To survey, or not to survey.....

Happy new year everyone. 

Hope you got through the turkey, Baileys and mince pies (other festivities/comestibles  available) and are  plugged in to take 2018 by the horns (other animal parts also available).

January often brings hope and expectation. Or a sense of tension and gloomy foreboding. Is such optimism/pessimism well-founded? One way to find out would be to do a survey. 

Meh.

In this blog, I’m going to look at when it's good to use surveys, their limitations and when you may be better off doing something else. Knocking a survey together is dead easy; crafting a questionnaire that delivers is a different prospect. It's like comparing Dairylea with Stilton. And so, ok, sometimes there's time for Dairylea......... if you're a maniac with no taste (see left).


Why survey?
Good question. One you should always ask yourself and/or your survey sponsor. If you don't have a  clear set of aims and objectives, scrap the idea. If the reason for doing the survey is anything like “It would be quite interesting to…..”, turn on your heel and FLEE.

Medieval fleeing

Surveys can confirm hypotheses and quantify suppositions, but rarely provide insight. If you're after insight or new ideas or have no idea of your audience views, you should chatting to them, not surveying. Listening rather than asking questions. Unless you have a good idea of your outcomes or potential outcomes beforehand, you’re on dodgy ground setting questions and expecting meaningful answers.

How do I confirm hypotheses and quantify suppositions?
Surveys. Though bear in mind, surveys are blunt instruments. The ideal market research survey should:

• be accessible (usually online - but do consider anyone without access to computers);
• be as short as possible - only ask questions key to the survey aims and objectives and keep open ended (non-precoded questions) to an absolute minimum;
• be piloted (tested). Get another set of eyes. I get fantastic help from colleagues in the CRM team, who have no problem telling me when questions or layouts don't work. I don't pretend to be Gary Gallup or Micky Mori, but would be happy to cast an eye over any survey you may be proposing - give me a shout;
• like I said, be AS SHORT AS POSSIBLE - if you don't need to know the views of different sexes or age/ethnicity groups, don't ask (it's good practice from a GDPR point of view);
• contain questions which are  clear and unambiguous, and easy to answer (remember, you can't explain things or correct any misinterpretations if you're not face-to-face). Here are some of the more common mistakes when drafting questions;
• be something which can be acted on – don’t ask about a la carte dining experiences if you’re researching locations for a grotty tea bar;



Keep questions as short as possible

• be something people will want to do. Offer incentives (very important for students; staff aren’t allowed a prize draw, although you could think about a charity donation based on the number of responses).

Check and share data early – an early check will confirm if an online survey is working. It may also prepare managers for any shock results if you're lucky/unlucky to get them.

Be prepared to be ignored
So. There will be times when minds are already made up. I recently carried out a survey looking at different choices of artwork. It was a good survey and gave a clear "winner". Inexplicaby, the third most popular design - there were four in all - was used. Fine. This isn’t a sob story. It does, however, illustrate how running a survey can be a waste, both in development and time spent by respondents. It’s always worth asking “will these results matter”? If not, why are you surveying?

Who do I survey?
You should be looking to represent the views and taking time to identify and reach a specific audience. This could, for example, be students, applicants or staff. It could be a combination of groups. Think about each audience's expectations – students and applicants may expect an incentive (immediate gratification), staff may be looking for feedback down the line to engender (so encouraging engagement and buy-in).

When should I survey?
Again, think about your audience and plan.

• For most audiences avoid holiday times;
• Check what else is going on, survey-wise. If you want to survey students, you need to contact the Student Experience team;
• If you want views on an initiative, it's best to ask the questions after the initiative;
• If you want to test the initiative, you may think about repeating questions before and after the initiative and check out the difference.

Anyway, this is all basic stuff. Quant 101. But then surveys should be basic.

Thanks for reading. This weekend, I'm going to have a shave and a haircut. i might think about The Ramones. Here's The Ramones with a top tune: