Donald Trump’s ascent to president-elect of the United States was a complete shock — that is, according to mainstream media experts and journalists, many of whom initially considered his campaign a joke.

Now, just as Americans are divided, the media is even more polarized, with journalists forced into a moment of self-reflection as they gear up to cover the new administration. Let’s take a moment to consider how the media business fared during the campaign and election cycle.

Ad Spending

Losers: Hillary Clinton’s campaign; the broadcast industry
Winners: Donald Trump’s campaign; earned media

Compared to the 2012 elections, 2016’s presidential race proved to be a loss for the candidate who invested in running expensive television ads. As of late August, Hillary Clinton’s campaign and pro-Clinton PACs had spent more than $90 million on political advertising following the primaries while Donald Trump spent nothing, as Variety’s Jan Dawson reported.

By late September, it was obvious this unevenness had affected the broadcast industry: Sinclair Broadcast Group Inc. took back its political-spending forecast for the year because of the unpredictability of the campaign, the Wall Street Journal’s Austen Hufford and Rebecca Ballhaus reported. The move sent broadcast industry stocks tumbling.

Following suit, Gray Television Inc., which owns stations in 51 markets across the U.S., also “withdrew its guidance for political advertising revenue for the rest of the year.” The primary reason above all else was the uncertainty from Donald Trump challenging the status quo — in this case, not putting in the ad dollars expected for a presidential campaign. Instead, Trump largely took to Twitter and free media to get his point across, and the simple messaging associated with his campaign helped enable its efficacy.

As Hufford and Ballhaus wrote, “Industry watchers are divided over whether the lower spending is a short-term anomaly caused by the abnormal nature of Mr. Trump’s campaign or a sign of a broader rollback of the importance of television advertising.”

One thing is certain: A frugal campaign that results in a victory is a win.

Pollsters and Data Journalists

Losers: Nearly everyone who predicted a win for Clinton.
Winners: Nate Silver for hedging; Allan Lichtman

From the New York Times to the Huffington Post, nearly every major media outlet using data science had specified that Donald Trump had a slim chance at winning — and so everyone in that crowd lost. “I’ve believed in data for 30 years in politics and data died tonight,” Republican strategist Mike Murphy declared in a tweet. “I could not have been more wrong about this election.”

But as numerous journalists and analysts have written since the elections, the wrongness has more to do with the failure of interpretation rather than a direct failure of numbers. Writing in Vanity Fair, Nick Bilton argued this is nothing new. “Polling has grown less accurate in recent years as society has changed,” he said. He cited how landline ownership in the U.S., a traditional polling method, has diminished, alongside the willingness to pick up the phone when an unfamiliar number calls. Then there’s the silent majority: “What’s clear now is that most of those who decided not to speak publicly about what they really planned to do in the voting booth were Trump supporters,” he wrote.

In an op-ed for the Guardian, data journalist Mona Chalabi argued that the obsession with “predicting opinions rather than listening to them” is to blame. Chalabi, who used to work at FiveThirtyEight, said that data journalism can be consumed by its own arrogance, to the extent that simply saying “it’s complicated” is not good enough. And Americans, who crave simplicity and a desire to know, wouldn’t have accepted that answer anyway.

Nate Silver at FiveThirtyEight took heat for giving Hillary Clinton less of a chance than other sites, so he earned a slight repreive post-election. And then there’s the case of Professor Allan Lichtman, who uses a simple “13 keys” approach with true/false answers, and predicted that Trump would win. He now has picked the right winner since 1984 (though he was off on the popular this time, as Clinton won that).

Regulators and Mergers

Likely Losers:

Donald Trump regularly accused the media of being out to get him and claimed the elections were rigged. Now it looks like his administration would likely block certain mergers and acquisitions that would consolidate media power, notably the $85.4 billion AT&T/Time Warner deal.

The point of acquiring Time Warner — which owns CNN, which Donald Trump has frequently touted as biased — was for AT&T to diversify its revenue sources. As Jim Puzzanghera wrote in the Los Angeles Times, “the deal would transform telecommunications giant AT&T, which bought DirecTV last year, into the nation’s largest entertainment company.”

But Trump says the deal is precisely the kind of power structure he’s fighting against, as “it’s too much concentration of power in the hands of too few.” Comcast’s 2011 acquisition of NBCUniversal is just as bad. “Deals like this destroy democracy and we’ll look at breaking that deal up and other deals like that,” Trump has said. “This should never ever have been approved in the first place.”

Still, the stock market has rallied since the election in the assumption that Trump will not follow through on a lot of his promises, so we’ll see if he backs up his rhetoric with action.

Social Media

Losers: Facebook and other social media for allowing the spread of fake news, and filter bubbles
Winners: Facebook and Twitter on election night with massive audiences and interactions

What happens with too many algorithms and an excessive number of news sites available on social media? Confirmation bias becomes the norm and it’s easy to miss out on other views, for one. That’s long been a concern as Facebook, Twitter and other social media sites — but especially Facebook, with its enormous power and reach. It’s in no small part why many Americans may have been shocked at the election results if nothing on their News Feeds hinted that such an outcome could have been likely.

But in the lead-up to the election — even as it helped Americans register to vote — Facebook’s series of missteps have shown the social giant is struggling to come to terms with the role it played in the U.S. election. Critics argue that Facebook allowed fake news, hyper-partisan websites and conspiracy theories to run rampant and spread untrue and misleading information. As Mark Zuckerberg remains firm that 99 percent of Facebook’s content is authentic, BuzzFeed’s Sheera Frenkel reported that an anonymous internal task force has formed within the company to take on fake news. And both Google and Facebook recently announced plans to stop allowing ads on sites with fake news.

Meanwhile, social media continues to dominate during big news events, and election night was no different. Twitter had a record 75 million tweets about the election that night, while Facebook had 115 million people generating 716 million likes, posts, comments and shares related to the election, and 643 million video views too. So while social media has been under the microscope for the spread of fake news, it still has massive reach and is becoming an ingrained media habit that’s hard to supplant.