What? Now the presidential race is over, it’s clear in hindsight there were huge marketing stories driving the outcome.
What? Now the presidential race is over, it’s clear in hindsight there were huge marketing stories driving the outcome.
November 9, 2016
Whatever else can be said about the 2016 presidential election, it was a triumph of marketing for President-elect Donald Trump and a failed campaign for Democratic nominee Hillary Clinton. It was an even bigger miss for marketing research in the form of election polls, which largely predicted a high probability of a Clinton win heading into November 8.
With that in mind, Marketing News reached out to some industry leaders for their thoughts on the election. Specifically, we wanted to know:
1.What did the Trump campaign do right from a marketing perspective?
2.How did the Clinton campaign fail to properly connect?
3.Why were the polls wrong?
Trump connected emotionally with his audience in a way the Clinton campaign never quite managed. Making an emotional connection is a timeless marketing best practice. … No amount of logic or list of qualifications could overcome the emotion of the Trump movement.
The polls, the media and the pundits missed it because they are concentrated in major markets and our information sources now simply reinforce our own opinions. It takes resources, time and interest to understand the diverse audiences that make up the U.S., especially the rural audiences.
Trump did a number of things right from a marketing and positioning perspective that enabled him to secure the nomination and then win the election.
1. He established a perception of authenticity that resonated with his constituents.
Trump’s “shoot from the hip” style, while viewed as a liability amongst his detractors, actually probably helped him with his supporters. We saw examples of many candidates trying to” appeal to younger voters” by doing things like creating zany YouTube videos. Remember Ted Cruz’s “how to cook bacon on a rifle” video?
2. He embraced TV, which proved that it is still king in politics.
Trump’s PR strategy enabled him to get something in the ballpark of two times the “earned” TV exposure of Clinton, despite being vastly outspent by Clinton on the paid advertising front. His constant TV coverage also created a ubiquity effect for Trump … A massive percentage of our survey respondents reported seeing ads from Trump across multiple channels, even when he wasn’t running any.
3. Strategic use of social media.
Trump’s favorite social media platform was Twitter. He used it to get the attention of the elites who gravitate toward Twitter, in particular, members of the press, political junkies and other influencers. It was incredibly effective and cheaper than issuing press releases (it was free actually).
4. He consistently hit Clinton where she was weakest.
Since the start of the general election we asked Americans about which attacks were the most effective against each of the candidates. The top answer in regards to Clinton was consistently that “she broke the law by using a private e-mail server,” and Trump kept pounding on the issue.
[Clinton’s] digital sophistication and massively outspending Trump on television was not enough … to counter Trump’s message to his own constituency. Message undoubtedly played an enormous role here. Trump connected with a massive swath of Americans who feel disenfranchised and are disillusioned with the establishment in Washington, and his message resonated with them. In that sense, Clinton, or any other Democrat, may never have really stood a chance.
The most reasonable theory/explanation we have [for the inaccuracy of the polls], in my opinion, is that Trump brought out a lot of new voters who didn’t vote in the last cycle, which skewed the weighting that pollsters applied to survey samples.
Trump’s classical conditioning effects for four months have been:
1. Clinton equals corruption; Clinton equals murderer (Benghazi); and Clinton equals old and tired and can’t get anything done in Washington. As a consequence, people “hate” Clinton—it’s visceral.
2. Washington doesn’t work. Those Washington insiders never accomplish anything. Trump to the rescue.
3. People are now very afraid of immigrants, GLBT, etc. Trump will protect U.S. citizens and their jobs from these populations.
Clinton’s problems include:
1. Implicit gender bias. Women and men do not like strong women. People don’t say this—it’s not socially desirable, but research documents this.
2. Connecting with people. She was perhaps too late and light on the children and family message. There are not images of the soft side of Hillary Clinton.
3. Last-minute, unsubstantiated FBI accusations reinforce the “Clinton equals corruption” message. Though she was cleared, there was no time to “recover” with Election Day so close.
The inaccuracy of the polls was due to:
1. Closet Trump supporters who would not publicly disclose their support for Trump but voted for him.
2. Undecided voters, not considered in the polls, who decided to cast a ballot for Trump (per above).
Two of the most motivating emotions are anger and anxiety.
Using social media in addition to his rallies, [Trump] tapped into the anger and anxiety around terrorism, abortion and loss of status among white, older males. Then he was able to tap into the negative side of trade agreements by relating stories of unemployed people as a result of NAFTA (think Ohio, Michigan, Indiana and Pennsylvania).
Individual loss is much more powerful and amenable to stories than is a 2% improvement in overall GNP as a result of trades or a lower cost of many products. These facts are not amenable to stories. How powerful are these emotions? Enough to offset an annualized growth of 9.41% in the Dow during Obama’s presidency, a cutting of unemployment to almost half, and a real increase in the median household income over the last two years. No stories here. No emotions.
Hillary really did not connect to the younger Bernie Sanders backers.
To them, a vote for Hillary was like stepping back into the Clinton/Bush era. She didn’t give them a reason to support her. It also didn’t help that she chose a moderate middle-of-the-road white guy as her running mate. This lack of edginess told the millennials that their vote was not important.
We won’t know the real root cause of flaws in the analysis for a while, but the current outcome of the presidential election only amplifies the importance of getting the data and analysis right. Whether it’s for politician predictions or choosing the best target market segments, channels and marketing investments to drive business growth, data matters. Marketing leaders can be the vigilant drivers of this data and analysis.
[Clinton lost because of] “predictive intelligence.” The Clinton campaign relied on predictive analytics to make assumptions about how people voted. Instead, they should have the used machine-learning approach and intent signal monitoring to pick up shifts happening in key counties and states that their polling methods missed.
We always advise marketers that if consumers feel nothing, they do nothing. This was a very emotional campaign. Rationale evaluation of the candidates seemed to be the farthest thing from voters’ minds. Regardless of the issues, Trump seemed to be able the land an emotional message to his supporters more effectively than Clinton.
There wasn’t enough storytelling in what Clinton’s presidency could mean to people.
Among Trump’s “distinctive assets” was the “Make America Great Again” slogan. Clinton’s messaging never gelled as a fluent, memorable slogan, and even campaign material was complex in terms of her promises.
The polls were definitively were measuring the wrong things.
Humans are woefully unreliable witnesses to their own behavior. Much in the same way that we steer marketers away from asking purchase intent questions, in the case of the polls, voters might have overstated their “undecided” status out of some reticence or even embarrassment at their intention to vote for Trump.
Trump managed to use social media to get his message across directly, unfiltered by the media. By combining this with a general anti-establishment, anti-media message, Trump made his followers in particular distrust media/establishment opinion in favor of his own messages. The social media profiles of politicians and public figures are very carefully manicured and follow best practices; Trump breaks all the rules. He says whatever he wants and uses an almost conversational tone, particularly on Twitter.
Trump has been a master of tailoring his messages and approach and created a brand that resonated with his followers and beyond.
Trump also attacked Clinton on her alleged corruption with a series of hashtags that came off as advertising taglines like #DraintheSwamp and #BigLeagueTruth that his followers adopted widely. Additionally, part of Trump’s success on social media has been down to his very “authentic” voice.
Clinton ran a traditional campaign … that didn’t stand up to the “wild West” of social media. Clinton depended on surrogates, most noticeably celebrities, to carry her message in social media. By depending more on traditional television advertising, Clinton couldn’t keep up with the constant barrage of tweets put out by the Trump campaign.
This election embraced social media, and the pollsters weren’t ready for it.
It is fair to say that Trump would have had much more difficulty gathering support had it not been for the massive reach of social networks. Social media may have played a role in creating a kind of scandal-driven (as opposed to issue-driven) campaign where topics such as Trump’s attitude towards women, Trump’s tax returns and Clinton’s e-mails tended to dominate discussion as opposed to actual policy issues.
[Trump] deeply understood core market motivation: frustration.
He got well beyond the surface and the polling data to the depth and conviction around the need for change.
He had a clear and powerful message.
He stayed on message and did not react to the all the challenges and external noise.
Clinton did not accurately read the consumer sentiment.
She did not understand the lack of true brand health of Hillary (the intense distrust, even dislike). Hubris for a marketer is death.
She had no message in a world of intense dislike and distrust of politicians.
Finally, she underestimated Trump.
She assumed [the public’s] distaste for him and his behavior was enough for him to implode. Assume your competitor is really smart and has it right).
The first thing that Trump did right was event marketing.
He outworked Hillary doing four, five or six events in a single day, sometimes in multiple states. These personal appearances … gave him the chance to get an unfiltered message to his backers. Plus, many of these events were picked up and played on live television. It helped him overcome the paid television advantage held by the Clinton campaign.
It’s not fair to say the Clinton didn’t connect. She won the popular vote. Hillary’s problem was that she is a very well known brand and many of the events in the campaign reinforced the negative parts of her brand image.
The polls stopped being objective measurements a long time ago.
They have become marketing pieces for campaigns and the organizations that pay for them. They are meant to be instruments of influence not illumination. They have some work to do to rehabilitate their reputations.
[The Trump campaign] knew their target.
They knew that target’s dissatisfaction with incumbency, spoke their language and made people feel like they were part of a movement—highly motivating and translated to votes.
[For Hillary Clinton,] … it was a failure to motivate, excite and mobilize.
One of the issues the Hillary campaign will have to reflect on is over-confidence and how they drastically misjudged the upper Midwest. They also failed to see that this was not a traditional campaign and traditional tools like heavy media spend and ground game were not going to be the deciding levers.
[The polls] underestimated turnout for Trump and most likely did not source a truly representative sample. Something clearly went wrong here. The traditional polling methods are broken. I’m guessing that the established polling infrastructure undercounts rural voters. The media also forgot the concept of ‘margin of error’ and kept talking polling in absolute.
The election results represent a marketing revolution in American politics. The Trump victory offers marketers of all products and services, in both the for-profit and nonprofit [sectors], lessons:
Lesson 1: Follow the marketing concept. Trump was talking about what mattered most to American people, namely jobs. Hillary spent too much time trying to tie-in Trump’s negative and tarnished image with raising children. In the end, jobs trumped family.
Lesson 2: Use technology strategically. Trump’s use of Twitter and Facebook allowed him to stay connected to 25 million voters, supporters of his, on a weekly basis throughout most of the campaign. Hillary took her base for granted and did not take advantage of the unfiltered nature of social media to stay connected with her loyal democrats, many of whom defected and ultimately cost her the election in some key states leaning towards her.
Lesson 3: Develop a unique brand identity. The Trump brand was unique in an election setting where voters were looking for the outsider from Washington to change the course of direction for the country.
Lesson 4: Create a winning advertising strategy. The Trump brand demanded the attention of the media and afforded him the ability to run his advertising campaign by spending close to half of what Clinton spent in advertising. Ultimately, voters bought his hard-nosed message.
Lesson 5: Build a relationship with your customers. Voters who supported Trump were constantly reinforced with his use of Twitter in particular, something that Clinton did not do, which cost her the election.
Lesson 6: Be prepared to engage in crisis management. Trump navigated his crises better than Clinton did hers, and in particular, the last-minute announcement by the FBI was not handled well by Clinton—and was probably the key reason she lost—as it reinforced her negatives that Trump pounded away at throughout the race.
While Hillary Clinton presented a logical case for how she would move the country forward over the next four years, Donald Trump appealed to the emotions of those who felt disenfranchised and left out of the country’s current economic boom.
Trump was able to create messaging that resonated with his supporters. He labeled his opponents with monikers that were highly memorable (“Crooked Hillary”) and repeated them time and time again. Clinton’s decision to ignore this and “go high when they go low” was understandable, but it ignored the marketing truth that repetition creates awareness and eventually an immutable connection.
The fact that the polls were so wrong should make us question the effectiveness of how we approach marketing analytics. If we focus on the wrong pieces of data, we come to the wrong conclusions. Don’t take old assumptions for granted. Just because we haven’t previously seen a Black Swan doesn’t mean there isn’t one that could dramatically impact your business in the future.
In every survey [we conducted], it was clear that Trump supporters were voting for Trump rather than against Hillary. This result stands in contrast to much of the mainstream reporting and punditry in this election cycle.
Trump was selling a clearer story about what to expect from his presidency.
Clinton took a what she thought would be a free ride, expecting people to vote against Trump, rather than invest in a clear and compelling brand story selling the electorate on why they should vote for her.
In this election, white male voters behaved like a minority block, and Trump was their Obama.The polls always under-represent likely voters making $50,000 or less. The assumption is translated into under-counting minorities. But I believe the data also under-represents working class white males, a group that doesn’t necessarily have a great track record for voting in general elections.
The [poll] models didn’t change their assumptions reflecting the worldwide populist movement. Brexit and the Colombian vote against the peace agreement with FARC were just the latest indicators of the rise of populism and disgruntled working class majorities around the world. These events should have played a larger role in the assumptions of the models, but I suspect that the bias of American exceptionalism may have overwhelmed the better judgment of the data scientists.
The polls did not get this right for three separate reasons:
1. It was a really close race, and the numbers were within a margin of error across many states, so in the end, no one knew how they would go.
2. There is a social desirability or the shy-Trump effect—those who were supporters did not want to admit it. We see this in marketing research as well. People may overstate their intention to buy something and then not end up buying it. In this case, the shy Trump supporter did not want to overtly acknowledge his or her support, thereby skewing the poll numbers.
3. The polls did not account for the larger turnout from rural areas, which primarily accounted for the surge in Trump support.