Five Web Analytics Lessons from the Obama Campaign Coakley Should Have Heeded
Martha Coakley’s failed Senate run has led many to review the key differences in strategy between her campaign and Scott Brown’s. It’s pretty clear that Brown took the web seriously and Coakley seemed to largely ignore many of the social media and web analytics strategies and tactics that swept Obama into office.
Ad Age and Social Media Today have taken a look at (and criticized) Coakley’s overall digital and social media’s approaches. This post narrows in and serves as a reminder of web analytics best practices which can mean the difference between coming in first or a distant second.
The Obama campaign tracked digital metrics obsessively yet ingeniously with great success. The techniques they used are just as applicable today as they were over a year ago.
Dan Siroker was a Google employee, who in December of 2007 decided to take a leave from Google and assist the Obama campaign in setting up a web analytics methodology. He was there for a month, came back to Google for a stint, then in July quit permanently to join the campaign full time through to the general election as director of web analytics.
As a measure of their success, of the $656 million raised by the Obama campaign, $500 million of that was raised online compared to $201 million the McCain campaign raised overall. ($201 million does not include the federal campaign funds McCain accepted).
In November of 2009, Siroker gave a presentation at Google about his experiences and the campaign’s web analytics wins. In it, he outlined five key lessons learned, which are excellent guidelines for any web analytics initiative be it for a political campaign or business marketing.
Lesson 1: Define Success with QUANTIFIABLE Success Metrics
Setting goals such as, “drive traffic via PPC” or “use the email channel to maintain a dialogue with our audiences” are not specific enough to gauge success.
Instead, Siroker’s team set quantifiable goals related to cost per click, signup rates, money raised per email recipient, and many more. He said, “We always had more to do than we had man hours to do it.” As such, having very targeted goals was essential to ensure time expenditures could be justified.
Lesson 2: Question Assumptions
The team utilized Google Website Optimizer to determine which landing pages and creative content pulled its weight. As an ad-hoc survey, Siroker asked the audience of a series of images and calls to action, which combination they thought was best received by splash page viewers and resulted in the most sign ups. It turns out the audience’s assumptions were wrong (so were the analytics team’s originally) and it required multivariate testing to get to the truth (“Learn more” and the “Family Image” were the winners).
Getting the version right meant over 4 million more people signed up for the email list between the time they ran the experiment and the election!
Lesson 3: Divide & Conquer
This lesson demonstrates the power of segmentation. As Avinash Kaushik puts it, segment or die! Data takes on a whole new meaning when you can know as much as possible about the people who generated it.
This was illustrated by the team’s donation button experiment. In it, they sent donation solicitations and tested different donation button verbiage (Donate Now, Please Donate, Why Donate, Donate and Get a Gift, and Contribute) with each audience (never signed up, signed up but never donated, previously donated). The results varied significantly based on who the visitor was when they got there.
Winning combinations:
- Never signed up: Donate and Get a Gift
- Signed up, never donated: Please Donate
- Previously donated: Contribute
Monetary results: tens of millions of dollars in additional donations.
Lesson 4: Don’t Reinvent the Wheel
This lesson goes to show that it’s not how much you spend on technology; it’s how much you invest in talent. One would think that a presidential campaign which broke all sorts of war chest records would use its means to select the most robust and most expensive analytics tools on the market. In fact, Siroker’s team used many free and open source tools. These included Google Analytics, Website Optimizer, Trends, and App Engine, TextMate, Open Reports, and DbVisualizer.
Lesson 5: Take Advantage of Circumstances
After Sarah Palin made a remark at the Republican National Convention in an attempt to undermine the perceived value of being a community organizer, the Obama campaign immediately sent out an email to their list protesting this notion.
The recipients responded. From that one email the campaign raised $10 million.
Think of your own organization. How close are you to putting these lessons into practice? Do you have examples of challenges or success stories?
Don’t pull a “Choke-ley” and underestimate the power and benefits of digital marketing, social media, and web analytics. Know where you stand before you’re passed up by a Brown blur.
Interactive Industry News & Events, Internet Marketing, Social Media, Uncategorized, Web Analytics, Web Strategy Consulting
Great review of Siroker’s campaign video and important lessons for all web analytics practitioners. Unless we can not measure it, our success and failures would be as good as, or as bad as of anyone else. Thanks.