"The data tells a story, and once you know the story you can change the ending." — Debbie King, CEO of Association Analytics
Data, even "big data," has become a buzzword in both the business and association space. Everyone wants data because, as Debbie King said, it can help organizations change. It can help them evaluate an industry, develop strategies to increase revenue or member retention, and tell them what benefits members want to see. But how do you get data? Your membership management software, online community, and website are a few options.
Possibly even more popular, however, are surveys. Surveys provide direct data and feedback from your members, many of whom are known professionals in their fields. They're an excellent tool for benchmarking research and ask members what benefits they want to see in the future.
Using surveys to collect information is only one part of a larger process involved in getting actionable data, however. After the data is collected you have to decipher the results, otherwise all the work that went into your survey will come to nothing. You can't use data that you don't understand.
Unfortunately, all that raw, unorganized data that you don't understand can seem overwhelming, especially if you're not sure where to start reviewing the data or how to analyze it correctly. If you don't do it right and miss a step, you could end up with data that sends you down the wrong path—costing you a lot of time, money, and brand damage.
While data analysis can seem difficult, it becomes easier when you start with small tasks. Work with your data piece by piece and shape it into information that's easy to interpret. The ultimate goal is for your data to be useful so your association can improve, or your members can understand your industry and be better at their jobs.
Here are four tips that will help you break down your survey analysis, making it easier to transform your data into useful information.
A simple, manageable jumping off point, cleaning your data will help ensure that the rest of your analysis is useful and accurate.
Begin by getting rid of duplicate responses and evaluating incomplete surveys. If someone filled out enough of your survey to make it meaningful, keep the data. If only one question out of ten was answered, however, throwing out the response may help maintain data integrity.
Putting in the time to scrub your data up front will save hours on the back end, allowing you to slice and dice data more quickly, and with more agility, later in your analysis process.
You can and should use all the data from your survey as an overview of responses, but you should also group data. By grouping your survey responses based on relevant populations—often known as filtering or segmenting responses—you can get more information out of your data.
That extra information comes from knowing if different types of people responded differently to your survey. The percentage of members who love your email newsletter may different by age, for example. Your data, based on groups, might look something like this:
Just by grouping your data based on age, you gained three additional percentages to work with. You can use that extra information to personalize your offers, communicating with and providing value to different member groups in the ways they prefer.
Other relevant populations to group by might include job title, length of membership, engagement level, and other demographic data.
Rows of numbers and data points take more time and effort to analyze, so present your results to executives and board members in table or graph form.
Well-organized tables with exact data points often allow for more detailed numbers analysis. Graphs, on the other hand, are a little less specific. They are much easier to read at a glance though, and allow you to quickly identify trends. They tell a visual story.
Expert Tip: It's easy to skew the results you see in graphs, so be careful to accurately display results. One example of how results can be skewed is through the use of incorrect scales and intervals. If you use age ranges like under 40 and 40-49, for instance, you will have more responses in the under 40 section just because it's a larger interval. To avoid this, use equal intervals like 20-29, 30-39, and 40-49 instead.
What is data? Numbers and responses. Nothing more, nothing less. It's your job to humanize your data and make it relevant by applying context. Why did you create the survey? How did you distribute it?
If your survey was on event satisfaction, use that as the base for your context. Tell your board what the event was about, who it was for (such as national or chapter members only) and other similar information. Anything that's relevant to the questions you asked in the survey, why you asked them, and what could have contributed to the responses is context you should include.
The right contextual information will help make your data and its analysis relevant to your association, industry, and members.
Even the best survey data will be useless if it's not properly analyzed, with the results presented in a way that's easy to understand. Take the time to do analysis right, starting by making sure that you're working with high quality, clean data, right from the very start. Then filter your responses and present them visually with the proper context so everyone understands how your results are relevant.
When you're finished, you can start identifying trends in your industry or areas of strength and weakness in your association. Put that information to good use. Take action to improve your association, educate your industry, and prepare your members for the future.