I have been reading all sorts of posts on the performance of the different campaigns, including Colin Delany’s post on TechPresident about John Edward’s $1M funding surprise in five days and the insurgence of online fundraising after a successful win in Iowa and/or in New Hampshire. Well, I have always had the online stats from those first two weeks after Kerry won Iowa and New Hampshire and thought maybe I could share them with you.
 |
Performance-based Metrics
One of the things we were doing from November until January was trying to unlock the secrets of fundraising and online performance. I spent untold hours pouring over data and online donations to determine the potential success formula. We learned that more pages on a contribution page is bad (went from a 20% conversion rate to an 80% conversion rate), one page minimum on the forms, fewer elements are better (you do not need them to check off each line on the FEC disclaimer) and watched how the conversion rate (contributions/unique visitors and finished contributing/started contributing) improved. But nothing could prepare us for the avalanche that happened on the day after Iowa.
The very next day, as the graph shows, we were having incredible performance on the donations. At the time, I got besieged by one of the communication staffers who asked me for metrics on performance. After a bit, we rushed out numbers and found our press going through the roof. Everyone wanted to know how we were doing. I remember a staff meeting a couple days later where someone from the senior staff commented on the ATM machine finally was working with the Internet. Within a week, we had hit $1M on online donations – web and email.
New Hampshire and Super Tuesday
We were seriously worried about the software doing on contributions since it was relatively cheap and had not been tested under serious load conditions, but when New Hampshire went to JK, I was there all night long watching the server load, making sure we were okay. And, we made it quite nicely. It wasn’t until SuperTuesday that we discovered the fallacy of low-cost solutions in an enterprise world.
The evening that SuperTuesday was announced for JK, we were in a large auditorium and were watching the celebration of JK becoming the presumptive nominee. It was astonishing, and I had someone watching the server to make sure all went well. The very next day, MeetUp.org decided to point their 2M members to our server. According to our software provider, we were assured everything would be okay. It could handle the load. Uh huh.
Within 15 minutes, the server suddenly froze and gave up the ghost. We restarted the server and tried again. Less than 5 minutes and it stopped. Interesting, at the same time, Nicco Mele of the Dean Campaign called wondering how we were doing. We discussed performance – and even commented on the sudden uptime challenges – and he said that their (the Dean Campaign’s info) had always said there was a pent-up energy for donating once the nominee was chosen. I had to dash off, and fortunately for us, we had a backup plan in place to handle the donations – making them a more “batch” process rather than a “real-time” (authorization right at the time of request). Once that system was in place, we easily took in another $1M within 24 hours.
It was an amazing time, and I wonder how Joe is doing over at the Obama campaign right now…
And for the political/data junkies, I offer the following PDF with the graphs for each of the online efforts during those weeks. Click here for the PDF.
Tags: Online Fundraising, eCampaigning, John Kerry, Campaign 2004, Online Fundraising Performance
Pingback: Questions for a Researcher | Political Gastronomica