Yes, market size really does matter!

By November 2, 2016 July 25th, 2017 Cool Analysis, Philanthropy

Peer Performance Insights: A New Perspective on Philanthropy and Fundraising
Post #3: Understand Your Markets, Part Deux:
Yes, Market Size Really Does #@$#@ Matter

In our first post we told you why we believe we need a new perspective on philanthropy and fundraising. In the second post we showed you that philanthropy is local, and all markets are not equal. “So what?”, you said. “Does it really matter?”, you said.

“Yes, Market Size Really Does #@$#@ Matter,” we say. Here’s why:

One-third of what you raise is determined by where you are, or where you plan to go.

Think about that for a moment…  one-third for market size ALONE. We aren’t (yet) looking at other market factors such as demographics, socioeconomics, culture, ratio of Red Sox to Yankees fans, and the implications of deflategate on local philanthropic giving. Nor are we (yet) looking at important organizational factors like, I don’t know, your mission, strategy, how many fundraisers you have, or how good they are! That’s a little crazy, no?

Crazy but true and we’ll show you how we arrived here… but a few notes before we do:

  • We are about to seriously geek out a bit, so brace yourself. If you love scatter plots and regressions, get ready to have some fun. If you are a normal human being and just want the bottom line before the Walking Dead comes on, bear with us… this is important. If you are busy and/or overconfident in our abilities, skip the steps and go to the “so what”.
  • For those of you who are serious geeks, know we recognize we are using incomplete, imperfect data of variable quality and comparability; and we are playing a little fast and loose with the laws of statistics. Below we make the case this is ok – we are helping folks make the best possible decisions based on available data, not proving the existence of the Higgs boson, but welcome your feedback and thoughts to the contrary!

Let the geeking begin!

Step 1: First we take the data we showed in our last post – Philanthropic market size by metropolitan statistical area. See Figure 1 for a snapshot of the top 50 markets from big to small (note, we have this for all 381 metropolitan areas, but that makes for a tough graph to read). This data is going to become our X-axis in a moment, so hold it in your head.

 

Figure 1: Local Markets are Not Equal

Step 2: We collected philanthropic revenue data for a set of 13 multi-site organizations raising money in multiple metropolitan areas. Specifically, we obtained philanthropic revenue raised by each organization within each metropolitan area. For example, Nonprofit A raised $11.4 M across 8 metropolitan areas in the following amounts: $4.3 M in New York, $0.6 M in Chicago, so on and so forth. Keep this in your head as it will become our Y-Axis in a moment.

Step 3: The moment is here! We create a scatter plot with each of Nonprofit A’s sites becoming a “dot”. For example, Nonprofit A’s New York office would be ($27.4 B, $4.3 M), and their Chicago office would be ($10.9 B, $0.6 M), so on and so forth for the remaining six markets. This gives us Figure 2, a plot of Nonprofit A’s revenue by market for each of its 8 offices.

 

Figure 2: Philanthropic Revenue by Philanthropic Market for Nonprofit A (Observed)

 

Step 4: We performed a linear regression analysis to quantify the relationship between revenue and market size as you can see represented as the line in Figure 3. We’ve included two stats in the label, r-squared and p-value. The former, r-squared is a measure of “fit” between the observations (dots) and the predicted values (line) and indicates, for this example, market size “explains” 69% of the variation in revenue for Nonprofit A. The latter, p-value, is a measure of significance, and suffice it to say anything below 0.05 is solid, so we are fine at 0.01.

Here’s where I will again highlight the limits of this data and analysis… I wouldn’t call this capital “S” science. Data on philanthropy can be sketchy, compiled from various sources over various time periods, and we’ve made assumptions to fill in the gaps. Our approach is empirical in nature, but we also rely upon our judgment and experience to make some analytical leaps that might make some cringe.

For example, we need to be careful in how we think about correlation and causation. We know the world isn’t so simple as to say “big” automatically equals “raise more”. That said, this feels mostly true or at least highly related, matches our intuition, and that of a dozen or so nonprofit leaders with whom we’ve road tested our thinking. So, we make the case “Good Enough!”

 

Figure 3: Philanthropic Revenue by Philanthropic Market for Nonprofit A (Predicted)

 

Step 5: Rinse, wash, and repeat steps 1 through 4 for the other dozen organizations.

So What?

Once we’ve done this for our entire sample, we find the following:

  • All thirteen organizations have a significant correlation between revenue and market size
  • The correlation or r-squared for the sample ranges from 10% to 75% — a big range!
  • The mean and median r-squared of the sample are both ~33%

It is from this last bullet that we derive our rule of thumb:

One-third of what you raise is determined by where you are, or where you plan to go.

This insight can shine a powerful light on philanthropic potential and performance for markets in which you currently operate, and new markets to which you are considering expansion; particularly when you combine market data with benchmarking data from local peers.

We’ll start to explore the “how to” of using this data to make expansion and growth decisions for all your markets, and to set and assess revenue targets within individual markets in the next set of posts, coming next week, same Bat Time same Bat Channel.