Tara McGowan, founder and CEO of the nonprofit ACRONYM, invested $75 million in digital advertising this past election focused on defeating President Trump and helping progressive candidates. She argues in this week’s Brand Insider interview that Democrats regained the mantle of digital innovators in this cycle.
The Dems generally evolved digital GOTV efforts, upped their social influencer game and focused on suppressing disinformation. McGowan’s team put special emphasis on advancing measurement and optimization -- parts of the standard marketing equation that political campaigns often lack the time and tools to manage. It turns out tracking viewer emojis played a key role in understanding how video creative resonated with swing voters.
McGowan is a longtime evangelist of digital marketing in Democratic campaigning, as she also ran digital advertising for the massive Priorities USA PAC during the 2016 cycle.
Political wonks will want to lean back for the full podcast
version, where McGowan reflects on how the consultancy/agency structure in political campaigning skews media allocation and stifles innovation. It echoes eerily sentiments we hear increasingly in
the general consumer brand world.
McGowan will also be among many other political operatives featured at MediaPost’s virtual event on Jan. 26-27.
This will be the eighth year MediaPost has hosted this after-action report on what was learned and missed by political marketers in the last cycle. Team Biden’s lead media buyer, the Trump campaign’s CMO and digital leaders from the DNC, RNC, RGA, NRCC, Biden for President and more will be joining us.
MediaPost: For many general marketers, this election cycle tugged at an existential question about the real value of media and advertising in this environment. There were record levels of money spent by Democrats, and it is unclear whether such mind-numbing levels of ad exposure really moved many needles.
Tara McGowan: We live in an attention economy, where everybody lives on different social media platforms, and within those platforms they live in echo chambers that are really designed by their own engagement through algorithms. And so you really need to take a much more diversified approach to media and advertising to reach your audience where they are -- and politics was slow to adapt to this new environment.
I am very, very pleased and proud that the campaign spent more on digital advertising in this cycle than ever before. But even that isn't enough. I do believe there's a time and place and audience for TV buying, but there is no way for us to understand diminishing returns.
So how people [determine] how much they should spend is really about competitives -- how much the other side is spending. We know now that the most money you spend as a campaign or candidate does not mean that you are the surest bet to win -- did actually used to be the case. You could see polls move based on the number of points you purchased.
That's no longer the case, and for good reason. Because the vast majority of the American electorate, they are not getting their news and information from TV anymore.
It's a very difficult question to answer: Does media matter? I think, yes. But what media you purchase and what content and content strategy you deploy on those platforms is really critical. And also, what KPIs do you pay attention to? There has been very little accountability in media spending to know that it works.
MP: There has been a lot of discussion about whether digital GOTV efforts could fare well against traditional door-knocking. What did we learn about the relative merits of each this cycle?
McGowan: We've been advocates, long before the pandemic, of bringing organizing tactics online as well, because again, your response rate is much higher. With door-knocking, you might talk to three out of 10 people whose doors you knock on, if you're lucky. That's a really poor return on that investment of that person's time.
Whereas, if you text-message those 10 people, your response rate is going to be closer to 8 or 9 out of 10. Even before the pandemic, we knew the data backed up that this was a more effective approach.
I do believe that the conservative movement really effectively started pushing their narratives and organizing networks online more quickly than progressives did. And yet, I think that we not only caught up but leapfrogged conservatives in this past election.
So, of course, [we used] SMS and peer-to-peer text messaging. The other thing was influencers: micro and celebrity or national influencers, and micro influencers, whom I think are a key part of the story of 2020 and innovation in politics. That is, finding people who have followings, who are trusted messengers of people in their community -- and having them deploy your message. That is incredibly powerful because [the] messenger, as we know, matters sometimes even more than the message.
MP: What are the key KPIs for gauging influence and persuasion in these environments?
McGowan: Persuasion, that's the toughest to measure, as any brand marketer knows. You need to be able to survey a sample of your audience.
We are very deep believers in in-field research.
Most of the tools to measure persuasion are survey or panel, which means that you are presenting an ad to a closed panel of representative voters of your audience. They are watching it and then they are responding to it.
We don't believe that's as effective, because when you are scrolling through your Facebook feed, it's [a] much more noisy environment. We need to understand if people are seeing the ads, when they're getting served them, and then if they are having that desired effect.
And so, we really do believe in serving our audience the ads in the native environment on Facebook, and creating randomized treatment control experiments that allow us to withhold a control group that is not served the ad. Both of those audiences, the treatment and the control, are surveyed and asked questions about, for instance, their support of a candidate or a horse race question. This approach has been around for a long time, and you can design those experiments.
Where we innovated in the space was that you used to have to take six to eight weeks before you could even get results. You have to be more nimble. We built a platform called Barometer that allowed us to run these experiments on a weekly basis and make it more cost-efficient by surveying our audiences directly on Facebook. [We continued] to push out different narratives to them and understand which ones were moving which segments of our audience, and we shared those learnings with the community.
It also led us to a second platform we built based on certain emojis on Facebook that voters would use to respond to a video. For instance, there's a haha emoji, there is an angry emoji, a sad emoji. It would predict if an ad was persuasive or if it was having the opposite effect or backlash effect, based on the threshold of different emoji uses.
We found that if an ad had a very high threshold of haha emojis, and it was not a funny ad, that predicted that ad was actually pushing people closer to
Trump, when our desired effect was obviously to push them away from Trump. We took those ads down very quickly.
And on the other side of the coin, if an ad had a high threshold of angry and sad emojis, that predicted that it was having its persuasive effect. And that was really helpful because then we could run a ton of different ads, and boost news content on Facebook, and quickly only put the most money behind the ones that were working.
MP: Ad executives right now are asking their underlings, “do we have an emoji metric?”
McGowan: Yes, we called it a haha ratio.
MP: Let's spend a second talking about disinformation and how countering it was an important component of the Democrats’ effort.
McGowan: The challenge right now is that social media platforms and their algorithms have made conspiracy theories and disinformation spread much more rapidly. When it is identified or reported or taken down, it has already done a great deal of damage.
Most disinformation is spread organically through organic posts, memes, text messages, in message groups and Facebook groups.
Two different things need to happen: offense and defense. And I would say that the most resources and energy on the Democratic side this past cycle were spent on defense -- monitoring disinformation narratives where they were spreading; trying to find the root of them and then working with the platforms, when the platforms were responsive, to get that disinformation curbed and those posts taken down.
And when the platforms were not responsive, to go public. The challenge with that is, the number one rule with any misinformation is that you never, ever want to amplify it. You never want to repost it even if you are debunking it.
So there was a great deal of work and infrastructure put together to monitor and shut down disinformation narratives. However, it's everywhere, and it's incredibly hard to find every piece of disinformation and understand which ones are going to gain more traction. I believe that we still do not know the impact of QAnon, the narratives that they spread, and the influence that they had on voters’ opinions.
MP: What went right and what went wrong or needs to be reevaluated for political marketing in the next cycle?
McGowan: I think that we made a great deal of progress. Again, I think that we out-innovated the right. I think what often happens is that when one party or entity makes a great deal of innovation and progress, they tend to rest on their laurels. That creates the space for their opponent to out-innovate.
That happened to Democrats after the glory years of Obama when we really were the most innovative party on the internet. Trump took back that mantle, and I believe that we have now taken it back. So there was a great deal of innovation and experimentation that happened.
When it comes to
things that we need to change, I really do believe that there needs to be a much deeper understanding and acknowledgement of the role that the right-wing media and disinformation plays on the psyche
of the American electorate and that paid advertising alone is not sufficient to counter that.
We need to be building media infrastructure and properties. We need to be deploying influencers who have real audience reach and trust. And we need to be taking an all-of-the-above approach.
We can no longer just invest in television, digital and radio advertising and organizing if we want to build real power and have an impact. I do believe that the consultant culture is holding us back from that, and their relationships to key stakeholders and party leadership.
I am hopeful that the ground is shifting, because 74 million Americans voted for Trump. This was not the Democratic landslide that the polls or the pundits made us believe it could be, and that is because a significant portion of the American people believe the disinformation that is reaching them on social media platforms. We are not sufficiently countering that.
I want to see more deep investment in infrastructure that reaches voters and Americans every single day of the year, the way the right has. And again, not just rely on what I described as the icing of the cake: the reinforcement and advertising. We actually need the cake.