M+R had a fascinating post* last week about political fundraising. It highlighted insights from the digital team who sent out fundraising emails for the Obama campaign. While political fundraising is its own animal, I do think many of these insights apply to all forms of fundraising. So whether you’re a political activist or a nonprofit fundraiser, or of the red or blue or purple persuasion, you will find this interesting.(The whole post is here. These are some highlights along with my commentary.)1. It’s hard to predict what will work – so testing matters. There were 18 very smart people on the email team alone, and they often predicted the wrong winners among versions of emails. And just when they figured out what worked, it stopped working. So they tested again. Keep testing!2. The best segmentation was based on what donors did – not how they voted or their demographics. Segmenting their message according to the ways people responded worked far better in yielding strong fundraising results than any other variable. What have people donated in the past? In response to which appeals? Segment accordingly.3. Length didn’t seem to matter a lot, until the end of the campaign, when shorter did better (reminds me of my advice to write very short appeals on December 31!). What did matter was the content and relevance of the message.4. For fundraising, setting a big goal for number of donations worked, but little, very local goals (we need six more donors in Washington, DC) did not. Those only worked for advocacy. Interesting. Something to test?And my favorite finding? The best appeals also had the highest unsubscribe rates. Like Mark Rovner always says, evoking passion means you get strong opinions on all sides. Bland is safe – and gets NO reaction.For more findings, check out the full post, “Surprises from Obama’s New Media Staff.”*Hat tip to Jono Smith of Event360 for sharing the post.