For the past few months, I’ve been knee-deep in a local political campaign. We were hired to do social media for a dark-horse candidate running for mayor of my hometown, Atlanta, GA. “My” candidate was barely a blip on the radar when we jumped in to create a social media presence in mid August. Yes, we were late-to-the-dance, very late, but we always love a challenge! There were six viable candidates running for the non-partisan (non-party affiliation) office. Our client ranked #4 and had virtually zero social media presence. Of the three leading candidates, one had a very sophisticated SM presence, one was so-so and the third was mostly non-existent until the last few weeks before the November 3 election.
I will spare the suspense and tell you that my guy finished fourth out of six. The eventual winner (following a runoff held on December 1) was the candidate that had the best SM program. Did that win the race for him? Probably not – it was plain old “get out the vote” activity that caused more people to vote in the run-off election than in the regular election, BUT, the difference came from communities in Atlanta’s “intown” neighborhoods that common sense would say are among the most “wired” in the city. Social media was most likely a factor in these key neighborhoods.
What I learned from this campaign is interesting, and as far as we can determine, broke some new ground in the use of social media in local political efforts.
When we started this project in August, the eventual winner of the election already had the maximum 5,000 Facebook Friends on his personal FB page and had several thousand more on a “fan” page (ironically, this candidate polled in the single digits until just a few weeks before the election and was running a distant #3). The candidate that was polling in the #2 position had about 2,000 FB Friends and the candidate that was leading in the early polling had a SM presence next to nothing – about where our candidate was when we started.
We knew that we had to be aggressive – that we would have to be proactive in building a Facebook and Twitter presence. We would have to actively recruit “friends’ and “followers” rather than waiting for them to come to us. While our original plan called for a blog, the campaign communications people were not up to the task of keeping it up-to-date, so we eventually focused on Facebook as a message platform and Twitter to drive potential voters to the Facebook page.
Facebook worked well as a platform as we could attach videos, links to newspaper and magazine articles, links to political blogs, photos from events, etc. and we could do instant updates via an iPhone.
We met with a political consultant to get the lay-of-the-land in terms of who the key demographic / psychographic profiles elicited those most likely to vote. Atlanta is a city that is about 55% African-American and 45% White / Latino. Historic voting patterns in Atlanta city-wide elections showed this ranking of those most likely to vote:
1. African American women; 2. white women; 3. white men; 4. African American men; 5. Latinos.
We decided to concentrate our efforts on African American and White women, the top two “most likely to vote” groups. This was based on the concept that we had little time and few resources and needed to get the biggest “bang” for our effort. Furthermore, we knew that those aged 35 – 60 were more likely to vote than those younger or older.
We identified those women who fell into our target age group among the current 100 or so Facebook Friends that our candidate had in his small network when we started. Then, using the concept that most of their “friends” would be people like themselves, we went to these women’s “Friends” requesting that they become a friend of our candidate. We included a personal message stating that our candidate was running for mayor and attached a link to the candidate’s website. Since this was a City of Atlanta election, and only residents could vote, we did not pursue “friending” anyone who’s profile showed that they did not live in “Atlanta.”
We really worked this, and a small staff of three spent hours and hours “harvesting” friends in our target population from new “Friends” as they came on board. This was a laborious process – click on the profile of the potential “Friend,” copy and paste the personal message and click “send.” Even though we were using broadband and other fast connections, there was a lag as Facebook’s software processed the request.
Next, we knew that just because you post a message on Facebook, not all of your friends will see it. We determined that key policy statements should be posted multiple times in an attempt to have it impact the maximum number of people. We combined this with user info as to when Facebookers are most likely online. We focused on three dayparts: 1) early morning / early at work to reach those who log in to their Facebook accounts at home in the morning or as soon as they get to work; 2) late afternoon to reach those who may log in to their FB page while waiting fro the workday to end; 3) evening to reach those who access FB from home between 8pm and bedtime. We would send a specific message “early morning” on Day 1, “late afternoon” on Day 2 and “evening” on Day 3. Time sensitive messages would be sent three times on the same day at the specified dayparts. Of the hundreds and hundreds of FB messages that we sent to thousands of people, we only received on complaint regarding multiple messages.
We also used Facebook applications to personalize our message. One of these was the “birthday” feature. A list of “Friends” upcoming birthdays (today, tomorrow, the day after tomorrow) appears on the right side of your FB Profile page. Each day, we would send personal birthday wishes to each Friend whose birthday was tomorrow. Although we used one of 5-6 standard messages that we created, we always used the recipient’s first name so that the message was more personal.
One of the most effective uses of Facebook was to pose questions or solutions to issues being raised during the campaign. This was most effective when our candidate’s Friends passed the 2,000 mark, as we had enough Friends in the network to elicit comments. We would create on-going discussions and have our candidate participate by making comments to keep the conversation going. Actually, our candidate was not doing this personally, it was done by us with specific input from the candidate’s senior staff.
We would also monitor the live News Feed and, when appropriate, our candidate would post comments to on going discussions. At times, we would use the “chat” feature to contact Friends one-on-one.
As our network grew, we received the ubiquitous requests to join all kinds of groups, support all kinds of causes, attend all kinds of events and become a fan of all kinds of people. Our policy was to NOT accept any of these invitations as they could become an “issue” in a political campaign, since we did not know anything about these organizations. I cann0t stress how important it is to monitor this activity! We learned a lesson when we accidently ”Twitter Followed” a person in the “adult” business – who we immediately unfollowed!
By the end of the campaign, our candidate had the second largest Facebook and Twitter networks of all the candidates. We believe that had we had the opportunity to get involved with this candidate much earlier, we would have made a bigger difference. The winner of the election was the candidate that had the best Social Media presence, and defeated, in a run off, the candidate with the least impressive SMM effort. Draw your own conclusions.
Political candidates must realize that they can often reach as many people in a social media group with multiple messages much more efficiently and cost effectively than with traditional media.