Once a Republican bastion, registration changes in the city of San Diego now more closely mirror the overall trend throughout the state of California. Ahead of the recent mayoral race in San Diego, party registration in the city gave a heavy advantage to Democrats. The registration breakdown: 40 percent Democrat, 26 percent Republican and 34 percent other.

With the election of Democratic Mayor Bob Filner in 2012, unions and other liberal groups determined that demographics were destiny, and that the city would default to Democratic mayors in the future. Mayor Filner’s scandal and subsequent resignation offered Republicans an opportunity to prove them wrong in a November 2013 primary and February 2014 run-off.

Once Filner resigned under heavy pressure from members of both parties, the center-right quickly consolidated support around Republican San Diego Councilman Kevin Faulconer, a socially moderate, fiscal conservative who represented a heavily Democratic coastal district. Faulconer immediately geared up his campaign team, which included our firm, REVOLVIS, as general consultant.

Campaign manager Stephen Puetz, pollster Competitive Edge Research, and fundraiser Golden State Consultants also came on board, while CRAFT Media, Magma Creative and Monument Communications produced the TV spots. Smart Media Group did the buying.

Labor and the Democrats, certain they would take the seat one way or the other, ended up splitting their support in the primary. Half of the Democratic elected officials and many of the city unions stood behind Nathan Fletcher, a former GOP legislator who in the span of just over a year ran for mayor, changed his party affiliation to independent, lost a race for mayor and then changed his party affiliation to Democrat.

Other elected officials, the Labor Council and the Democratic Party stood behind David Alvarez, a 33-year old first-term councilman who represented a heavily Latino Democratic district.

It was likely that only one of the Democrats would survive the primary to battle Faulconer. Fletcher started out with a significant lead, and was viewed as the more formidable run-off candidate against Faulconer. As such, the Faulconer strategy was simple: face either Alvarez or a hobbled Fletcher in the run-off.

Thanks to contribution limits, third parties and outside groups played heavily in the campaign. Faulconer allies included the Chamber of Commerce and the pro-business Lincoln Club. The campaign was also able to directly coordinate with the local Republican Party. Fletcher had an independent expenditure committee and labor. Alvarez had unions and the Democratic Party, which he could also directly coordinate with.

Fletcher faced withering attacks throughout the 45-day campaign. Alvarez and his allies moved to the left, with advertisements geared towards capturing more liberal voters. The Faulconer campaign geared our positive advertisements toward the center, some featuring popular former Mayor Jerry Sanders.

Turnout in the primary was 37 percent. Without a real voter base, Fletcher came in third with 24 percent, winning virtually no precincts. Backed up by massive labor spending, Alvarez came in second with 27 percent. And having faced minimal negative campaigning, Faulconer led the pack with 42 percent.

The Run-Off

The unions and Democrats quickly united around Alvarez, counting on a 14 percent registration advantage. The conventional wisdom suggested that 1) most of Fletcher’s voters were Democrats and independents, and would move to Alvarez, 2) Faulconer’s vote was artificially high since voters had not been exposed to negative messages about him, and 3) union ground operations were superior across the board.

We believed we could be competitive with Fletcher voters, who were mostly moderate. Given Alvarez’s overtly liberal message, we believed we could lock up 90 percent of the GOP and concentrate resources on moderate Democrats and independents. To win these segments we would need to make sure that Faulconer was viewed as nonpartisan, good government, moderate—not the caricature of an extreme Republican that Democrats love to run against.

November public polling put both candidates in the mid-40s, separated by a point. Our internal polling put the race at 47 percent for Faulconer to 43 percent for Alvarez. Both candidates had plenty of soft support and with remaining voters looking more like Alvarez types.

These considerations led to a three-pronged strategy: 1) play on their turf, 2) own the center and 3) turn out the base. The Alvarez and union campaign strategy was fairly clear: motivate and turn out so many progressive voters that we would be overwhelmed.

Play On Their Turf

San Diegans tend to divide the city by Interstate 8. North of the 8, voters tend to be older, whiter and more Republican. South of 8, voters tend to be more diverse, younger and considerably more Democrat. Historically, Republicans do very little physical campaigning in the Asian Pacific Islander (API), Latino, black, Hipster/Urban Pioneer and LGBT communities that dominate south of 8.

We decided to change that, both to win non-traditional votes and to keep the Alvarez campaign defending their own base neighborhoods. In major executive races earned media often means as much or more than paid media, so we planned to focus our press shop on this aspect of the campaign, and to physically run a ground operation in those communities.

Black community leaders largely supported Fletcher in the primary. We immediately began to meet with them and committed to being different Republicans: With their support we would open a headquarters, staff it and run an aggressive campaign in their community.

We also found we had more policy issues in common than any of us might have thought. Several of these leaders endorsed Faulconer, and provided the volunteers we needed to be competitive.

Similar efforts and results were conducted in the Latino, API, and LGBT communities. The campaign rolled out coalitions of independent voters, Democratic voters and former Fletcher supporters. We wanted to ensure that no one would conclude Faulconer was your typical Republican.

The other demographic piece of the puzzle: One-third of San Diego is Latino and David Alvarez represented an opportunity to elect the first Latino mayor in the city’s history. This was a source of pride for the community so we understood we needed to work hard to earn their vote. We launched the first television ad of the run-off on Univision and Telemundo, with Faulconer speaking Spanish to camera. This effort signaled that Faulconer cared about the community and opened the door for a multipronged effort to engage and build support. An outside group, the Hispanic 100, funded pro-Faulconer radios ads featuring former United States Treasurer Rosario Marin on Univision radio, and we ran an aggressive Spanish-language earned media and outreach campaign. The latter included a Latino field program funded by the local GOP, which supported four Latino Spanish-speaking staffers going door-to-door and coalition-building. On Election Day, Faulconer won an estimated 35 percent of the Latino vote.

Own the Center

Among the likely electorate, liberals led conservatives by a small margin, with moderates making up approximately 15 percent. Given the Alvarez/labor strategy of motivating and turning out progressives, and the fact that we would probably lock up conservatives easily, we were aggressive in wooing moderates and soft liberals.

This led to positive campaign themes like “There is no Democrat or Republican way to fix a pothole”—themes that stressed Faulconer’s experience at getting things done over seven years on the council. By comparison, we highlighted Alvarez’s relative youth, inexperience and partisanship.

The other dominant theme was Alvarez’s affiliation with government unions, entities unpopular because many voters thought they had brought the city to the edge of bankruptcy over the previous decade. By the end of the campaign, the unions had pumped more than $4 million into the Alvarez effort—85 percent of all money spent on his side.

We decided to make it clear to voters where Alvarez’s support came from, and lead with the message that he would become a proxy for labor at City Hall. Playing on his turf made it difficult for Alvarez and labor’s paid media to paint Faulconer as the “extreme Republican.”

Turn Out the Base

An analysis conducted after the primary revealed that 20,000 high-propensity Republicans failed to vote. We knew this group needed to be a key target in the run-off, which is where our social pressure strategy came in. Several weeks out from Election Day, we launched social pressure mail, phone, text and field programs. Our primary election focus had been on mid-propensity voters, assuming the high-pros would turn out regardless. But we adjusted our focus to the high-pros in the run-off.

Our social pressure campaign started with mail, and we made a concerted effort to be a bit more aggressive than the current wisdom on social pressure suggests. We sent mailers, with the voter’s name in bold letters, which read, “John Doe, public records show you failed to vote in the primary.” The mailers featured either the former mayor or local party chairman. We knew that most of the studies on mail-centered turnout programs advocated a softer approach, something much more subtle. But we decided that bold graphics and the name of the voter would attract more attention. We also figured there would be some blow back from such an aggressive approach, but in fact it was minimal.

In addition, we had 20,000 volunteer hand-written post cards, 7,000 of which were hand-written to apartments or no-phone-available households. This was augmented by pre-election mail that was personalized and included the voter’s polling location.

Absentee turnout made the success of our program reasonably clear. In the primary, Democratic non-Election Day absentees came in at 40.9 percent (66,448 voters) and Republicans at 34.9 percent (57,636 voters).

In the general, with the unions spending heavily, and our more targeted programs working over our target file, Democratic non-Election Day absentees came in at 39.8 percent (66,134 voters), while Republicans came in at 35.9 percent (59,683 voters). In raw numbers, Democratic non-Election Day absentees actually went down a few hundred votes while Republicans went up about 2,000 votes.

The campaign was able to walk 10,000 homes and call 15,000 each week over the four weeks leading up to the election. Personal follow-up calls were made after the door-knock. Precinct leaders also used text messaging to remind target voters in their precinct to vote and included the address of their polling location.

The local GOP funded and organized a 300-person per day paid walk program for the four days leading up to and including Election Day, while the campaign turned out 250 volunteers on the Saturday before and 150 on Election Day. These efforts helped us walk our target universe twice over the four days—more than 40,000 homes, and 27,000 on Election Day.

On Election Day itself, the campaign and party were able to deploy poll watchers to 300 precincts, checking off those who already voted and walking the houses of targets who had yet to vote. Workers left household-specific door-hangers that included specific polling locations and poll hours, and then checked off those who had already voted again at the polls. After this initial precinct walk, the list was returned to headquarters so volunteer phone banks could make follow-up calls.

Labor was rumored to have deployed 1,000 foot soldiers on Election Day, but Faulconer had done such a good job of outreach to traditional Democratic voters that it was hard to motivate them to turn out to vote against him. Internal polling showed Faulconer pulling 20 percent of Democratic votes (including 30 percent of those exposed to the anti-labor message). Public polling showed him pulling one-third of Latino voters.

Absentee votes eventually accounted for 65 percent of the total votes cast, and while Democrats had a slight advantage in real numbers, the GOP over-performed and Faulconer built up a 13 percent advantage going into Election Day.

According to our internal tracking, we held a 5 percentage point lead a little more than two weeks out. Around that time, unions went negative on television and closed the gap for a few days. Unfortunately, it sounds as though they polled in the period after their negative ads sunk in, and before our negative ad went up. Once our negative sunk in and re-established the gap, they went all in.

A full 20 percent of Democrats and more than 60 percent of independents supported Faulconer. Voters clearly viewed Faulconer as an agent of nonpartisan good government, de-coupling him from their views of more partisan Republicans.

In the closing days, we doubled down on television, radio, pre-roll and tele-town halls geared towards low-information voters, hoping to blunt what we thought would be a labor-driven Election Day surge for Alvarez. In the end, Faulconer over-performed on Election Day as well to win 53 percent to 47 percent.

Jason Cabel Roe, Duane Dichiara and Stephen Puetz are partners at the California-based Republican firm REVOLVIS.