Long ago, when I first started working in market research, nearly all our surveys were carried out over the phone. The trouble was, there was just so much we wanted to know about consumer behaviour. This led to surveys of 100+ questions, with some amazingly complex routing options to cover every possible scenario about how families used technology in their homes.
The call maker had to phone up a participant, go through half the survey, then at the end of that, make an appointment to call back at a future time, so they could go through the rest of the questions, after both the interview and the respondent had had a stiff drink and/or a long lie down.
Once we’d got our target sample, we’d have to cobble together both surveys into one spreadsheet (that Excel at the time only let you use 256 columns per spreadsheet didn’t help), then begin our analysis. What fun!
The New Way of Surveying
If you heeded our CEO Paul Hudson’s warnings in his recent blog post you’ll already be familiar with the term ‘survey overkill’, and I think we can definitely describe the olden days of surveying as a classic example of overkill.
The great thing about online research and moreover the fact that our participants are never offline, segueing through the day from smartphones to desktop to laptop to tablet is that you can do more surveys. In fact, little and often is the key to successful modern surveys – not only does the way we are online now mean we can do this, it means we must.
| Tweet This | |
| "The keys to successful agile surveys are a careful new balance of frequency and brevity." |
But before I outline the four ways to go about this, let’s talk about what agile is and what agile isn’t. Agile research using surveys doesn’t mean:
- It’s just like ‘normal’ surveys but more of them
- That you don’t need to plan and think about what you’re asking
- Just a case of making it up as you go along
How Do You Make Research Surveys Agile?
1. Little & Often. Thankfully, gone are the days of the 200 question behemoth of a survey I cited in my earlier example. The best way to make sure your surveys are agile now is to hack them right back to just what you really need to know. Even better, work out what you really need to know up front and spend your time crafting maybe 10 beautiful, concise questions rather than pruning down from 80 mediocre ones.
2. Two Week Iterations. The classic agile methodology for Scrum and the like relies upon two-week iterations. And, you know, this isn’t a bad idea for surveys, either. We would generally recommend leaving a survey open for 8-10 days, so that gives you time to plan and write your survey, run it, analyse and present the findings. Sounds scary? Remember, we’re talking about bite-size, manageable and very specific sets of research objectives here, not asking every possible question you can on the subject. Missed something? Ask it in the next iteration. No problem.
| Tweet This | |
| "Don't worry if you don't have time to ask every question in an agile research survey - it can be asked in the next iteration." |
3. Pre-Screened Participants. If you’re using a little and often iterative approach to your market research surveys, the last think you want to be doing is to have to ask your participants to wade through swathes of screening and demographic questions each time they complete a survey.
Having a carefully profiled community panel means that you not only know you’re talking to the right people, it also saves having to ask them irritating, unnecessary questions every time you survey them and keeps your survey lean and mean.
4. Mobile Friendly Surveys. The relationship between agile surveys and mobile is a bit of a symbiotic one: little and often is the best way to conduct mobile surveys (and therefore, given so many people now complete surveys on their mobile given half the chance, all surveys). Meanwhile, surveys on mobile are the perfect way to get contextual, in the moment feedback.
This means that you have to be using a survey platform which makes running surveys on mobile simple and offers a good user experience so that the quality of your data is not affected no matter how your respondent completes the survey.
Have you managed to make your market research surveys agile? If so, let us know what changes you made and how you approached agile methodology in the comments below.
