April 12, 2019
Survey Design – Create quality without a research background
Research in any field needs a clear purpose to be successful – business is no different.
For non-researchers, the idea of crafting a survey can seem anything from utterly daunting to fairly trivial – the truth sits somewhere in between.
The design of a survey is imperative to its success: a good survey greatly improves the quality of data provided, data which could ultimately be used to make investment decisions worth millions. On the other hand, survey design is a skill anyone can learn, whether or not they have a background in research.
Throughout my career I’ve had thousands of surveys cross my desk, many from non-researchers, and I’ve found there are a few finer points of survey design that can significantly affect the success (or failure) of the project.
Survey Design Questions to Ask Yourself
Who are you targeting?
Consideration of audience is the critical first stage of the project – although it happens long before pen hits paper on the questionnaire, the decisions you make have serious implications for the language and information you will use later.
A project targeting a younger audience for instance, requires more informal, succinct content to engage a shorter attention span. B2B respondents on the other hand, are typically more highly-engaged and willing to digest sophisticated language and concepts – you can, therefore, push the length and formality of your questionnaire.
It’s similarly important to bear in mind how well-informed your audience is on the topic, and thus how much, or how little, information you need to give in your survey to elicit a positive response.
Ultimately, there’s no limit to the potential variables at play, but operating with this idea in mind, you ensure your survey design strikes the right chord.
What are you trying to achieve?
Research in any field needs a clear purpose to be successful – business is no different. Setting hypotheses before conducting a survey ensures everyone involved has a concrete idea of what the study is attempting to prove or disprove.
In addition, brevity is key – it’s certainly important to understand what you want to find out at the top level, but also to realise that one survey can’t give you all the answers. Maintain focus on one at a time and the answers that project can give.
This, in turn, extends to the questions you ask – being frugal with the number and length of questions is a skill, perfected it allows you to zero in on the exact information required, avoiding wasted costs and superfluous data. Put simply, cut out the questions you’re not sure will give valuable insight.
The final “objective-related” factor to consider is how you want the data to be split. Demographic, Firmographic and even Psychographic variables can be used to slice your results for easy access to insight. Understanding this before the project goes in-field can save a lot of hassle after the fact.
At the end of the day, your sample are people and they’re donating their time for this – it’s up to you to make the experience as smooth as possible!"
What to Consider When Producing the Survey
To roll out a somewhat dated phrase, it’s time to put pen to paper. High-quality, effective surveys all share five key attributes:
- A logical flow–you can’t bounce back and forth from topic to topic expecting respondents to keep up. Questions should lead naturally on to relevant questions – confusion breeds disinterest, disinterest breeds incorrect answers and incorrect answers breeds angry stakeholders!
- Neutral questions –Confirmation bias is very real, and it’s easy to allow your hypotheses to influence the wording of questions. This, of course, defeats the purpose of conducting the research in the first place. If the hypothesis is true, it will come out without the help of leading questions.
- Short and snappy –In a world of distraction, drop off ramps up over time and quality will fall with numbers. Always stay below 20 minutes and remember: the younger the audience, the shorter their attention span.
- Asking the right people –Consider your target audience. Aim for the most relevant respondents but don’t get too niche, if you’re trying to understand drivers to purchase an iPhone X, you could target iPhone X owners, but to widen the pool you could also consider respondents who own competitor devices. i.e. why did you decide against an iPhone?
- A team effort –Make the most of the expertise you have available, between your team and your survey provider, you can ensure all bases are covered, avoiding any unwelcome surprises.
Basing your survey around these five qualities is a recipe for quick success – once you’ve finished, be sure to try it yourself, before you send it out to the world. At the end of the day, your sample are people and they’re donating their time for this – it’s up to you to make the experience as smooth as possible!
What is a Typical Checklist?
In addition to general theory and best practice of survey design, a checklist can be a real helping hand when it comes to covering all bases with the questions in your survey – especially when you’re trying to stay tight.
The following is a great framework of themes to ensure your survey stays on track and achieves what it sets out to achieve:
- Demographic / profiling– Who are you and what do you do?
- Screening questions– Are you the right person for this survey?
- Behaviour– How do you interact with my topic of interest?
- Brand / product– what are the brands / services you like / use?
- KPIs– what drives you in this area? What do you like? Dislike?
- Guest question– used to answer specific hypothesis that doesn’t fit within the other brackets.
This will obviously vary greatly between surveys, but in the same vein as the classic IEDA sales funnel, it provides a solid structure, around which you can build the individual questions and themes you require to make your project a success.
Get in Touch
Make a start by telling us a little about what you're looking for.
If you’d like to speak with one of our experts right now, please call: 020 7632 3434 (UK) or 646-856-9750 (US)