In this blog post we highlight seven things to think about when running an online survey for B2B.
There are plenty of tools available that allow you to create and publish your own online survey, without you needing to go to a specialist market research agency or to host it online yourself.
You may want to check out PCMagazine, who have put together a chart comparing ten different tools, and DemandZEN, who have compiled a list of their favourites.
Here are some things to look out for when choosing the best one for you:
Most tools will support mobile devices, but it’s worth checking. You want it to be convenient for respondents to complete, and making it accessible across devices is key to this. Some tools, like SurveyMonkey, even have a mobile app.
You may consider customising the survey with your company logo and branding. This will help it to look credible, and if you are including prospects in your sample, it will also help to build awareness of your company. Most tools will give you the option for customisation, but the extent to which you can do this and how easy it is to do varies across tools.
All tools will allow you to export the data you have collected from the survey. But depending on which you use, there may be options to export it as a CSV, Excel or Word file, or you may be able to export it to more advanced programmes like SPSS. Consider which program you want to use for your analysis and find a tool that will allow you to export to that program.
If you have never used an online tool or scripting programme to create a survey then you may require some level of hand-holding. Some tools provide more support than others so be sure to do your research and pick one you feel comfortable with.
The questions you ask will be determined by your research objectives – the specific topics or issues the research plans to investigate.
Here is a quick guide to the main types of questions you will most likely use in your survey:
Typically, a survey will consist of structured questions with closed answers, meaning respondents select their response from a set list. Questions can be ‘single code’ where the respondent can select one answer only (e.g. ‘Where do you live?’) or ‘multi code’ where the respondent can select more than one answer (e.g. ‘Which of the following brands are you aware of?’).
Closed-ended questions don’t allow for respondents to expand on their answers or provide detailed responses in their own words. You may therefore also want to consider including some open-ended questions in your survey, where respondents can type in their own answers.
For example, if you’ve asked ‘How likely would you be to consider purchasing from x brand?’ as a closed question, you may want to follow up with ‘Why do you say that?’ as an open-ended question. Avoid too many open-ended questions though – it will extend the length of your survey and analysis can be time-intensive. As a guide, in a 10-minute survey you may want to include two or three.
It is likely there will be questions you want to ask just a subset of your sample. ‘Routing’ allows you to do this by directing a respondent through your survey based on the answers that they give. If a respondent provides a particular answer then they are directed to a particular question in the survey.
For example, you may ask ‘Have you ever purchased a product from x company?’. If they answer ‘yes’ you may want to go on to ask ‘What type of product did you purchase?’. But if they answer ‘no’, you will want them to skip this next question and route them through to the following one.
It’s important to find the right balance between asking enough questions to capture all the data you need but not making it so long that your respondents lose attention. The ‘sweet spot’ is generally considered to be around 10 minutes.
If running your survey online, it’s useful to include a progress bar so respondents can see how far through the survey they are. Most online survey tools, for example Survey Monkey, will give you the option to include this. Of course the shorter the survey, the more likely people will be to complete it - so keep your questions relevant.
Before you send out your survey, test it. Ask other people in your business to complete it as though they were a respondent. Ask them how long it took, and whether any questions were confusing. Any issues can then be ironed out before you launch the survey.
It would be unfortunate to have gaps in your data because one of the questions doesn’t make sense or because people aren’t being routed through to the right questions.
If you are sending a survey out to your customers, it’s a good idea to be courteous and give them a heads-up first. Send them an email to let them know you are doing some research and would like them to take part. Tell them how they will be contacted, who by, and when. And give them an opportunity to opt out if they wish – you want to keep your customers onside.
Many things can impact on survey response rates, including the quality of your email contacts, the relationship you have with them, and the length of your survey. It goes without saying that the higher response rate, i.e. the more people you can get to complete the survey, the more robust and credible your results will be.
Your email invitations will be key to engaging potential respondents with the research. Here are some tips for optimising them to maximise response rates:
It would be frustrating to go to the effort of putting together a B2B research survey only to find that you have few responses and high drop-out rates, or data that doesn’t tell you anything useful.
The better your preparation before launching your survey, the more successful, and insightful, it will be.