses
- Use screening questions to double check you are talking to the right people
- Leave the easy to answer questions (e.g. demographics) to the end if they’re not needed for screening at the beginning
- Keep the questionnaire short
- Don’t ask two question in one; be clear about what you are asking
- Avoid or explain technical terms