Top Tips for Collecting Audience Surveys

Audience surveys are a valuable tool for organisations that want to learn more about their target audience. It might be part of an evaluation, monitoring your strategic objectives and KPIs, or you might be required to collect audience data as part of your funding requirements. However audiences are busy, and getting your data collection off the ground can seem difficult at first. These top tips should help you focus on what matters and get results.

If you are just getting started, you might first want to check out our Introduction to Audience research here. 

Image credit Paul Holmes

Design

Keeping your survey short and to the point will make everything else much easier. Stay focused on your organisation’s strategy and key goals and try to avoid adding questions if possible. Being able to say that your survey is 5-7 minutes long means that most people will agree to complete it, while surveys that are 10 minutes or longer are much less attractive and result in wasted opportunities.

Clean, minimal design is important – keeping the design and language brief and clear will avoid overwhelming respondents, reducing cognitive load and helping respondents complete your survey more quickly and attentively.

If you have similar questions or repeated question sets, keeping the structure or type of response asked for consistent will avoid confusion, but it is important to clearly differentiate each question – bolding or underlining key changes in a question’s subject or topic can help.

Mixing it up with a variety of topics and question types will generally help keep respondents engaged.

They’ve agreed to tell you what they think, so you should consider what an audience member or visitor might want to communicate to your organisation, and what an interesting survey might look like to them. There’s nothing worse than spending a few minutes on a survey but not getting chance to express yourself. If you don’t already have questions that allow this, an ‘anything else?’ open text question is an easy way to give space for responses about things you might not have thought of.

Make sure you carefully test your survey to review the language used, that the questions asked and answers available are relevant to a variety of survey takers, and that any logic or any dynamic elements within the survey are all working correctly. You should also make sure that it really takes the amount of time that you claim.

Consider distribution channels

There are a lot of ways to get your survey out there, although you should consider that different people can be found in different places. It may be helpful to combine a few different approaches and channels to maximise your response rate and get a good mix of people.

If you’re conducting a survey of your existing audience, sending surveys to ticket or events bookers is a great way to connect to them. You can time a survey email to come through to them straight afterwards – you’ll always get a better response rate if this is within 24 hours of a visit or event while it’s fresh in their mind. If the system allows you, you could also send a follow-up reminder email 2-3 days after the visit if they’ve not already completed the survey.

You might also find that a more hands-on approach will pay dividends. You could make a QR code or short link available in your venue – for example, on a receipt or ticket, posters around the venue, or re-usable cards in your exhibition space or auditorium. It’s worth briefing front-of-house staff to point people specifically to them and make an appeal.

However, one thing to consider is supporting people who have access needs or are less comfortable using technology – you should have front-of-house staff ready to assist people to complete the survey if they need it.

If your survey has a particular research focus, like developing new audiences, or a programme of events or workshops for young people, then you’ll need to find other channels, spaces or organisations where you might find the groups you’re looking for. For non-attenders or potential audiences, some organisations might need to turn to research agencies, fieldwork suppliers, or online research panel services.

Image credit Neil Pledger

Face-to-face

Alternatively, if you have the resources available among staff or volunteers, you could conduct surveys face-to-face with respondents. Conducting surveys face-to-face allows you to make a direct appeal to visitors and boost your completion rates, make a connection with them, and allows for a degree of quality control and probing for more information during the survey.

When considering face-to-face interviewing, you should think about any training and buy-in for these colleagues, or whether you need to bring in a fieldwork agency or dedicated staff to deliver this. The DCN can offer dedicated support to train staff to carry out interviews effectively, or help you find suppliers to deliver fieldwork or research services.

Making an appeal

Once you’ve chosen your distribution channels, it’s important to make sure your survey is being given the spotlight necessary to maximise completion. If you’re emailing people after a ticketed event, make sure it’s not competing with other actions or signposts. Ideally, the survey link would be sent in its own dedicated message rather than buried in a general email.

It’s important to clearly frame the case for the survey too; people’s time is valuable to them. You can say that the survey really helps you understand what your audience need, understand what’s working for them, and to find out who isn’t visiting and what isn’t working. You can thank them and acknowledge them for taking the time to stop and help, for having the patience to answer all the questions.

Think about the quality of the data

The data you collect from your survey needs to produce a dataset that is robust and useful when it’s analysed. There are a few main factors to consider here. 

• Sample size – you need to gather enough surveys so that you can have confidence that any trends you identify in the data are statistically significant. The more surveys you gather, the more reliable it is and the greater the confidence that any outliers have been smoothed out. It also makes it more likely that you’ll be able to usefully look at specific groups within your respondents (e.g. online audiences vs in-person, visitors to different shows or exhibitions, or different demographics or segments). You can refer to the Illuminate targets for minimum and suggested targets for different audience sizes.

• Random sampling – if you are directly appealing to and approaching people to ask them to do your survey face-to-face or using a QR code, you should consider that your interviewer or colleague may also have subtle biases in the types of people that they ask. Maybe they are more comfortable approaching a particular gender, or they presume that families with young children will be too busy to help. To counter this, professional interviewers use a counting approach where they pick a number appropriate to the visitor flow (say 3, or 5), and then approach every 3rd or 5th person they see. If it’s a specific person within a group, you can cheerfully explain that you must count and approach a specific person to make it randomised. If you have a large enough team, another consideration is that you might want to make sure you have at least three people responsible for recruiting respondents, so that you can identify if one interviewer is doing something unusual and any differences in approach or profile are smoothed out.

• Across your offer – ideally, the composition of your survey respondents would reflect the different groups within your audience. A key consideration can be around making sure the audiences of different events, shows or activities are represented in proportion – so you aren’t gathering too many surveys from one activity and very few from another.

• Stratifying by day and time – you might also consider making sure you are spreading your survey across different day types, and times of day. For example, a daytime weekday audience is likely to have a very different demographic profile to an evening or weekend – with the mix of ages, families, and working people varying. A general rule of thumb is about 60% weekday, 40% weekend, but you can check your visitor counts for an exact split as it does vary.

Keeping on top of achievement

Finally, you’ll want to dedicate some time to checking the number of complete surveys regularly – at least once a week. Most importantly, you’re looking to find out how many surveys you’re getting per day or week, and then make projections about how many you’re likely to end up with. 

A simple sum to work this out: 

Average number of surveys per day / week achieved so far 

X 

Total number of days / weeks in the research period 

= 

Projected total number of surveys 

It can be very useful to maintain a spreadsheet that will keep track of this for you as you fill in daily or weekly survey totals. If you’ve recently seen an improvement in survey response rates, you might want to use an average over a recent period to make projections for the future. However be aware of any seasonal dips or gaps in your programming that might bring them down again. 

You can then see if you’re above or below your target, and whether you need to do anything to boost completion rates or share and congratulate your team if they are driving these surveys and encourage them to keep going. 

Another great motivation for your team to work hard at gathering surveys is to regularly share data or findings from the surveys, especially if they relate to their job roles, or events and programming on site. This will help them see the value in the surveys, as well as give them useful food for thought and feedback.

Privacy and sensitive topics

Questions that are more sensitive or feel unrelated to your interaction with the audience (e.g. detailed demographic questions) might need a short text explainer to reassure respondents that this data helps you understand who is visiting you, and therefore who is not, so that you can grow your audience.

It may also help to reassure them that all data collected is anonymous. Usually, survey data is analysed in aggregate, meaning that you look at statistics for the whole dataset rather than looking at the way that specific people have answered.

Data Protection

If you are collecting contact details or any information that potentially allows you to identify a specific living person as part of the survey, you will need to make sure you are doing so in a manner that is compliant with data protection legislation; primarily the Data Protection Act 2018 (sometimes referred to as UK GDPR). Most organisations will already process personal data and will likely have privacy policies, responsible staff, and processes in place.

If your survey is collecting personally identifiable information, you will need to point to your GDPR-compliant Privacy Notice before and during the survey. Make sure that it does reference the scope and purpose of your survey activities and any platforms that you use, especially if they are processing personally identifiable information and contact details. This is a legal requirement and fines for non-compliance can be severe, so it’s important to make sure that this is up to scratch and that you have Data Processing Agreements in place with any suppliers or platforms that will also be processing people’s personal data on your behalf.

You should think carefully about how survey records containing personal data are stored and consider regularly expunging them from your survey data platform or database to make sure that the individual responses stay anonymous. In general, you should aim to collect the minimum amount of personal data required for your research or other activities, and typically privacy policies will specify how long it’s going to be kept before it’s destroyed.

What next?