Article as it appeared on CampaignsandElections.com
By David Mowery
The simmering tension between traditional media consultants and their new media counterparts is a popular panel topic these days. Given a chance, most consultants will yap for hours about the fight between the digital folks and the TV and radio folks. In reality, this conflict isn’t really a consultant problem. It’s a campaign management problem.
Take my recent experience with Todd Strange, who serves as mayor of Montgomery, Ala. Strange earlier this year decided to run for a third term and tapped me to put together the campaign team. I managed his successful 2009 race and since then, like many in the industry, I’ve been trying to keep up with ever-changing digital strategy.
The good news was that the mayor was open to new ideas, but he also made it clear that he wasn’t running again to lose. I have been able to use his campaigns in the past to test tactics that other candidates wouldn’t have the time, budget or vision to try.
Back in 2009, we built our own database of voters, and used email and phones to attempt to have a hard count in an election of 39,000 people. While we never got a total hard count, we did best a seven-person field without a runoff, despite another candidate getting the endorsement of Bobby Bright, who’d left the mayor’s office when he got elected to Congress.
In the less-competitive 2011 race we released an intro video designed to dissuade anyone considering entering the race. On Election Day, Strange won more than 80 percent of the vote.
This year, I wanted to try to figure out the best and most efficient use of old-media-versus-new-media strategies and tactics, and what the proper split in terms of communications dollars would be. The industry seems to have settled on 20-25 percent of the budget going to digital, which doesn’t really tell you anything. The first thing you have to decide on is, who you’re trying to reach, how they’re best reachable, and how effective each medium is at reaching your targets.
For Strange’s 2015 campaign, we started with an overall budget of around $750,000. That was eventually whacked down to $450,000, although we actually spent around $550,000. If that sounds crazy for a town of less than 200,000 people, consider that one of our opponents, former Congressman Artur Davis, spent close to $700,000.
Here’s how the budget mainly broke down: $140,000 went to TV, $40,000 was spent on mail, $25,000 on phone contacts, $5,000 on billboards, $5,000 on local advertising and probably $3,000 on the radio. We also spent $60,000 on Facebook, Pandora and cookie-targeted ads.
But here’s the key: we chose our vendors, decided on a turnout model, got TV and radio rates, tested message delivery online early, and identified our win number and how to reach it before anyone else had announced.
I bought my data from L2 and used VoterTrove as our VoterCRM vendor. While I have worked with (and will work with) other data vendors in the future, I knew I needed hands-on and immediate response when I had a question, as most of my staff consisted of business associates of the mayor and field operatives, not typical campaign staff. I did hire a chief technology officer to sort and tag the data, send emails, and troubleshoot. This was essential as a lot of the day-to-day data entry doesn’t need to be done at senior staff level.
The rest of the consulting team remained unchanged from 2009: Ourso Beychok did the mail, and Chism Strategies did the phones, while a local firm handled media production and placement. I kept the digital: targeting, social media management, and content creation in house. I worked with several local creative firms to design and help me produce the shareable content every campaign should be using in 2015.
The misnomer that TV and digital are in competition with each other is laughable – if done correctly they should work in concert to drive messages to different audiences using similar creative. Facebook and cookie-targeted ads can show as much as a one-third boost in awareness of your advertising. You can’t have one without the other.
We produced a total of five spots: one longform narrative and four highlighting our topline issues from our polling, and the last a blatant play for the solid conservative voter, featuring the candidate’s grandchildren.
We released the longform intro at our kickoff fundraiser, then emailed it to our contacts. Afterward, we targeted our most likely voters using only Facebook. We drove 132,231 impressions to 32,567 unique people with 4,617 watching past 30 seconds. Video views cost us a whopping $0.13 per view and more than 2,000 people identified as likely No.1 voters for roughly $4,200.
We also placed an ad on Pandora, along with banners and audio during this initial phase. We were able to secure roughly 2.1 million impressions to our target audience over two months.
After we were driving views and running Facebook ads for about a month, we conducted a poll. The Davis campaign had spent about $130,000 on TV and radio during the same time period. Our poll said 45 percent of our target voters had seen the Davis commercials on TV, while 28 percent said they had heard Davis commercials on the radio. Moreover, 28 percent said they had seen the Strange ads on online. The cost differential was 31-to-1.
As the campaign went on, we released each ad on Facebook, targeted to our likeliest voters, and made and tested banners and memes targeted to more granular audiences. We also had good success with a button contest, which was actually a grab for voter information. We were able to drive some 1,000 sign-ups via our website for the button contest — all people we were able to add to our No.1s list of people voting for our clients.
With two weeks left we partnered with VoterTrove to serve cookie-targeted ads to our most likely voter targets, and to unreliables who we thought might vote for us.
One of the bigger things to keep in mind is that while all media is used to persuade, two forms give you a chance to identify and interact with voters. The first is phones, and the second is social media. We used Facebook as a medium to gather data. We scored voters in terms of likelihood to vote, and then used their online behavior to tag them in our database as likely to vote for us, or likely to vote for one of our opponents. Respondents who said something like, “I’m voting for the mayor” were tagged as 1s, whereas those who commented “we need someone else” were tagged as zeroes and excluded from our GOTV contacts pool.
On Aug. 25, Strange was reelected with 57 percent of the vote. Davis, despite his name ID from his years in Congress and running for governor, took 27 percent. We found the right digital-TV mix to avoid a runoff and secure a third term in office for our client.
David Mowery is the founder and president of Mowery Consulting Group.