The velocity of asks.
It’s Friday night and you receive a call from your local food bank during dinner. They’ll chat about a few things that feel relevant and interesting to you and then make a pitch. You’ll decline. Something in the pitch felt off , despite the friendly tone and personal touch. Later you’ll realize it wasn’t a human but an AI bot. They had your facebook information and your giving history. By the end of the week — you will have had 50 of these.
Versions of this nightmare are already happening in major philanthropy. AI-powered grant writing sends thousands of grant letters out every day. Soon, there will be customized pitches that feel very real, designed beautifully, and delivered to your inbox. The velocity of the asks will be unmanageable. If attrition of major donors was bad before, welcome to the new future absent an intervention.
We’ve arrived at a pitch-and-be-pitched model of philanthropy that is wholly unhelpful if we’re going to get serious about solving real problems.
- My age
- My skin colour
- My geographic location
- My social skills
- My looks
- My connections
All benefit our fundraising by at least 50% more than someone with a different genetic, geographic, and social makeup.
Moneyball was about discovering helpful proxies, and building a team around the overlooked metrics that truly reflect a player’s performance over the long haul. Helpful proxies.
“Just because someone interviews well or is friendly doesn’t mean they can really perform the work that needs to be done.” Adam Grant has written extensively about these concepts in organizational psychology. And yet, here we are, with an outdated interview system still in the HR world and still in the fundraising world.
Given that we can now index, sort and scan vast swaths of data regarding what a charity does, how their leadership team is composed, their financials and more with AI, what proxies might be helpful to determine what type of doers of “good” we want to bet on?
If we like their charm and charisma, let us call it for what it is and realize that it may not lead to better outcomes for the people being served. If we give to who we know because we “know” 😉 we can trust them, we know we’re excluding those who statistically would likely be better able to perform the job but are barred from participating due to barriers in fundraising.
But now that we can measure so many things (like program performance metrics), the easy-to-measure stuff like fundraising ratios (which effectively depend on the charisma and talent of a few) and staff count is a hurtful and expensive distraction.
As AI agents soon make traditional fundraising a complete cluster___, how might we flip the model to use tools that could better measure real indicators of the change we want to see in the world rather than perpetuate pitched based system that favours false proxies over real ones.
//
PS I’m building WellFunded.io — Matchmaking for Major Philanthropy. I’d love to have you come along for the ride. Visit the page and join the wait list today.
PPS Seth Godin has written a lot about Proxies — check out his blog here.
