I'm digging the numbers

December 1, 2015 6:11 PM

I've been given the opportunity to do some development at work (at a nonprofit) - yay numbers!!! I have tons of data to work with...but what questions should I be asking?

I've been given free rein to play around with the data we have from previous years' donations. Ideally I'd like to use this data to cultivate more donations in the future. We have a lot of data to work with in terms of donors, their donor history, amount, largest donations, smallest, averages etc. etc. etc. and it is all easily exportable. I am not going to have any trouble analyzing the data, but I am trying to figure out what the most interesting questions would be to ask. Lots of new donations will be coming in over the next month - so there will be lots to work with!

posted by Toddles to Work & Money (6 answers total)

Ooh! Playing with development data is basically all I do all day! I'd start by looking for some of the following things:

-What are your attrition rates? How many donors give consistently year-to-year and how many lapse?
-Do you have as many new donors coming in each year as people dropping out of your pipeline?
-How have the largest and average gift size changed over time? Are they increasing, decreasing?
-Do you have any donors that always give at the exact same time of year? Can you time their solicitation mailings appropriately to save some costs?
-What was a donor's behavior like before their LARGEST gift? Did they increase steadily, jump drastically, did a particular event trigger their gift?
-Are there particular geographic areas or programs that are more successful than others? How do those differ?

Oh man. I could write about this for hours. Feel free to memail me if you want to talk any more about this. And have fun!
posted by JannaK at 6:35 PM on December 1, 2015

If you can figure out how/where people first heard of you and what precipitated each act of increasing (or decreasing) engagement/support, you will be a hero. You want to be able to map this in terms of physical proximity, temporal proximity, and social proximity (to other donors, to your own events and initiatives, to outside phenomena like holidays and natural disasters, etc.) Basically, anything that has influenced their behavior.
posted by SMPA at 7:02 PM on December 1, 2015

By the way, the temporal proximity thing is potentially very complex. You might want to get an intern to help you collect potentially relevant dates and date ranges. When I was analyzing FMLA usage, we had to figure out things like which pay periods ended "perfect attendance" calculation periods, which school district people lived in and when they had Spring Break, and the times when we had very heavy snowfall. Temporal influences could look like just about anything. You're going to have to try different things to see what actually matters.
posted by SMPA at 7:08 PM on December 1, 2015

What format is this data currently in? Do you already have standard donor reports in place through a 3rd party product? IE are you looking for the standard development metrics, of which you have none, or looking for off piste, interesting ways to interrogate data already in a donor management system?
posted by DarlingBri at 8:48 PM on December 1, 2015

Very good suggestions above.
Also: Include all participation, attendance, volunteer, etc. behavioral data of the donors as well as all non-donors, in the analysis. Consider effects due to actions and activities of the organization itself, as well as nonprofit sector-specific information and trends.
posted by lathrop at 8:49 PM on December 1, 2015

I'm not in the non profit sector but I have been tasked with relatively undirected data analysis. I don't think you're asking about exploration tools (if you are, pivot tables), so I'll contribute a fairly technical point:

If you're doing any kind of quantitative analysis (eg linear regression), you'll typically get more useful results by identifying which of your numbers are stocks/levels and which are flows/rates of change. Compare like with like; if you want to compare a stock to a flow, look at some variety of moving average of the flow. If this is relevant to you, feel free to MeMail me for some more detailed sources.
posted by PMdixon at 10:59 PM on December 1, 2015