Helping you learn and grow in your higher ed career – no matter how you do it.
As I mentioned a month ago just before the High Ed Web New England conference, Higher Ed Experts has chosen to support the 5 regional conferences this year and help them bring great keynote speakers.
Whether or not you get got a chance to attend the High Ed Web Florida conference next week in Gainsville, I also wanted to make sure that you would learn a bit from the conference keynote speaker, Daniel Ryan who was Director of Front-end Development at Obama for America 2012.
That’s why I asked him to tell us more about him, some amazing tips to raise online donations and his keynote.
1) You worked for the Obama Campaign. Can you tell us how you manage to get the job in the first place and what you did?
My journey with the campaign started back in 08. I was working at an agency in Chattanooga, Tennessee, managing the front-end developers there when I got an email blast from the Obama campaign looking for web developers. I applied and had a few interviews, but apparently got lost in the shuffle. There’s been a lot of press about how well we targeted folks with our emails during the 2012 cycle; that’s how they got me the second time around. Late one evening I received an email from the campaign looking for developers once again. The next morning I uploaded my résumé. About 30 minutes later I got a reply back from Kyle Rush, who later became my deputy. Kyle and I set up an interview later that week. A few weeks after that I had moved to Chicago and started working at the campaign.
I started as a Senior Developer but it didn’t take long before I was promoted to Director of Front-end Development. Kyle, Jeff Louden, Michael Renehan, Jason Rico and I formed the leadership of our team as we built up to two dozen staff. Our team was divided along the lines of the main functions of a campaign: fundraising, persuasion and turning out voters.
2) Obviously, the campaign managed to raise a lot of money online. Any lessons learned on the web development front that universities and colleges could use in their fundraising efforts?
Totally. I don’t think we had any secret sauce at the campaign, we just combined lots of smart smaller decisions together. Members of the development team, along with project managers, email writers, designers and advertising experts, formed a working group around optimizing our donation experience. We ran hundreds of A/B tests (using Optimizely.com) comparing various layouts, imagery and language.
If there’s one major takeaway I could give to folks, it is always follow the data. Never make assumptions without something to back them up with. I think the number of times we predicted a testing outcome accurately was far lower than when we got it wrong. We took that to heart and tested everything.
A few specific lessons we learned that I think are applicable everywhere are:
- Speed matters.
We switched our donate pages from a PHP/MySQL backed solution to static HTML hosted on Amazon’s S3. The exact same pages served up 2 seconds faster and we saw a 10% overall increase in revenue. Milliseconds can mean millions. - When your audience changes, run your tests again.
Many things that we tried early on with no significant difference turned out to be major conversion increases as we neared the election and our visitor base broadened. We tested the color of our buttons quite often and found that no color really outperformed another. Until one day in the late summer of 2012 when we did find a difference. If we’d not kept trying that basic test, we would have lost thousands and thousands of dollars. - Photography really sells.
We’d always had a photo of the President on our donate pages, but through testing we found that larger images performed better. Images of the President looking towards the donate form outperformed those where he was looking away. The biggest breakthrough we found was using a photo of the President speaking to a crowd of supporters.
It’s impossible for us to know exactly why this one worked better, but our best theory is that it spoke to our campaign’s ethos of being about our supporters and not just one man. - Don’t give the user choices.
Every time we presented multiple options for the user to choose from, the overall level of engagement went down. In other words, if we had three calls to action, they would get less click throughs combined than a single call to action. The same holds true to the number of inputs on forms. For every field we added, we would measure about a 10% decrease in conversions. There should be no optional fields on your donate form. If there’s a need for additional data (like why a person donated) ask it in a follow up after they’ve made their donation.
3) You’re going to present the keynote talk at the High Ed Web Florida conference.
I’m really looking forward to the High Ed Web conference. I’m giving a brand new talk there called “Human-first Web Design” that focuses on getting balancing organizational needs with the needs of your visitors. This was a real struggle at the campaign and I think a lot of the lessons we learned directly apply to academic websites. Specifically I’ll be discussing navigational structures, responsive design techniques and audience targeting. I think it’s going to be really great.