View profile

Why 2020 will be a Facebook election, too

Revue
 
America's longest sustained national conversation about CPMs entered a surprising fifth day on Tuesda
 
February 27 · Issue #92 · View online
The Interface
America’s longest sustained national conversation about CPMs entered a surprising fifth day on Tuesday, as writers and thinkers continued to analyze the dynamics of Facebook ad auctions. It’s a weedsy topic — even weedsier than this newsletter hopes to be on most days — but it involves an important question: does Facebook’s ad platform incentivize politicians to create more polarizing, divisive ads, because they spread faster and more cheaply than less inflammatory ones? 
By the end of the day, we would have internal Facebook data to consider, and some new ways of thinking about the question. Firm answers, though, remained a bit beyond our grasp.
Hillary Clinton kicked off the latest news cycle on Twitter, with a post that got wide pickup: “We should all care about how social media platforms play a part in our democratic process,” she tweeted. “Because unless it’s addressed it will happen again.”
At the Washington Post, Philip Bump talked to one of the people who ran Clinton’s digital advertising campaign, and found him unperturbed by the suggestion that Donald Trump’s ad campaign had been more cost-effective. The important thing to glean from this interview is that different ad campaigns are designed to produce different results, and are priced accordingly:
“Fundamentally, we were trying to do two different things,” said Andrew Bleeker, a senior adviser on advertising to Clinton’s campaign and president of Bully Pulpit Interactive. “It’s not that one was right and one was wrong, but they were different. They shouldn’t cost the same.”
Clinton ran a lot of video ads, aiming to boost turnout or make the case for her candidacy. Trump’s ads were often single images with an enticement to click. The ads were designed to do different things.
“When I look at cost, I don’t necessarily care at all what he’s paying,” Bleeker said. “I care: What would it cost me to get this video in front of people on television, or any other format? This is still a good deal.”
In a Twitter thread last night, former Facebook advertising VP Andrew Bosworth played down the impact of cost differentials in campaign advertising. “The only caveat is that our auction favors good creative which improves user experience,” Boz wrote. “However that benefit is on the order of +/- 10%, meaningful for advertisers but far from 200x.” 
And then, after I had worked on this story for four months, Boz just … tweeted it out. Or rather, he tweeted the campaigns’ average CPMs from the general election.
“After some discussion we’ve decided to share the CPM comparison on Trump campaign ads vs. Clinton campaign ads,” he wrote. “This chart shows that during general election period, Trump campaign paid slightly higher CPM prices on most days rather than lower as has been reported.” (See chart below.)
So on one hand, you had Brad Parscale saying he paid only “pennies” per CPM in some cases. And on the other, you have the former vice president of Facebook ads saying that Trump paid more than Clinton. What gives?
The first thing to note is that the CPMs shared by Facebook include only what the campaign paid up front. They do not include so-called “organic” reach — additional people reached by the ad campaign based on people liking, commenting, and sharing them. So it’s possible Trump had to pay a little more up front to place some of his inflammatory ads, but his fans picked them up and lowered the effective CPM in ways that are still invisible to us. 
The second thing to note — reiterate, really — is that Trump and Clinton ran very different kinds of ad campaigns on Facebook, making direct comparisons difficult. 
A former Facebook account manager wrote to me in response to yesterday’s newsletter saying that average CPMs ultimately would not tell us very much, precisely because of how different the ad campaigns were. “More important than average CPM for each campaign, the most interesting thing would be to have the CPM for each camp on the ‘overlapping’ audiences (swing states, etc),” the former employee wrote to me. “I think that the differences there may be significant but not wild. I would expect +50% cost difference on a given audience based on engagement but not more than this (ok, maybe 2x for some edge cases).”
Antonio García Martínez, who kicked off this week’s discussion with his piece in Wired, made a similar point on Twitter. “National averages won’t tell the tale on a very complicated story,” he wrote. “This needs to be broken out my market, action type, etc. to make any sense at all.”
Ultimately, then, it can be true that:
1. Trump paid more to reach 1,000 people on average than Hillary Clinton did, because the direct-response ads his team used are more expensive generally.
2. Trump’s ads reached more people, because they were picked up and shared by more people, lowering his overall cost relative to his opponent.
3. Facebook’s ad platform can reward more polarizing ads.
Whether all three or true, and the degree to which they are true, is still a bit beyond us. But perhaps it won’t be forever: Bosworth says Facebook is asking campaigns to let them share more data from the 2016 election campaign. And it will release even more for future campaigns. “We’ll be making data like this plus more available to the public when we launch our ads transparency tool in spring,” he tweeted.
In the meantime, Parscale rode this weekend’s press cycle to new prominence, and as of Tuesday he has a promotion. Trump named Parscale campaign manager for his reelection efforts today. Which means that like 2016 and 2018 before it, the 2020 election will be waged — and possibly won — on Facebook.  

Democracy
Here’s that Hillary tweet in full:
Hillary Clinton
We should all care about how social media platforms play a part in our democratic process. Because unless it’s addressed it will happen again. The midterms are in 8 months. We owe it to our democracy to get this right, and fast. https://t.co/aM3pRrZW4J
7:04 PM - 26 Feb 2018
And Boz’s:
Boz
After some discussion we've decided to share the CPM comparison on Trump campaign ads vs. Clinton campaign ads. This chart shows that during general election period, Trump campaign paid slightly higher CPM prices on most days rather than lower as has been reported. https://t.co/u0qgUQ02qM
12:05 PM - 27 Feb 2018
And that chart:
One of America's Most Notorious Neo-Nazi Extremist Groups Is Posting Freely to YouTube, Steam
YouTube Won’t Ban Neo-Nazi Group Chanting ‘Gas the K**kes, Race War Now’
How Black Twitter and other social media communities interact with mainstream news
Elsewhere
‘Success’ on YouTube Still Means a Life of Poverty
N.C. man livestreams his own killing on Facebook
Curling and U.S. snowboarder Shaun White were popular Olympics topics on Facebook — we just don’t know how popular
Pinterest has hired former Google and Square executive Francoise Brougher as its first COO
Despite backlash to the redesign, Snapchat downloads are up
Ex-Twitter CEO is shutting down his fitness startup
Alexis Ohanian and Tinder are petitioning Unicode for interracial couple emoji
Launches
Facebook launches program to help local news outlets boost digital subscriptions
Facebook is starting to tell more users about facial recognition
Takes
Quinn Norton: The New York Times Fired My Doppelgänger
And finally ...
How Tiny Red Dots Took Over Your Life
Talk to me
Questions? Comments? Pleas to stop writing about CPMs? casey@theverge.com 
Did you enjoy this issue?
If you don't want these updates anymore, please unsubscribe here
If you were forwarded this newsletter and you like it, you can subscribe here
Powered by Revue