Less than a week before this year’s Ohio Democratic primary, a local political analyst on WOSU-TV in Columbus laid out the conventional wisdom in the race for the state’s newly-drawn 3rd Congressional District:

“Mary Jo Kilroy is going to win handily for several reasons,” the analyst opined. “I don’t think it will be close.”

To say the odds were stacked against Democrat Joyce Beatty would be an understatement. Beatty is a former minority leader of the Ohio State House who was serving as Ohio State University’s vice president of outreach when she opted for a primary bid against former Congresswoman Mary Jo Kilroy.

Kilroy began the race as the odds-on favorite. She had 94 percent name ID and high favorability, but our numbers revealed vulnerability right from the start. Though Kilroy was better known, she lacked the depth of connection and personal intensity that voters who knew Joyce had. Indeed, Joyce earned a very strong 68-degree personal feeling thermometer on our zero to 100 scale, yet was identified by just a bare majority of voters—56 percent. Kilroy had almost universal name ID, yet her personal feeling thermometer was a less intense 63 degrees.

In an open-ended exercise among voters asking why they supported Kilroy from the outset, we learned that fully one-third of Kilroy’s supporters were with her simply because they were familiar with the name, and that they were widely open to the right alternative if they could find one.

A poll released seven weeks before the primary had Kilroy defeating Beatty by a whopping 34 points—48 percent to 14 percent. And even though Kilroy was widely expected to win the primary, she had still raised millions and was ready to spend it.

In December, the Beatty campaign quickly assembled a team that included media consultant John Rowley of Fletcher Rowley, pollsters Andrew Myers and Lauren Spangler of Myers Research, campaign manager Greg Beswick and direct mail consultant Achim Bergmann of the Baughman Company.

We knew from the outset that Kilroy’s strength wasn’t our only hurdle. After Beatty entered the race in mid-December, the window to pull off an upset actually shortened: the primary election date was moved from mid-May back to March 6. For an underfunded hopeful who needs to come from more than 30 points behind to win, that’s usually fatal.

A Crowded Field in a New District

When Ohio’s new 3rd District was drawn, there was hope that an African-American candidate would have a fighting chance in a Democratic primary given that African Americans made up 36 percent of the primary electorate. But with two credible African-American candidates in the race—Beatty and Columbus City Councilwoman Priscilla Tyson—this prospect looked bleak. In addition to Beatty, Kilroy, and Tyson, state Rep. Ted Celeste (the brother of former Ohio Gov. Dick Celeste) rounded out a field of well-known and well-liked Democrats.

Yet another challenge was rebuilding Beatty’s political organization. During her time in the state House, Beatty was widely respected for her legislative leadership and for helping Democrats move from a 22-seat minority to a legislative majority. She also hadn’t faced a hotly-contested race in a decade. It meant that Beatty did not have an active campaign organization or even good fundraising and email lists when her Congressional campaign began. Putting the campaign organization together would truly be running a “hurry-up offense.”

What we had going for us was our candidate. Beatty had deep ties to the local community and a strong track record as a legislator. A large portion of her former state House district was part of Ohio’s new 3rd Congressional District. Beatty had also worked in the nonprofit sector, started two small businesses and held a high-profile position at Ohio State. She entered the right race at the right time and we highlighted a message that resonated with key segments of the primary electorate. In all of our campaign communications, Beatty was our messenger. Our baseline poll offered a clear path to making it a tight race with Kilroy.

Despite the fact that Beatty trailed by a substantial margin at the outset, Kilroy lost ground after a balanced and positive message profile, while Beatty gained, leaving the two locked in a statistical dead heat that had Beatty at 36 percent (close to the same share of the vote she received on Election Day).

We were able to close the gap with Kilroy because the environment favored Beatty and her message. The frustration of the electorate was palpable in our polling—voters were tired of the status quo and did not want a candidate who would “go along to get along.”

We tested two slightly different Beatty profiles that showed a crystal clear path. The best profile honed in on the fact that Beatty would not “go along to get along” and would stand up to anyone in either party to create and protect Ohio jobs. It was the springboard for our entire media message. Our open-ended polling questions also revealed that Kilroy’s initial support was only skin deep. Five weeks out, it was clear we could make up ground quickly, but we needed to get on the air just as quickly.

The numbers suggested the attacks on Kilroy lacked salience. At the same time, we found that Columbus Mayor Michael Coleman was a powerful endorser for Beatty and was as popular with this primary electorate as any public official you could imagine. Coleman had nearly universal name recognition and his favorable number was around 80 percent.

The ‘Long Ball’ TV Buy

Our data showed Beatty’s momentum could snowball, but only if we made a splash quickly. So within a few hours of getting our benchmark poll analysis our media strategist John Rowley came up with an idea for a spot that would connect emotionally with primary voters. It featured Beatty delivering a speech in which she emphasized that she wouldn’t head to Washington and simply “go along to get along.” Her message was focused on fighting for jobs, standing up to the Tea Party and protecting Social Security and Medicare. At the end of the spot, we highlighted the Coleman endorsement.

We paired the first TV ad with an African-American radio spot that featured Mayor Coleman saying “We need change. We need a voice. We need Joyce.” As endorsers go, Coleman was everything we could have hoped for. And in an era of soft and often meaningless endorsements, it made a difference. He publicly endorsed Beatty, hosted fundraisers, mobilized volunteers, did robocalls, spoke at events, and even personally phoned voters.

It was only week five, but we spent what would end up being almost two thirds of our communication dollars with this initial buy. Our plan was to positively define Beatty, make up significant ground on Kilroy, and then do a tracking poll that we hoped would show real movement—the results would then help us fill Beatty’s campaign coffers to fund the final phase of the campaign.

The question of timing the initial media buy was a critical one for us. Come Election Day, we knew Ohio’s 3rd District Democratic primary would barely make headlines given the tight Republican presidential primary contest between Mitt Romney and Rick Santorum. We also knew that the presidential candidates and their allied Super PACs would soon begin causing clutter on the state’s airwaves creating an additional challenge for getting Beatty’s message out.

Our media buying team was confident we would get more affordable TV time by going up in week five given that there was less demand. It would also be weeks before the Super PACs starting advertising in the GOP primary.

In fact, even though Kilroy outspent us by about 20 percent on TV when all was said and done, we delivered the same number of gross rating points (GRPs) given that most of our buy aired before Super PACs drove up the TV costs. Almost all of Kilroy’s network TV buy was in week one as the Super PACs flooded the airwaves. The Beatty camp went up on network, cable TV and radio on Feb. 3. Kilroy wasn’t up on cable until Feb. 13, and she wasn’t advertising on network until Feb. 21. Our strategy worked.

The next tracking poll showed us surging from 33 points behind to just a single point back of Kilroy with 10 days to go before the election.

We had also made significant gains with our primary targets—white women and African Americans. Meenah Hulsen and Kristin Pixley, our media buyers at the Davis Group, were able to maximize a thin media budget (Kilroy outspent us 2-to-1 on cable) and hit our targets by zeroing in on cable zones with higher numbers of white women and black populations, and by using cable networks that resonated with our primary targets. We were also the only campaign to run sustained radio advertisements targeting African Americans.

Our media strategy made it a real race with Kilroy and our ground game took it from there. The final challenge was changing the turnout model on a limited budget, and our team led by campaign manager Greg Beswick, former state Rep. Otto Beatty (the candidate’s spouse) and Mayor Coleman’s political team executed a masterful and underfunded plan to exceed expectations on absentee and early voting. The campaign exceeded its persuasion goals in minority precincts and even garnered 30 percent in white majority precincts.

We also broke the mold and targeted so-called “drop-off” voters even in an early March primary. The campaign targeted 12,000 voters who previously voted only in even-year general elections, those who voted in last year’s collective bargaining referendum, and Democrats who cast ballots in the state’s 2008 presidential primary between then-Sens. Hillary Clinton and Barack Obama. We knew we needed to energize 10 percent of that vote to drastically change this election, and it made all the difference in a 1,400-vote win.

Guided by the results of our benchmark poll, we ran a positive campaign—not the typical tack taken by a candidate who’s down by 30-plus points in early polling. Rather than employing harsh negatives in an attempt to close the gap, we presented a subtle contrast by telling voters that we need a “new voice ... Joyce,” as opposed to Kilroy who had been in Congress and had run for office for 20 years.

The bottom line was that we were up against a Democrat with a rock solid progressive record in Mary Jo Kilroy. The concern was that attacks on Kilroy could backfire and only serve to energize more of her supporters, so we focused on soft contrasts and making the most of our limited media budget.

In the end, it was fitting that a win in the district where Ohio State football is king featured a hurry-up offense, a well-timed long ball and an effective ground game.

Andrew Myers is the president of Myers Research a national public opinion research firm. Kati Bumgardner is director of media and research at Fletcher Rowley, Inc.