Most of our research work involves in-depth qualitative interviews – over the telephone or in person. But when the need arises we design and implement online surveys for clients.
These can be exceptionally good value research activities, particularly if a client wants an overview of opinions among a large group of respondents within a particular target group.
There are a wide range of really useful websites giving hints and tips about different elements of online surveying, and I’ve published a sample of these at the base of this article.
But I thought it might be useful for fellow communications, PR and marketing practitioners if I listed some tips that you won’t necessarily find anywhere else.
Our surveys tend to be quite complex affairs and part of a wider body of study, so I’d recommend approaching an agency if that’s your requirement (I would say that, wouldn’t I!) However, if you are doing something fairly straightforward in-house then these may come in handy….
1. Incentivisation – we find that mixed models of incentivisation work best, in which a respondent has the chance to gain something personally (i.e. enter a prize draw) as well as raise money for a charitable cause by completing the survey.
2. Charity support – spending time considering which charity to support is a worthwhile investment; if it resonates strongly with the respondent group, it can have a really positive impact on response rates.
3. Split testing – in particular, where a target group is particularly large, it’s worth split-testing different email introduction and subject heading types as part of a pilot stage, and running with the most successful for the bulk of the campaign.
4. Pilots – running a two-stage pilot process can add significant value to a project. Typically these involve an: initial test of the survey with a friendly group of respondents (and it’s important this is neither you nor the ‘client’, as you will be too close at this stage to see errors or issues); split testing of invitation types (see above).
5. Watch out for betas – the functionality of most of the current crop of paid-for online survey tools is impressive. However, take care when using ‘beta’ test versions of sites as there is (naturally) a higher risk of glitches that could impact your campaign if you use them. Current issues with fonts in the SurveyMonkey beta email tool is a good case in point.
6. Invitations – there are lots of useful blogs about what to include in an email invitation, some of which are listed below. In crude summary:
- Personalisation [First Name] etc
- Thank you
- Why you’re doing it, who for and how results will be used
- Length of time to complete
- Confidentiality assurances
- Contact details
7. Subject headings – again, hints and tips abound (see below). In the end it comes down to good copyrighting skills. Brevity, relevance and an answer to the inevitable ‘what’s in it for me?’ question.
8. Invitation timings – again, it’s worth splitting invitation email dispatch between different days and times of day.
9. Reminders – in-house email management or survey software email tools will often build in reminders for you, and send only to those who haven’t completed the survey. Reminder timings should be closely related to deadlines – work backwards from the deadline when setting them, rather than forward from the dispatch time.
10. Deadlines – Beware the distant survey closure deadline, a gift to procrastinators. The vast majority of responses will be collected in the two or three days after an invitation or reminder receipt.
11. In whose name should the email be sent? The person or organisation best known to the respondent group, even if you are using an agency like us, as this will boost response rates while minimising problems with spam or junk filters.
12. Don’t spam – not only is it poor practice, it’s a waste of everybody’s time. Only send survey invites to people who have agreed to receive communications from the organisation or, if you’ve bought a list, contact the recipients through your own email client to get their consent before sending the survey invite.
13. Beware contact lists provided by online survey companies – I have yet to hear a positive story about these. (Please do get in touch if you have a different tale to tell.)
14. Watch out for public links – on one occasion a client contact accidentally posted the link to an invitation-only, incentivised online survey in a public forum. It was very quickly attacked by spam bots and it took quite a while to clean up the results. Where posting links to surveys, it’s best to publish to a closed group.
15. Place the most important questions at the front of the survey. Sometimes this may involve pushing demographic data down the running order.
Some useful links: