In this week's edition, we circle back to donations analysis to explore an oft-asked question.

A Word From Our Attorneys This entry is meant to aid ROI calculations for marketing, not for purchasing e-mail addresses. DIA users in particular should be aware that our opt-in policy prohibits mailing a purchased list from a DIA account.

Last month, the Tipsheet's three-part series (I, II, III) explored timing the release of your donations appeal. Today, we're getting a little bit more abstract. Using the entire year's collection of gift data from throughout the DIA universe, we're going to try to get at what a supporter might be expected to contribute ... and as a result, what they're worth to obtain in the first place. Numbers on the flip. To tackle this measurement, we're going to do basic arithmetic: (Number of Donations/Supporter) x (Amount Donated/Donation) = Amount Donated/Supporter Or in words, we're going to find out how many times the typical supporter donated over the course of a year, and how much such a donation was typically worth.

1. Donations Per Supporter

To get this number, we're going back to the same data set that generated the "propensity to donate" averages for our seasonal donation ratings. Those scores, you may recall, are a count of donations per 100,000 supporters among the organizations actually using our donations tools. I originally reported them as weekly and monthly medians to reduce day-to-day fluctuations to more representative scores, and I did so after removing some gravity-skewing major fundraising pushes by a couple of our largest organizations. In this version, we're going with the unadjusted original data, because rather than trying to get at some "underlying" seasonal propensity, we're trying to get an overall yearly propensity that would, of course, include donations made during fundraising pushes. So what we're doing here is simply adding the propensity scores for all 365 days.* A straight sum of all daily donations-per-100k-supporters gives us 1,990.72 expected donations for every 100,000 supporters -- or 1.99 per hundred. Since there's some interpolation on our supporter growth numbers, let's not pretend three significant digits of precision and just round it off to 2%, otherwise stated as a simple rule:

For every 50 people on a list for one calendar year, you will receive 1 online donation.

You could in principle reduce this to the idea of 1 donation for every 600 (50 * 12) supporter-months, but be careful: since the seasonal spike at the end of the year is so enormous, it's much more valuable to have people on your list in November and December than in July and August. More useful for our purposes will be to say that each supporter is worth .02 donations per year, on average.

2. Amount Per Donation

Now we need the second term in the equation -- the average donation. This is a very simple calculation: the mean gift of some 80,000 online donations we handled last year was $86.57. As with our previous number-crunching, these exclude paid event registrations, storefront purchases, instances of recurring gifts beyond the original payment, and offline donations loaded into the system, generally because they're much less easy to generalize as revenue streams from organization to organization. A true accounting of the value of a supporter cannot help but look at these other sources: an e-mail subscriber who sends a $50 check contributes no less than one who donates that amount online. But it's at this point that our general data will need to yield to each organization's assessment of the relative importance of online donations to its overall development picture.

3. The Calculator Says ...

That important caveat aside, our result is that in terms of online donations only, the average base value of a supporter is .02 x $86.57 = $1.73 per year. That's the floor, not the ceiling. Think about it that way the next time you're tempted to stuff the last event signup page into a desk drawer instead of typing in the names. For a more thorough calculation for people getting ready to wade into cost-per-click and conversion ratios, there are a lot of directions possible, but the obvious next steps to calculating a more complete ROI using an organization's own internal data are threefold:

• For the full value per year, add in the average value per e-mail supporter in offline donations, purchases, event registrations, and any other revenue stream;
• For lifetime value, find an attrition rate and/or average lifespan for an e-mail address, e.g., "the mean e-mail subscriber remains on our list for 8/18/48 months".
• Then, of course, find the cost per acquisition (expected or empirical) for whatever outreach program you're considering/judging.

Get out the slide rule and go to it, Poindexter. *The reason for using these propensity scores rather than straight donation counts is that they're already adjusted for the growth over the year of the burgeoning DIA universe.