PBS FRONTLINE: Brad Parscale was a digital and media advisor for Donald Trump’s 2016 presidential campaign. He is currently managing Trump’s 2020 presidential campaign.
Parscale tweeted Monday:
Watch my full @PBS interview. Haven’t seen an outlet dump a raw interview like this in quite awhile. It’s long, over 70 minutes. The reporter tries to force a narrative. I fight back and destroy all the media’s preconceived opinions. #StopTheBias https://t.co/hY68nbGmMV— Brad Parscale (@parscale) December 3, 2018
This is the transcript of an interview with FRONTLINE’s James Jacoby conducted on August 8, 2018. It has been edited in parts for clarity and length.
At what point did you decide really to make the push on Facebook? What was your rationale in thinking at that point?
When did I make the push to be on Facebook?
Yeah, meaning the Trump campaign, the digital campaign of 2016.
Well, I think that started back in 2015. During the early portions of the primary, I would say a couple months after Iowa, [Trump’s son-in-law and adviser] Jared Kushner and I were on the phone, and we were having an open discussion about the possibilities of how we help him in Iowa and New Hampshire, how we help Donald Trump get through the first early primaries. We had good conversations about that Facebook would be a strong avenue to push his message.
The campaign we ran on Facebook in the primaries was much different than what we ended up doing in the general election, but we had less budget then; we had a different goal. However, we knew that Facebook, the audiences were there; the people were there; the people we needed to touch [were there], and there was probably no better way to, per dollar, connect with them other than Facebook.
In the primary season, what was the strategy on Facebook, and how did it kind of shift going into 2016?
Shock and awe.
Shock and awe? How so? What does that mean?
Which means is, put [out] Mr. Trump’s message, let him speak directly to camera, and get it to as many people as possible.
And why was Facebook the ideal medium for that?
Low-cost CPM [cost per thousand impressions], large numbers of conservative voters, ability to broadcast all day, multiple times to the same audience, and the numbers were showing in the consumer side that people were spending more and more hours of their day consuming Facebook content and aggregated news feed.
What does that mean in less technical terms? An aggregated news feed—what is that?
Well, it means that their eyes are finite, right? Every day, so many people can only look and read so much stuff, right? That can never change. We can’t add more of that. The only thing that can change is what that’s viewing at, so people watch less TV; they read less newspapers; they want to read more Internet content. The force has been to read more aggregated content that’s being fed together. There’s only a few platforms that really aggregate content well together, Facebook being one of the best, probably the most diverse of all of them. Where Twitter is kind of a Wild West show, Facebook was concentrated at the point of 2015, was probably the most aggregated, concentrated aggregation of content so people can understand what was really happening on the political landscape. If you have any best place to show your content, it would be there.
Basically the fact that Facebook had in some sense become a source of news and information for voters was a big deal to you?
I don’t know if it’s about Facebook being—[it was] a source of aggregated content. Yeah, it was a place where the eyes were. That’s where they were reading their local newspaper and doing things, so being able to show a message directly from President Trump talking—or at that point Donald Trump—talking directly to camera was very important. I could get it right there not filtered by the media, not filtered by anyone. It was his face; it was the person you wanted to hear from talking directly to you. And Facebook allowed long-form video right into it – from the ad platform directly into news that’s aggregated from multiple fields.
So we could get our message injected inside that stream, and that was a stream which was controlling the eyeballs of most places that we needed to win.
When you say “shock and awe,” do you mean that in that you were bombarding people with content—?
Bombarding, yeah. Yeah, it’s not about what we were showing; it’s not shock as in the type of content. It’s the shock of: here is a considerable amount of content to just continue to show directly from the president or Donald Trump. I think shock and awe just in the military sense of “Let’s just go and flood the zone.”
Unlike microtargeting that took place in the general election, that capability wasn’t available. You would need an unlimited amount of money to do that nationally, but for a single state like Iowa, the ability to flood the zone is much more capable at a much lower budget.
But in terms of shock and awe of content, one of the things: Things go viral, right, on Facebook, and they’re sort of playing to the algorithm of what’s engaging content.
Was there thinking inside of the campaign that what’s more engaging on Facebook, for instance, is more shocking content, more incendiary content?
No, it had nothing to do with incendiary content at all. What I mean is that Donald Trump speaking to the American people was a shock to the system. They had for so many years received a concentrated message that had been filtered by the media into what they were being told they needed to do. They were slowly being convinced that they needed to be something they weren’t, and Donald Trump speaking directly to camera telling them this isn’t how it has to be; that I’m a guy that can stop this. We can stop the pouring in of illegal immigrants; we can stop government fleecing your money; we can be more efficient with your tax dollars. That was the shock.
The left wants you to believe that that was something that was above and beyond. No. It was just a message different than theirs, and it was a message about a conservative voice, and he delivered that directly to them in large volume. We didn’t have to make anything; we just had to let Mr. Trump talk directly to camera, and they listened.
You know as well as anybody that Donald Trump was an unconventional candidate who said things that didn’t appear scripted, may have been a little bit more like—it may have been a message that was more—
More resonated with the people who actually wanted to hear it.
Well, maybe it’s that, or maybe it’s that it was more outrageous to some degree, and that it plays well on social media.
It’s only outrageous to a few leftist media people who don’t like that because they want to take the country in a different direction.
I know, but are you telling me that inside of the campaign there wasn’t thinking about what actually triggers the algorithm, what actually makes things go viral?
… No? That—
That wasn’t thinking inside of the campaign at that point?
Nope. It was to sell Donald Trump’s message. He has never changed his message to equal an algorithm.
Right. But does it work in the advantage—? I’m just wondering, because you know how social media works, right?
You know that the algorithm favors content that’s engaging—
—favors content that is emotional to some degree.
So you’re saying that the messaging of the campaign did not play on that in any way?
Well, it did, because the American people—the reason why it went viral is because people were seeking that message. That’s what they wanted to hear. They wanted to hear someone [who] said they would drain the swamp; they wanted to hear someone that wasn’t playing to this media left. And what happened in this country and why Donald Trump’s rise to being president of the United States was—Americans wanted to hear someone talk to them direct. They didn’t want to hear anymore that—I can laundry-list a thousand things, but they didn’t want to be spoonfed anymore. They wanted to hear it directly from the person, and that’s why it went viral. He didn’t have to do anything. There was never any content we put up, and there’s never been a single screenshot of anything that’s ever been taken, with millions and millions and hundreds of millions and almost billions of views, because we never put anything up like that.
It was Donald Trump talking to the camera through the primaries.
When you say never put anything up like what? What didn’t you put up?
I don’t know. There’s been a lot of accusations that we somehow put something up that was incendiary or that in some way caused problems. We just never put anything like that up.
Facebook and Digital Advertising
One of the things that’s interesting is that back in 2016, digital advertising was kind of a dark space, right, especially on Facebook. There could be no public scrutiny of what political ads were on there. Only Facebook basically knew what was being advertised on its platform.
That’s not true.
How is that not true?
Because if you run an ad for $5 and your five friends share it, all their friends share it, so there is no hidden dark posts. Dark posts are only hidden from your profile page, not hidden from the world.
Right, they’re hidden from the—but the only place that actually has full visibility into what was advertised online at that point on Facebook was Facebook, correct?
Other than maybe the 100 million people that saw the ads.
Was it in any way seen as an opportunity to or the chance to experiment … with different messaging, for instance; that there wouldn’t be—I mean, granted people could share, right—?
I get that. But the rules are very different, and the game is very different than on TV. A lot more people see it; it’s exposed to the public more—
That’s not actually true. It’s actually less. If I run an ad in only eastern Washington on a local DMA [designated market area], who all sees it? Only people in eastern DMA or Washington, right?
So if I run a Facebook ad in that same DMA, and they share it with people in L.A., who just saw it more?
I can give you all that—mail, same way, less shared. Actually, social media dark posting is the most shared of all ad platforms when you choose a singular DMA.
So you think there’s a bigger opportunity for public scrutiny of what’s being advertised during a political campaign—?
From social media? Yeah, 100%. The difference is that when the left saw that we used it so well, they panicked and thought that somehow we twisted people’s minds, which didn’t happen, just because they couldn’t explain it—because they [for] once lost control. That’s the beauty of the Internet. Because the ads could be shared, because it was so open, it allowed it to expand and to become a movement. It’s the actual opposite of what they want to believe.
How does it change your strategy going into 2020 the fact that Facebook’s changed some of its policies about, for instance, they’re going to—?
Show all the ads?
Yeah, show all the ads.
It will just save me a bunch of money, because now everybody get[s] to see all my ads for free.
It doesn’t hamstring you in any way in terms of what ads you want to show to what people?
No. It just lets other people see my ads for free. I mean, it’s kind of like a gift.
Facebook and the Trump Campaign
Getting back to the transition from the primary to the general, what were you thinking about using Facebook and the power of Facebook targeting tools going into the general?
After the primary’s over? At the general, I recognized that the same tactics weren’t going to work. Not the fact that it was something different and odd; it was just the fact that to do the same thing and to flood the entire United States with ads, Facebook isn’t going to work. You’re going to need $200 billion. Who knows what you would need to control, put that many ads to that many people, right? It’s just not going to work.
We recognized that we really just need to show the ads to the right people at that point. That’s where microtargeting comes in. What microtargeting allows you to do is to upload Custom Audiences of people and say, “Well, these are the people most likely to show up; these are the right audiences we need to show up,” so we recognized that, and a little over 10 to 15 million people—it was about 14 million people in some audiences, and we recognized those ads. Now, were those ads only shown to those 14 million people? No, they were shown to like 100 million people. The reason is all those millions of people shared it with all their friends. But most likely friends tend to be in friends that are groups like themselves, so the other people that were seeing that tend to be the people who we also want to vote for Donald Trump. That’s the beauty of Facebook, right?
Friends usually run in similar packs. That’s not always 100% guaranteed, but that’s just how it happens.
Running in similar packs is like basically a filter bubble?
No, no, no. I’m just saying that friends usually—if you’re a liberal person, you usually have five liberal friends; if you’re a conservative living in the Midwest, you usually have five Midwest conservative friends. I’m just saying, when you buy—when you pick up your original Custom Audience of 14 million people and you show them ads, and if they share them, most likely they’re sharing them with people who are more similar to them, because people usually have friends that are more similar. That’s just an inherent psychology thing. People live in rural areas or city people, things like that.
But the audiences themselves were chosen through the Republican National Committee’s data they had been building over six years to try to make sure that they had the right data for the 2016 election. During that period of time, they were able to start to hone in on who and what people would need to show up to vote, and I was able to then onboard that into Facebook and show our audience, and then test and relearn ads and see what would perform best with these people mainly through fundraising and GOTV [Get Out the Vote].
Slow that down for me a little bit, because it’s really important. I just want to basically understand the mechanism of what you were doing. But before we get to that, did you see right away going into the general election that Facebook was going to be the primary tool for you guys?
Yes, I knew that back in 2015.
How is it that you then actually start to ramp that up? Bring me through the process of getting the data from the RNC [Republican National Committee]. How does that whole thing work?
I think first of all, that’s—I don’t know if I could do that in just this interview.
I don’t mean that negatively. I just mean that that’s a very long question of two years’ worth of work. In essence, what I recognized was the simple process of marketing. I needed to find the right people in the right places and show them the right message.
The start of that was, break down what Trump’s messages were, which he had made very public between tweets, videos, interviews and speaking with him in person, and break that down into: What are our main shells of content or our strongest points to help him win? Match that with the right people. Who are the people that are going to best write for these audiences? Then match that with the right places they live so that their votes matter the most. Do that math in reverse order, then produce as much content [inaudible.] Show them that, and if it works, and they react to it, donate, say they’re going to vote, share it, then you know you’ve created positive content. If they don’t do that, then you go back and you re-tweak the content to show it in the way that allows them to react to it the way you want.
And people will say that’s a different complete ad. No, that ad change might just be the difference of showing him standing up pointing versus him sitting down. It might be the word you use for the call to action. It could be a lot of things that could cause people to react. Anyone that believes they can anecdotally just decide that because their gut was once they made an ad and people looked at it, that’s the wrong way to do it. We use data and machine learning to learn which ads look better and then perform back to that. So it wasn’t like it was an alarming amount of intelligence to know what audience it worked for; the idea is to produce the right content, and I had the best candidate in the world to do that.
In terms of creating the audience and the data that you were bringing, explain how that process actually works in terms of the RNC data and what is actually a Custom Audience? What is it like to interact with Facebook?
This is probably one of the most underreported and misreported things. The data comes in; there’s public companies—or there’s companies; I don’t know if they’re public or private companies—but there are companies out there that sell data. This is consumer data. They’ll sell whatever they can scrape and find about you within American law, whatever those corporations can do, and they sell access to that. Then every state in the United States through different secretary of state laws provide voter data. How and when—different states have different rules of how and what data they share for voting. Most of it is pretty similar, but yes, some make it easier than others, and some make it more difficult and all this stuff.
The RNC scrapes all that. Both DNC [Democratic National Committee] and RNC do this. They scrape all that voter file data. All that data file goes back into a data[base], you match that with consumer data, and then you have a giant environment of data that’s out there in the world. You have polling data. So you have thousands of people or hundreds of thousands of people across the United States running for party offices, from mayors to city council members, to state party to state senate to state congress—to state representative, I mean, to actual senators and representatives and governors and all these people, right, and all of them hire pollsters. All of them hire people to go out and find data. The real trick of it is, is how do you get all that data and all that information all those people are answering about all those important questions, and you drive them all back into a central database.
When you decide you’re going to run for president of the United States, now you have hard-matched data with consumer data, matched with voter history, matched with very comprehensive polling data from all over the country. When you do that, you put that into a machine and then you start to learn. What you start to learn is how people react in areas, people, and individualize what they call hard ID’d. Hard ID’s are then matched with phone numbers, email addresses and everything. By the time all those pieces are put together, then you can actually pull out an audience. You can say I want to find everybody in this portion of Ohio that believes that the wall needs to be built, that thinks that possibly trade reform needs to happen, and so we want to show them a job on trade and immigration.
Then you take that; you export it out into Excel file—very simple. You import it into Facebook with PI in it, which is personal information that’s matchable data—you know, addresses, phone numbers, whatever you have—and Facebook has been in the job of scraping that all from you, and then it just matches them. Then you have a little button on your computer that says that audience, and you can use that for all your ads.
It’s the ultimate targeting tool then.
I don’t know; you tell me.
It’s no different than sending them mail.
You mentioned earlier—you said the “beauty of Facebook,” right? What is the beauty of Facebook then?
Speed. It’s the speed of doing the same thing we’ve already been able to do. Facebook isn’t doing anything else that you haven’t got through mail, phone and TV for 100 years.… You’ve been able to find out [where] you are at home for 50 years of politics, send you a mailer with a custom message on it, mail it to you, get you to call a phone number, get you to donate. That’s been like that. You’ve been able to buy DMAs. I mean, what’s addressable TV? Addressable TV is no different also. You can pick a TV commercial in this household different than the one next door. Those are TV commercials that are doing that. You’ve had targeted emails for over two decades now. You’ve had people call you on your phone that sell you solar panels while selling your neighbor a water heater because they think you need different things. Targeted marketing has existed for as long as there’s been marketing.
They put a billboard on one highway next to you when they know you live in that neighborhood versus one that’s downtown that might sell something completely different. Why? That’s targeted advertising. The difference with Facebook is, is the only thing that was different, was the speed at which you could do it, meaning you could show them one ad that afternoon and then change the ad by the night. Now [that] you have new digital billboards you can almost do that now, but everyone’s just trying to speed up that, and Facebook was the fastest and had the most eyeballs.
The Trump Campaign Ads
… OK. How many ads a day were you creating?
Tons. We made I think over 5.9 million ads between convention and general election and Election Day.
What was the approval process for those ads? How did it go through Facebook approval processes? What were the standards in place?
You would have to ask Facebook exactly what those approval processes were. I’m not privy to what they were there. They have their own process of how they approve ads.
But were you—?
I approved ads, and I had staff. Once ads became evergreen, then they would be run more. But you know, we might write 50,000 ads that only have very, very small nuances changed about them—maybe what the call to action is versus let’s say if we’re selling a hat, “buy” versus “have” versus “give”; red versus green versus blue. I would say there was only a couple thousand root narratives to all the ads.
And to your knowledge, were any ads ever rejected by Facebook?
Oh, I’m sure. But they probably weren’t rejected on—they would reject ads [based] on how big the word was over a picture. They have 1,000 reasons they reject ads. I don’t think we had any ads rejected for messaging, though.
… The Facebook employees that were embedded with the Trump campaign, what were they doing?
In essence, it’s like this: I asked Facebook, “I want to spend $100 million on your platform; send me a manual.” They say, “We don’t have a manual.” I say, “Well, send me a human manual then,” and that’s pretty much it. I wanted a human manual in my office.
And what does the manual provide?
Well, what do you ask a manual? You know, a manual for your car, I say: , “How do I open the doors? Where is the door lock? I can’t find it. I’ve always wondered how I change the oil in here. I don’t know where the oil cap is.” You look in the manual. If you didn’t have that for your car there might be things you’d never learn how to use in your car, right, and you might not be using your car as efficiently as it can be. Maybe there’s a secret button you push that makes the fuel efficiency five miles an hour better. You never knew that button was there.
You didn’t have the manual, and there, when you’d say the guy at the car dealership would tell you, well, he’s a human manual then. Well, he doesn’t tell you either. So how do you know how to push that button? So I wanted a human manual to tell me what buttons am I not knowing.
And what did you learn from the Facebook employees?
I’m sure they taught us all kinds of stuff. It wasn’t just me; it was my staff. They would sit there and be a human manual, so if we needed to get something done—or also, an on-call mechanic. Things break. When you’re using a platform, $100 million, and you’re pushing it for everything it can do, it breaks also. You might wake up one morning and the API doesn’t work the way it did yesterday because somebody pushed some code. They can pick up a phone in five seconds. I didn’t want to lose five seconds. If I would have chosen the way Hillary’s campaign did it, I’d have to send an email and make a phone call, wait a couple days and then have it fixed. I wanted it fixed in 30 seconds.
What other sorts of things specifically were the Facebook employees helping with?
That’s it. They were a manual, and they were there to be a technician.
But in terms of audience targeting, were they helping with that?
No. They would only help if the Custom Audience tool didn’t work.
…The Facebook employees were helping us use their platform, but they weren’t helping us make decisions on the campaign. …
I spent $100 million on a platform, the most in history. It made sense for them to be there to help us and make sure how we spent it, right, and they did it, right? I think it’s insane if you’re going to go ask somebody to spend $100 million on a platform and you don’t ask anyone to be there. I think that’s just crazy.
I understand that. I’m just wondering—for instance, one question is the ad approval process. Did they help with getting ads approved? Quickly?
I would imagine if we got one locked up, maybe through reasons that—they might call and say, ”Hey, why was this one…?” So they can tell us.
I don’t think they would have rammed something through, but you’d have to ask those employees. My guess is they would be there in case the ad-approval process broke down. They’d be there to call in and say, “Hey, why is this broken?” I would doubt the employees we had in our office would have the ability to manually approve ads.
What was their data good for for you in terms of like, how—obviously—?
Lookalike Audiences, maybe?
So explain how that works. What is Lookalike—?
Well, if you have one audience, let’s say 300,000 people, you can click a button, and it will find people who look like those 300,000 people.
Facebook does that for every commercial vendor in the country. So if you needed to expand the Custom Audience and maybe say, ”Hey, maybe Facebook has some people that look like this that we don’t know,” then you could say, “Hey, I really need a 600,000-person audience. I only know 300,000.” Upload 300,000, see how well it [Facebook] can find people, and then the testing would allow us to see if that was accurate or not.
How accurate was it?
Facebook Lookalike Audiences are pretty amazing. I mean, it’s why the platform’s great. Now, that’s probably one advantage over a mailer, because U.S. Postal Service doesn’t tell you that. The TV channels will do that some, the addressable TV and some of them. But the Lookalike Audience is probably one of the most powerful features of Facebook.
Facebook and Russian Disinformation
You’re describing a very powerful tool; obviously a very coordinated, well-run campaign. I mean, during 2016 we know that other actors, bad actors, were using the same tools that the Trump campaign was using. Do you think that Facebook bears any responsibility for the fact that bad actors were also using these powerful tools to mess with the American electorate?
You know, it’s a complicated question. I would think that Facebook would probably want to verify who’s running ads first. I think once they verified them, then the burden changes some. Even the United States government can’t and police force can’t manage every single person, every human every second, right? It’s an impossible task for anything, especially something as big as Facebook. Could they have maybe questioned why a Russian company with Russian IP addresses are buying ads on a platform? Sounds kind of suspicious to me also.
You know, maybe it was just a little amount of money they spent that just went under the radar, you know, $6,000 or whatever in the last six weeks or whatever, eight weeks. I mean, I was spending that per half-second or millisecond probably, so if you take that over six weeks, they were spending .0001 cents per second. I can get how maybe it would fly under the radar. I think they’ll do a better job now, though, of probably verifying initial ad vendors.
So this problem, – does it go away? But I would imagine it’s more to do with the small amount of money they spent. If they would have [come] and spent $100 million, and they said they were going to, or they even started to spend that kind of money I would imagine. I think one of Facebook’s biggest mistakes was they were in a conflict of marketing versus ethics. To actually explain to everybody that $6,000 in ads wasn’t doing much also makes you sound like your platform is not that powerful to do something for cheap. So if you want to get rid of every small vendor who’s using your platform in America, go on TV and say, “Well, that actually isn’t very much.” Facebook was caught in a catch-22 to say Russia couldn’t do a lot with that little bit of money, or whoever these actors were. Also says that all the small businesses across America, you can’t actually do something effective for that little amount of money. So they were kind of caught in a little paradox. What they’ve almost said is: ”Yeah, $100,000 and $100 million are the same thing. Even $100,000 to $6,000.” They only spent less than $10,000 in the same period in time that I spent $100 million.
Have you been satisfied with Facebook’s response to interference on their platform during the 2016 election?
I think of all the platforms, Facebook has done the best job, I think, of trying to even the playing field. I think they are a PR disaster.
I think they’re trying to do some things. I think they’re also now so worried they’re going too far, and I think it’s one of the problems also. I think owning that the situation happened and, you know, make some changes to make sure it doesn’t happen again I guess would probably be the best way to do it.
But do you feel like there’s been a proper accounting of what actually went on on Facebook during 2016?
No one’s ever given me that information.
Well, do you know—presumably you’re well read in on what—
I mean, only what I see in the public. By the way, though, I don’t believe everything I read anymore since 90 percent of the stuff that’s written about me is false, so it’s hard to believe what [is true about] everybody else's once you’ve had it happen to you.
Do you think that Facebook has any responsibility for the fact that no one knows what to believe anymore, because it’s “fake news” on their platform that spread around or that there’s—?
Yeah. I mean, I don’t know. I mean, CNN’s on Facebook every day, too, and I would tell you that half the stuff that comes out of CNN is fake, so why is Facebook still allow[ing] CNN on there?
But you personally, do you think that we know all we should know as the American public about what happened on Facebook, this really important medium for both content and for advertising to American voters? Do you think we have the complete picture of what happened there?
I don’t know. I think that a lot has come out. The problem is, again, Americans don’t understand how the system works, and they think this was some huge thing. I explain it like this: If you took Madison Square Garden and you said that was a salad, and you had to eat the whole thing, and you took three pieces of salt, and you put it on the very top of that, like literally table salt, and then ate that whole thing, and if you could say that you tasted that, then I think that’s about what’s happened.
And who are the three pieces of salad?
No. I’m saying the whole bowl of salad is all of Facebook during the election campaign, all the noise that was happening. From what I’ve read, it’s like putting three pieces of salt on it is how much these actors put into the system. And to think that that influences that entire bucket of salad’s flavor is crazy to me. Until someone shows me a piece of evidence that shows it wasn’t three pieces of salt, it was dump trucks’ worth of salt, I‘m going to think it was three pieces of salt inside a giant salad bowl the size of Madison Square Garden and you’re never going to taste it.
As far as I know, Facebook has not released a repository of advertising that happened on their platform during 2016, whether it was by the Russians or the Trump campaign or any other actors who were—
Well, Trump campaign is not an actor. Trump campaign is the official campaign of the president.
I understand, but dark-money groups or anyone else—
Those are also, if you’re talking about super PACs and (c)4s and all those things, that’s part of the current political system. The only thing that’s been accusatory is what these actors did from this Internet Research Company or whatever, is the only one that I know about and the only one I’ve read about.
What I’ve read about is during the last couple months of the campaign, there was only a fraction of money spent, well less than $100,000, somewhere down under $10,000 range. There’s been different accounts of how much that is, some saying $6,000, some saying $8,000, but the fraction of that of probably what was spent on Facebook was probably close to $500 to $600 million in the United States by legitimate organizations. And to try to say that that $500 or $600 million, that $6,000 somehow influenced $600 million is the biggest piece of just malarkey I’ve ever heard.
Even though Facebook is actually a very effective targeting tool?
Yeah. They can’t say that, though, because that’s like saying … some restaurant in Wichita, Ks., puts $5,000, and all of the world the next day is going to know how great that restaurant is from $5,000 advertising. Just doesn’t happen. It’s not possible. I don’t understand it. If someone shows me the proof that somehow it influenced somebody, I haven’t seen it. But I think the media wants you to believe it, because they want to believe that somehow none of this is legit; that these guys somehow faked all—everybody out for $6,000. And I think it’s a big joke. I mean, I can barely fly from here to Hawaii and back for $6,000. I’m not going to change the entire American electorate for it.
You also know how Facebook works. It’s not just advertising; it’s also organic content and groups, and —there’s clear evidence that the Russians and the Internet Research Agency were creating a lot of organic content that had a lot of reach.
I have not seen a lot of that content. There’s been a few posts about that. They would have had to made a lot of friends of real people to really create a lot of influence. I think it would be a considerable amount of work. I haven’t seen that evidence that they have had any kind of real influence [on] any person’s vote. I just haven’t seen it.
Cambridge Analytica and the Trump Campaign
The Trump campaign paid Cambridge Analytica nearly 6 million bucks. What did you get for that money?
Well, $5 million, that was a TV buy, so you’ve got to wipe off $5-million. [Then-campaign CEO] Steve Bannon made a purchase of TV advertising that ran on the East Coast. They have a television division where they literally just place television. They were paid I think approximately around $800,000. … We received staff—the reason I hired Cambridge Analytica was actually not for Cambridge Analytica at all. I didn’t know about their company. I didn’t know anything about them. They had hired some of the [Gov.] Scott Walker (R-Wis.) digital team in 2015. These were the guys that had really helped Scott Walker re-elect, and there was a guy named Matt Oczkowski. When I met Cambridge, I thought they were full of crap, but I met Matt Oczkowski and I really liked him. He had experience running a very well-targeted and well-run digital campaign. I actually wanted to hire him without Cambridge, and he said, “I’m under contract.” So I asked him for an employment contract, so I hired them for staff only, and each one of the payments between then and Election Day were for staff only. Then Matt worked on my team with four or five of his people, and they mainly ran polling, visualization and support staff to all the things we needed to do to get things done digitally. I actually didn't hire Cambridge Analytica for any data work. Or hire them for any other data.
… Do you regret any affiliation with Cambridge Analytica?
Hindsight is 20/20 now. I don’t regret, because I’m sitting here today, and Donald Trump is president, so I’m not going to regret that decision. You know, it’s sad what Cambridge Analytica executives did, but I barely knew them, didn’t even have any contact with them, and from the day I hired them till after the day of the election, I actually never talked to any of those guys.
What do you make of the fact that they used the same tool Facebook—one, they have user data on people that they got unethically from Facebook, and then they target—I mean, what do you make of the fact that these types of companies are running, in some cases, disinformation campaigns using Facebook?
I have no idea what those guys do. It’s not anything I do, nothing we did on the Trump campaign. Who knows what people do out there in the world? There’s billions of people out there, and I couldn’t tell you what they all do.
Does it piss you off, though, that there are probably some of the same voters that you were targeting were also targeted by groups that were spreading misinformation or disinformation?
I was completely upset that Hillary Clinton was putting misinformation into my people about what she was going to do.
I’m just saying. Hilary was, in her campaign, put out lots of things they said they were going to do and misinformation about President Trump. You know, that’s a political campaign. In a political campaign, you take your position; you stay strong with it. And the Clinton campaign put out tons of stuff that was not true about Donald Trump and who President Trump is.
I think that to try to say that conservatives and Republicans were somehow trying to use Facebook in an organized manner to be misinformation is crazy. I didn’t have any part of that. Our campaign didn’t have any part of that. I would actually say we ran the least amount of negative ads I’ve ever seen in a presidential campaign. We ran almost all positive. We had some contrast ads to different things of statements she made, but for the most part we just tried to get our people to show up.
What were your favorite targets in terms of demographics, income?
I would say we had a few different targets that were very important, and I’d say one was low propensity to turn out Trump voters. You had a lot of Trump people and Trump supporters that maybe only showed up one or two last of the four elections, not because they didn’t care about this country—they love it. They’re just hardworking middle-class Americans who on that day their kid has basketball camp, or they have to pick three kids up from school; it’s like, “I don’t have time to get over there.” A lot of people live hour to hour, paycheck to paycheck in this country, and some of them struggle just to put food on the table, and sometimes the thought of getting out to vote feels disconnected for them.
I think the idea was for one major audience was for Trump to connect to those people and show them that he had a plan to help them and that he could bring real change, and not this “You’re going to feel better about yourself” change, “The economy’s not going to get better and you’re not getting any better, but we’re going to make you feel better” change. Donald Trump brought change that was going to make their lives better, and that’s what we wanted to get. So that was one of the most important audiences.
I think after that was an audience of people, of Republicans, who didn’t like that they were losing their establishment hold on control. Those were the Republicans that had showed up to vote every time and weren’t happy because they felt if all these new voters show up, they would lose their influence over the party. … That’s the “power group,” I call them, and that was another audience. We needed to get them convinced to show up, because if they didn’t show up, Hillary’s in power. That was worse than the other situation for them. So that was another audience. Those were probably two main audiences that you had to get to show up to vote, and they were slightly in conflict back then together. I think things have changed now.
Wait. Those audiences were in conflict or no?
Well, I’m saying those people are. I’m saying the one didn’t like that new ones are showing up, and the other new ones didn’t like the establishment ones. We need them both to vote for the same guy, because they didn't really want the other one.
That’s a binary choice there slightly. They’re both Republicans. They both actually believe in the same plan; they just—one believes in a methodology of lifelong Republican swamp, and the other one believed in a guy that was coming to bring change and would attack a Republican just as much as a Democrat if it stopped real change for these middle-class voters.
What were the most successful negative advertisements you guys did?
We really didn’t run a lot of digital negatives. We probably ran some. We ran some TV advertisements that were negative. I think it was about her being in establishment for 30 years, that it was going to be the same. I think the most successful campaigns we ran against her were probably that it would be four more years of the same from Obama and that in all that time, things hadn’t changed for them and gotten better, and if you want four more years of that, vote for that. I think that was probably the most successful campaigns we ran.
Facebook and Microtargeted Ads
There’s been reports that you targeted to some degree African-American ZIP codes.
Yeah. So tell—
No. We didn’t write any kind of suppression and/or campaigns in any way that ran to—I don’t think we actually—we didn’t run on a campaign, I believe. I would say I’m nearly 100 percent sure we did not run any campaigns that targeted even African-Americans, which people think is crazy. We individualized this campaign instead of segmenting that way, meaning we were—I don’t assume that an African-American and a white person are different because they live in the same neighborhood. I would take what they do and how they consume information. The things they buy, the jobs they have, [would] be more important than the color of their skin.
So I didn’t see that as a choice that I needed to do. I—
Right, but there’s one thing about targeting on ethnicity, and it’s another thing to target ZIP codes that might be predominantly African-American more.
I don’t think we targeted by ZIP codes.
I mean, I targeted by individualized. Microtargeting’s not ZIP codes.
OK. So explain the difference for me then.
Well, ZIP codes would be assuming that everyone in a ZIP code is the same. I didn’t do that. All of our Custom Audiences were by the person, so consumer history, voter history, polling data all matched back.
Now, are certain information from demographics and geographics probably built somewhere into those models? I would imagine so, somewhere in the underlying belly, but we didn’t say, ”Hey, let’s just go target this ZIP code with this kind of message.” That’s just not how we did it. Our Custom Audiences are imported. State-based would probably be the main geography, because people in Ohio vote for us; I’m not looking for people in Nebraska. Well, I had Nebraska, one zone, but let's take Oklahoma. There was no Custom Audiences with Oklahoma people in it, right? States matter, because we’re an electoral system, not a popular vote system, right?
But no, we didn’t target like that. That was an anonymous source that gave that. I don’t know who that anonymous source was, and I’ve been saying I never did that since.
So …it’s completely false that African-American voters in Pennsylvania were targeted with ads about Hillary’s record with superpredators?
Yeah, I don’t think we ever ran the superpredator ad. We had—I guess one of my staff members—we told them to be creative, make whatever they wanted, and then we would say no or yes to things. I guess this reporter thinks he saw an ad that was an ad where she was talking about being a superpredator. I don’t think we actually ran the superpredator ad.
I do think we ran radio ads with her saying that in certain DMAs [designated market areas]. I don’t think we ever ran social media ads like that. It is something she said. It's something she said that I believe is extremely racist, and we did run it. It’s something she said. We didn’t actually say anything to it. We just ran what she said on radio.
Do you have any concerns about the nature of the targeting? Just regardless of the Trump campaign, but just how this tool, right, which is clearly a very powerful tool with all of the data that there is out there about voters and individuals, coupled with what Facebook offers, do you have any reservations about—?
Zero, zero. The reason why is I can’t stop somebody from mailing me, I can’t stop somebody from calling me, but I can turn off Facebook.
And do you think there’s no difference between the fact that you could have run in 2016 an ad to one person and the message could have been completely different to another person?
Nope. No, because when you get on Facebook, that’s what you’re buying. So why are you on Facebook? Facebook’s product is to get on and get information that is targeted to you so you can simplify your life, and then you get on there, and they show you ads that are simplified to you, and then you’re upset with them. Seems pretty hypocritical to me.
So you don’t have to use Facebook. There’s no law that says you have to be on it. You have to have a mailing address. Most of us have to have a phone to survive. But there is no rule that says you have to have Facebook. And when you simply sign up, it says you’re going to be targeted with ads for usage of this platform, and then people are now upset that that’s what it did.
The Problem with News Feed and Ads
Do you think it was clear in 2016 that there was a substantive difference in a Facebook feed between ads and content—
It says “Ad” on it.
—and sponsored content?
It says “Sponsored.”
And you think that an American voter is going to know the difference between that?
Yeah. I mean, I’ll tell you what’s worse, is you open up The Wall Street Journal or, actually even worse, you open The New York Times, do you know the difference between an opinion piece and a news piece?
Yeah, you do.
Yeah. Sure. There’s an op-ed page; there’s an editorial page.
There’s now online. There’s one word, one word underneath their title that says “Opinion.” What’s the difference between the one word that says “Sponsored”?
I think that people know that when, first of all, it’s not—if you look back to how things were done in 2016—
No, 60 percent of the content now in major newspapers and on major outlets now are opinion pieces.
Thirty years ago it was less than like 10 percent. We have turned into an entire media company and media outlet system that is mainly— we’re on opinion and commentary instead of news. You turn on any major news media tonight, and you watch it, and you can see how much of it is commentary and how much of it is news. That has flipped. I think that what has happened—
Well, the news industry would say in response to that that—
It’s what makes them money.
Well, it’s the only way to get engagement on the Facebook—
To make money.
No, on the Facebook platform, because what engages people when Facebook is your news source is opinion and is polarizing and—
I would say what the Trump campaign put up online was 100 times more real than all the news articles that were put up.
Are you serious?
How the—in what way? I mean-
Because I read the articles, and they were false, one after another after another, because I actually was in the room. I was with him; I listened to him; I was in all the meetings. And piece after piece after piece was full of anonymous sources, single-source content, fake stories over and over again. Every day. Every day. It’s why I record this conversation. It’s why everything I do now I have to think that everyone’s going to do something fake against me.
There were also troll farms, or there were also kids in Macedonia whose articles— fake articles about candidate Trump at the time or about fake articles about Hillary Clinton at the time—were going viral on Facebook, right; that were getting more play than articles from The New York Times.
I don’t know how much play all those got. To make the argument, though, that those kids in Macedonia made more impact than The New York Times is insane.
I think that the numbers bear out that the fact that some of those articles got shared a lot more than a New York Times article—
I think you could go back in time and say lots of conspiracy theories and weird things have made a lot of—JFK, 9/11. I mean, you want to go down all of them, we can just go back. And there was no social media [with] JFK, and can you imagine one American that doesn’t know what that conspiracy theory is?
I think that it’s—humans naturally are mesmerized by conspiracy theory, because when they can’t answer something they don’t want to believe should have happened, or they don’t understand why it’s happening, so they have to inherently put structure to any though process, right? So I think they naturally read these things, and sometimes they want to believe them, because—and the people who want to believe them the most are the people that don’t like you the most, or those who do believe you like you the most. I’m still yet to convince that independent voters and swing voters are consuming [a] considerable amount of that stuff. I think that noise has been there for as long as human history’s been.
There’s been satire magazines; there’s been content produced for millenials, you know, or millenniums—it’s like however long we’ve been creating this content, that’s probably the wrong word, centuries, right, of content that’s been created, that is designed to confuse people and to be entertainment. Who knows why people create it? I just think it’s existed forever. And I think that to ask for a platform to control that 100 percent only asks for one thing: to block free speech. I think the reason why we have free speech in our Constitution is it’s on no one but yourself to determine what’s right and wrong.
Anytime you try to stop something, what happens when you stop the wrong one the one time? That’s the reason why we have free speech.
It’s one thing for to stifle speech; it’s another thing if the system that they’ve created, that Facebook has created, rewards content that is engageable and not necessarily true, or engaging and not necessarily true.
Well, if people don’t want to see that, then they shouldn’t get on Facebook. I mean, it’s a free country. It’s a free country to make a platform the way you want to do it. It’s a free country to say the things you want to say.
Do I think there should be anything on there that hurts people? No. Now, that’s always where that thin line gets to get. But I also think there’s a lot of conservative voices in this country that actually have real messages they believe that I don’t think are fake. They believe—Whether that’s religious conservatism, whether that’s social conservatism, whether it’s whatever they believe, it doesn’t make it fake. It doesn’t make it painful; it doesn’t make it hurtful. Do I think there are people that do stuff like that? Yes. Do I think white supremacy groups should be on there? Hell no. I completely am disgusted by anything that has anything to do with that.
But it’s a very thin line. You have to be very careful to try to block it. I think when you do that, you set yourself up for a much bigger problem, and I think that’s censorship. If you go back to early European times, and you go to any time, people start taking away your guns and they start taking away your voice, that’s when problems happen.
… [On Facebook] There’s more of an opportunity for opacity of who’s advertising and how they’re targeting people.
Well, do you know everybody who's advertising in Des Moines, Iowa, right now?
I don’t, but there’s a way for me to find out, right? There’s ways to find out that are much more public than having to ask Facebook for that data.
That’s not what you said. You didn’t say that there was after market services that are greater at doing that. The TV networks- there’s after market services now because they recognize that that made it a business to try to sell what all the different commercials are and monitor them all. So what I’m saying is somebody could have easily done that on Facebook too. Facebook now is saying they’re gonna do it themselves but there is nothing different of me sending 50,000 ads on Facebook. You say the question is do you feel like there’s a certain amount of hidden or not opaque to showing those ads? And I said, no, that’s not different than me mailing 50,000 people. I actually think it’s more, because the odds of taking those 50,000 people and actually showing people your mail before they read it and throw it away, I would actually say the chance of these 50,000 people sharing it was much higher.
You’re saying there’s actually things you could do that are much more microtargeted, that are much more hidden if you wanted to. To say that Facebook had created a system that somehow is a secret evil thing is a very incorrect statement. The only reason anyone is upset about this is that Donald Trump is president and used a system that was all built by liberals.
I disagree with that.
No one ever complained about Facebook for a single day until Donald Trump was president.
No, that’s not true. There were plenty of people that were raising warnings about the idea that this is a completely unregulated digital advertising space.
OK. It went from 1 percent to 99 percent. [crosstalk] When I got on TV and told everybody after my interview of what we did at Facebook, it exploded. Then when Cambridge [Analytica] did their stupid thing with whatever they do with data, which had nothing to do with our campaign, and people even exploded more. The funny thing is the Obama campaign downloaded the entire social graph and used it, then went on TV and newspapers, and they put it on the front of a magazine, and the left and the media called them geniuses for doing that.
Yeah, but there were also people that had major qualms with that no matter what your political persuasion.
Do you not see the hypocrisy that the difference is—this other company go gets the social graph, does the same thing? By the way, the same time period. It’s now four years old. Doesn’t even get used on a presidential campaign, and they need to testify. These people are put on Wired magazine as the geniuses that brought you Obama. These people were said: they’re the evil of earth, and they did the exact same thing.
Well, no, they didn’t do the exact same thing.
They both—these guys downloaded the entire social graph.
Right. They did, but they got permission from Facebook to do so, and Cambridge Analytica did not.
OK, but why in 2000—in his [Obama’s] re-election , why didn’t one media outlet attack them and say: what they did was wrong, that they got in people’s, you know—they didn’t—?
No, I’d say there was a big failing all—
It’s because they wanted Obama. The media wanted Obama.
That may be your interpretation. I think that this is more about corporate accountability than it is about our political persuasions here. I think this is about the fact that there was data—
I think it was wrong when they both did it.
Right. That’s different.
I think they’re both wrong to do it. I don’t think Facebook should–I had never openly used a social graph for anything. I have never imported that and used it for any of my political and/or commercial stuff that I had ever done.
I just—I don’t know. I just—first of all, I don’t think it works. And I think content’s more important than that.
Secondly, I don’t know. I never really thought it was worth the effort. And the second thing is, I don’t know if it’s actually wrong if Facebook says that’s the rules and people understood it. I do think it’s wrong if people didn’t know they were having that done to them. And I do think the place that Facebook was really sketchy probably was that they didn’t let people know how their social graph could be exported. I don't know if back then I knew when I used Facebook exactly what they were sucking in. You know, it was years ago. I mean, this data is like four or five years old I think now, right? I think it’s almost six years old now.
I mean, they shut that down back in like ’13 or ’14.
Right. There’s a federal consent order against Facebook for having shared that data.
Yeah. I never think anyone should have taken anything from them that they don’t know is being done. I do think that Facebook does have an inherent thing that they’re using your data to better serve you, though. Now, sharing that, that’s a different problem. But I don’t understand why the Obama campaign didn’t get lit up for doing it. If everyone’s so upset about it, it seems to be in the appearance from a third-party view of my own that because it was used to what they thought was good purpose, that made it fine.
It goes back to the same liberal argument that raising energy prices is fine as long as we get to a greener world. Or we do all these things that damage is OK as long as it’s for political purpose. So it looks like that for a lot of Americans, they read those articles and say: “Well, Obama did it. Why did they put the guy that did it on the cover of Wired?” By the way, the person who whistle-blowed that—after the 2016 election, everybody was getting weird—was someone that worked on the Obama staff and said, “We did all this plus more; why is everybody freaking out?” That didn’t get almost a single article written about it.
Well, it’s gotten some coverage, and we’re certainly going to cover that, but regardless of it being Obama—that’s your theory about it—but still the company was giving out that data without users knowing about it.
I agree. Yeah, it sounds like it. I mean, I wasn’t a party to either. From what it seems like the outside was, Facebook made a mistake, and they tried to fix it. They knew they shouldn’t have shared that much data, and they’ve been trying to fix it ever since. They tried to go out and get rid of everybody’s data that had done it, and I think—look, I don’t defend either group. I don’t defend Obama or Cambridge Analytica. The both—If they used it in an improper way, either of them, then that’s wrong. If they broke the terms and conditions of Facebook, Facebook should deal with them.
I didn’t use it. Didn’t have any party of me. I didn’t ask for it, and I had nothing to do with it.
Do You Trust Facebook?
You think Facebook’s been a responsible company?
Hmm. Tough one. I think Facebook has done a lot of things to try to make a lot of money.
And you’re happy in 2020 to give them more money?
Most likely, if that’s where people’s eyeballs are.
Even though they haven’t necessarily been totally—
I think they’re making the changes now. I think they’re a lot—you ask me “ever.” You said, “Have they ever?” I would say that they’ve probably had some issues back in the day. I think they’re a young company that was growing fast, and I think they also—sometimes tech and all these things run into each other. I think right now they’re doing everything they can, and if they’re smart, they’re going to make sure that they make the most ethical platform they can.
But I also think that’s a thin line to go into censorship, restriction, and try to be police. I think right now you have YouTube, Twitter and Facebook almost trying to join together and see if they can create what’s conversational health, and I don’t think I want them to create my conversational health. My conversation might not be the same and agree with you politically. I imagine you and I sit here today of completely different political [persuasion]. But do you want someone to actually regulate what I say to you? No. And I don’t want them to regulate what you say to me either.
There are certain things that need to be protected against, but blocking conservative views is not one of them. And if I believe that God has a place in my life, then I’m not wrong for saying that. If I believe—
I don’t think they’re blocking content like that.
Yeah, they are. They blocked a lot of people’s content—stuff like that. I have had multiple people send me their accounts blocked, show me the post, and the post was a religious post that they believe was hate speech. In my view, there was nothing hate about it. And I think that that’s wrong. A lot of people have different views in this country in different ways.
How confident are you that the 2016 [sic] midterm elections, right, Facebook, still a main place for advertising, still a main place for content for American voters, do you feel any more confident that American voters are protected from disinformation and misinformation on Facebook?
No. First of all, every voter in this country should recognize that their neighbor might come and tell them something that’s wrong. They should judge everything with their own mind, and they should make their own decisions, and they should do their own research. They should believe all things, including mass media, with skepticism, and they should think themselves and make a decision based off that.
I think that’s been around for as long as mankind has existed. And to think that somehow 10 years ago we were getting perfect information and today we’re not is just a falsehood. Look, it goes back to my argument again that there’s nothing better the left would like to do right now is to make you think that all these platforms and all this information is wrong. But they’re right.
I don’t think anyone’s trying to- that’s- that’s- that’s a big statement, all right. I’m talking about the fact that you know, there’s undoubtedly this is-Facebook has become a primary information source for American voters. It was a place that was taken advantage of in 2016, right?
I think it’s an aggregator. It’s an aggregator.
And it was taken advantage of in 2016 by bad actors to spread false narratives and divisive content to American voters.
Do you think any false information was spread by mail, phones or TV?
Of course. But what I’m saying is that mail, phones and TV were different media. There’s one major medium right now, OK?
I would still say TV’s bigger than Facebook.
It still is?
Yeah. We spent more money on TV. TV advertising still trumps it, pun intended. You know, TV—
Maybe in ad—
No, ad and media. I would still say TV new