This series might be called a series of “reality posts” since we are following LST in real time through its strategic planning process. To see earlier posts, go to the Categories drop down box and find the Strategic Planning Series.
As I reported in the last blog in this strategic planning series—which was about 90 days ago—LST sent out over 9000 surveys to all its different constituencies in order to help us know how to plan for the future. We received about 15% of the surveys back and if each survey averaged 50 questions, then we are talking about approximately 70,000 bits of information for us to analyze.
The ease with which you can get useful information out of your surveying depends primarily on two main factors:
A good online survey tool like www.surveymonkey.com will make the task much easier.
- A good tool itself provides some basic analysis of the captured data.
- The data analysis must be presented in both table and graph forms. You need it in both because different people in your organization process information differently. Your accountants will love the tables with all their numbers, but your public relations people will often understand the graphs much faster. Your board members will likely have different ways of processing information as well.
- Filters should be available to further refine the data. For instance, in LST’s Donor Survey, we asked “Have you personally done LST either out of the country or through FriendSpeak?” This question would seem to require a simple Yes or No, and would result in a percentage of Yes’s and No’s, but because of other questions that we asked, we could find out how many men vs. women said yes, how many of those who said no contributed over $100, and much more. Your online tool must allow for sorting and filtering your bits of information.
The demographic questions you build into your survey are the second big factor in making your information easy to understand. You only get out of a survey what you ask!
- If your survey is 50 questions or less, probably the first 7-10 questions should be the demographic questions that are important to you, i.e., the categories that could be important for sorting the answers. Here were our questions for our donor survey:
- Gender?
- Age?
- Average household income?
- Size of home congregation?
- Have you ever done LST?
- How many non-profits do you support?
- What has been your single largest contribution to LST?
- Largest annual contribution?
- What prompted your most recent contribution?
What these questions have in common is that they are going to provide sorting and filtering categories for all of the remaining questions in the survey.
In this particular survey, there are a total of 25 questions. After these first nine demographic questions, the remaining ones are mostly satisfaction or preference questions: Example:
- “On a scale of 1- 5, how satisfied are you with the way your donation is used by LST,” and
- “When giving to LST, how do you prefer to contribute?”
Inevitably, you will discover that you failed to ask one demographic question that you really needed to ask. We got all of this information back but realized that we could not separate our regular donors to our general fund who have given to us perhaps for years from those donors who had only given once to support a particular worker, often one of their family members. The difference between these two types of donors is huge—and we simply failed to ask the one right question that would allow us to separate them in our analysis.
And, finally, you are going to need to let many eyes look at the results if you truly want to understand what your surveys are saying. We gave all of our staff and all of our board members complete access to all of the raw data online, so that they could see everything, sort and filter the numbers as they wanted to, and then we asked for them to tell us what they saw as well.
One tip: you can be more open with the data if the surveys themselves are anonymous.
Next, we’ll talk about actually beginning to draft the strategic plan.
Mark's Blog 