Pod Save America - How to Reach the Unreachable Voter

Episode Date: August 23, 2026

How can Democrats reach the elusive swing voter? Danielle Butterfield, executive director of Priorities USA, one of the most important PACs in Democratic politics, joins Dan to talk about Priorities' attempts to reshape how Democratic campaigns think about targeting voters on the internet. She breaks down their research into how voters are looking for information on search engines, why campaigns need to rethink how they talk to voters on YouTube, and how Priorities USA is getting out ahead of 2028 by trying to frame JD Vance and Marco Rubio in swing voters' minds before the presidential election begins. Hate listening to ads? Become a Friends of the Pod subscriber for ad-free episodes of Pod Save America, Pod Save the World, Lovett or Leave It, Runaway Country, Offline with Jon Favreau, and more—plus exclusive content, including bonus episodes of Pod Save America. Subscribe now at crooked.com/friends, on Apple Podcasts, or through the Pod Save America YouTube channel.You can request a transcript by emailing [e-mail]. Include the podcast name, episode title, and air date. Please allow 48 hours for delivery.

Welcome to Potsave America. I'm Dan Feiffer. Today I'm joined by Danielle Butterfield, the executive director of Priorities USA, one of the largest, the most important super PACs and democratic politics. Priorities USA does a lot of essential digital work in the party, building infrastructure, investing in research and data, and running ads to reach the voters' campaigns often struggle to reach. They are in the midst of a massive study about how less engaged voters are getting their information on TikTok, YouTube, Instagram, and even AI chatbots. The platforms that will decide the 2028 election, whether we like it or not. Priorities has already started a $30 million campaign to define the 2028 Republicans like JD Vance and Marco Rubio. And I invited Daniel on the podcast to unpack that research and talk about what they've learned. We dove into their findings, talked about how the media environment

has changed to the 2024 election, and broke down some of the ads they're already running. It was a great conversation, and we'll get to it in a minute. But before we do, please consider becoming a paying subscriber to Friends of the Pod. Subscribers get access to the single best show produced at Crooked Media, Polarcoaster, with me, Dan Pfeiffer. I was forced to say that by our producers, but I really do like the show, and I think you'll like it too. This week on Polarcoaster, we discussed Angie Nixon's upset win in Florida, new polling in the Michigan Senate race, and a strange fake polling scandal on the L.A.

May U.R. race. And while I have you out there subscribing to truly great content, check out my substack message box break down what is happening in politics. You can find it all at crooked.com slash yes we, Dan, or subscribe to Friends of the pod at crooked.com slash friends. All right. Here's my conversation with Daniel Butterfield. Daniel Butterfield, welcome to Potsave America.

Thanks for having me. All right. I'm so excited to talk to you. Thanks for doing this. Priorities has been working on a whole much of things that are really interesting to me. I want to talk to you about the study you've done on less politically engaged voters. I want to talk to you about how you're thinking about ads. And I really want to talk to you about why you are already spending money to try to win the 2028 election in the middle of the 2026 election.

We're going to do all of that. I want to level set on a couple things first. For people who've been around politics as long as I have, they remember Priorities USA as sort of the Democratic Super PAC, spending tens of millions of dollars to elect Barack Obama, Hillary Clinton, spending money for Joe Biden. That's not what Priority USA is now. So maybe talk a little bit about who you are and what you guys are doing.

So our audience understands. Yeah. So I joined the Priorities USA team back in 2017, fresh off the Hillary Clinton campaign. Priorities, you know, up until that point had been operating to use, like you said, as a presidential super PAC. But in the aftermath of 2016, one of the most urgent kind of quickest realizations that folks had about why we lost was just how little Democrats had invested on digital compared to television. And so the priorities team got some funding to kind of stay open beyond the presidential and really investigate why that is. So that's when I joined the team. I became our digital director alongside a bunch of other really smart people. And we kind of made it our mission to really diagnose and understand why the Democratic Party is so severely underinvest in digital, but then also like to actually close the gap to, right, to raise a bunch of money to spend on digital. And so flash forward to where we are today, I like to say that we are anchored as a presidential organization, but we really have gotten to a point where we're using the midterm elections and special elections.

to really prepare and get better and innovate and try new things. So that by the time the next presidential election rolls around, we're like that much more ready for it. So that's how we would think about us today. I know the presidential elections seem like a million years away for a lot of people. But as I like to point out, that if we follow the previous schedule, exactly one year from now, there will be Democrats and Republicans on stage in debates.

Actually, like this thing will be often running the day after the mayor. terms. So people should be prepared for that. And so just also to level set here, you know, as you point out, like in 2017, you're worried about the digital gap in working on it. After the 2024 election, there was this sort of widespread consensus that Democrats had once again been outmaneuvered digitally. There was all the talk about all the podcasts that Trump did and how they were, you know, using the internet to target these less engaged voters. I think that there was a general sincerity among a lot of people in the party to try to close the gap, even if they didn't know exactly how to do it or how the gap got there.

18 months after that election, where do you see things? Has the party made real improvements here? How much work needs to be done? Yeah, I mean, we obviously have made improvements, but we're not where we need to be, right? I think in the aftermath of 2016, it was quite easy to say, wow, Democrats are really not spending anything on digital. In the aftermath of 2024, Democrats outspent Republicans 2 to 1 on digital. So it's no longer a numbers game. It's actually like several layers below the surface, which makes it a lot more difficult to diagnose and solve.

And that's why, you know, it's no surprise that the last couple of years post-20204 have been a really tough set of fundraising years for Democrats because the solutions to why we lost in 2024 really require us to dive deep and solve some big, bigger cultural. problems that we have ingrained in our party. And so, you know, I think if I'm going to diagnose that problem in one sentence, my read is that while Democrats are spending a lot more money on digital than they were back in 2016, we're still running digital programs to mirror what goes on on television. And I think that's the, that's the crux of the problem that I'm seeing today. Could you say a little bit more about what that means? Yeah. I mean, I think, you know, 20 years ago, when a campaign was setting out to reach voters, they had a handful of options of how they were going to do that. They could buy, you know, broadcast cable television points. They could knock

doors, build an organizing program, send out mail pieces. Digital really wasn't on the table. When you're setting out to, you know, run a TV program, you're picking out DMAs that are efficient based on where your state or congressional district is. You're building really fancy models to determine which door I'm going to knock and which door I'm going to skip. You're, you know, building mail pieces that, you know, same thing. You're going to mail to one house and not to another. But I think that what we have really failed to realize is that digital opens up so many more possibilities to be able to reach voters in ways that bleed so much far beyond where you live,

what your DMA is, and what your voter file score is. And we're really failing to lean into the creativity that digital affords us when we do that. And I think put in a much simpler way, nine out of the 10 top YouTube ads in 2024 were just 30 second ads that ran on TV first. So we are quite literally, practically, taking 30 second ads that we run on TV and then putting them on YouTube and expecting people to watch them. And that is quite literally lighting money on fire. I mean, I've been struck by this. And as you know, we've been sort of both together and several scurming about this for a long time. Anyone who watches YouTube would understand that a TV commercial run on YouTube makes no sense.

It would be like someone putting a print ad on your television when you were watching it and you would just be supposed to read it. And they're obviously incredibly smart people working in digital spaces across the party in the Harris campaign and some of these super PACs in the Biden campaign in 2020. what accounts for the fact that we're sort of like jamming this anachronistic square peg into a much more modern round hole? I mean, I think I cannot adequately answer this question without addressing the consultant-driven nature of decision- Go. I mean, let's get right into it. Yes. Right. Like, it is just, if you want to look at profit margins for a second, what I'm basically declaring that a campaign should do costs a lot more than what it costs to run a TV campaign. such that if I'm asking a campaign to cut 20 ads to put on the internet instead of one ad to ship to everyone, that is obviously going to eat into profit margins. And the folks that have been around for

a really long time are going to resist that. So I think that the pressure has to come from a different place. That's part of what priority sees as its role in this ecosystem is pushing the hard questions and trying to get past profit margins as the core goal. It's ridiculous that we even have to have. that conversation. But it's also part of why, like when I joined the team in 2017 and, you know, was offered the job as digital director, part of my negotiation was I want to build an in-house team because I don't want to have to worry about, you know, working with a set of consultants that's ultimately going to be trying to, you know, pay the bills. I want to hire people that are purely focused on trying new things, innovating, and that has to be their number one focus.

You know, I don't want to get on here and, like, consultants are in it for the right reasons, right? So many Democratic consultants want to win these elections, but they also have to pay bills. And when you are forced to make decisions between paying bills and winning elections, that's naturally going to create a attention that we just can't ignore. But anyway, all of that to say, you know, I think it is extremely easy to look at the reach numbers that YouTube affords where you're any given week, 80, 90 percent of voters are reachable on YouTube. to say, oh, that's just an extension of TV.

Just like TV is a one-size-fits-all, reach broadcast. We literally use the word broadcast to describe TV. If you're treating YouTube like a broadcast channel, though, you're just missing the point, right? And I have a lot of good friends that, you know, work on the Google sales team, love them to death. But their sales pitch is YouTube is TV. I just fundamentally disagree with that, right? Like if you are treating it like a broadcast channel, you are fully missing the point. The YouTube is TV thing.

I'm going to get to your study a minute, but you've got me down a rabbit hole. And I knew that would happen in this conversation because this is a issue of particularly nerdy interest for me. I mean, YouTube is TV. And this is a thing I've done the round since 2024 meeting with various politicians. And I've tried to explain to them. And I use YouTube as TV for the purpose of understanding that for people to like actually, especially these politicians, understand that young voters watch YouTube on their TV. Right?

So they come home, they turn it on. And what's different is they're not looking at TV guide or the literal TV guide that shows up in your Xfinity or cable thing is that the algorithm is showing them things based on, you know, a lot of data mining and the things I subscribe to. But where the point you're making is where the analogy fails is that it is not consumed in the same way. And sort of like you can't just put a TV show on YouTube, which is why YouTube does. not invest in scripted television. Like, there's like a lingua franca of it. And if the ads, if you're running TV ads on YouTube, even if it's being watched on a literal television,

that's not what people are expecting. So they are more likely to dismiss it. Is that right? Yeah, just I think it's an attention span problem, right? And it's a function of, you know, if someone is sitting and you're correct that a majority of people when they're using YouTube are literally using watching it on a smart TV. And that is different than YouTube TV,

which is like a subscription service. that people pay for. No, it's literally YouTube.com being streamed on a television device. But like, how many of those people are also holding their smartphone and playing games? Like, I fit into that category. Like, it is extremely tough to get me to pay attention to one screen, and I know I'm not alone in that situation. So you both have an attention span challenge. And then you also have a challenge slash opportunity that people are using YouTube in very different ways. It's a battleground, right? It's people listening to podcasts. It is people playing games. is people watching clips that their friends send them.

Like, it is people listening to music. It's just, it's so many different things that if we're going to try to squeeze it into one category, we're really feeling. And this does open up a really good, you know, segue to a lot of the work that Prairie is doing to really understand, like, you know, I'm pretty tired of media consumption research looking like, let's slap a question to the end of a poll and ask how many people are watching YouTube. I need to know much many, many more details than that to really understand how to craft a media strategy. So that is the work that priorities has been investing in since 2024.

Actually, way before 2020, we were doing this in the middle of 2022. And that's, you know, the work that we're excited to chat about. Yeah. So you've like I said, you've been studying how people actually get information in this crazy, insane world we live in. And not just how people like you and I get information because we kind of get information the same way people have always got information. Like we're going to sources to get it. we're following politics closely. But, you know, as we've talked about sort of endlessly on this podcast and elsewhere since

2024 is the swing voter is not necessarily a well-informed centrist white man from Wisconsin. It is someone who doesn't pay that much attention to politics and is sort of kind of, you know, surfing the news at best. And we're very, we're not very good at reaching those people or understanding those people. So you've been looking at them. So before we get into like what you actually found, can you walk us through the process by which you, you try to understand the media habits of these less politically engaged individuals. Yeah, absolutely.

But just to underscore what you just said, in our post-2020 research, we actually found that beyond age demographic, geographic indicators, that your consumption of YouTube was more predictive of your vote choice and your vote swing from 20 to 24 than anything else. Meaning that you were more likely to vote for Trump, the more YouTube you watched? Yeah, exactly. So basically the cross tab was at men. Men who are watching more than three hours of YouTube had some of the highest swing of any other, you know, cross tab that we were looking at. And that it also applies to people who are listening to podcasts.

It also applies to women. It's just, it's pretty consistent that your internet consumption patterns are very predictive of how much you swung from a persuasion standpoint. So our team, you know, took that information and said, okay, like, we got to be media consultants for the 21st century. Because if no one else is going to do it, like let's let us step up and play that role. And so we basically have this project that our team has been working on that really takes a look at. We have an always-on survey operation where it's actually really cool. We actually, in a normal kind of survey environment, we pay people to send us a screenshot of their screen time apps. I'm sure you guys have logged in to see how much time you spend on Twitter every single day.

But we're actually paying people to send us a screenshot of that and then validating it against what they self-report to where they spend time. and then using that as an understanding of what new apps are popping that should be on our radar. How much is YouTube consumption changing over the summer? How much are people changing their subscriptions between Hulu and Peacock, depending on what reality TV show is on, which is very common. And the thing that I think our team was highly surprised to see how much people change up their subscriptions, which is very good information. So basically, part one is an always-on survey operation to just get a baseline of like,

where are people spending time? We then do kind of quarterly, sometimes monthly deep dives on specific topics that our team wants to dig into. So last summer, we did a deep dive on audio consumption to really understand specifically how are people consuming music, podcasts, etc., what platform, what are some patterns that we can take into account from a media planning perspective. And then we also acquire, you know, third-party data that often is very kind of observational and behavioral in nature so that we can actually observe what people are doing. And the search data that Peter Hamby wrote about over the weekend is that third category, where we paid to acquire people's search data.

And then we did a really long-term kind of study to kind of better understand what patterns we could pick up there. Positive America is brought to you by Haya. Back to school season is here again, and we want to give your kids their best shot to be prepared for the new academic year ahead. That includes their health and wellness. That's exactly why Haya exists. to give parents a real solution in a market flooded with kids' vitamins that prioritize candy-like appeal over actual nutrition. Some children's vitamins on the market today contain up the 7 grams of sugar per serving

and are stuffed with artificial additives and petroleum-based dyes. Haya took the opposite approach, zero sugar, zero gummy additives, just clean nutrition. Haya looked at what modern kids are actually eating or not eating and formulated around those specific nutritional gaps. Working alongside pediatricians and nutrition scientists, Haya created a powerful chewable multivitamin that packs a blend of 12 organic fruits, and vegetables plus 15 essential vitamins and minerals into every chewable. I know a lot of little kids who love high of vitamins who've been taking them for a long time.

And the parents tell me, unlike other vitamins, it's very easy to get them to take them because they taste good and they're fun to take. And you can get to decorate the little bottle with stickers that comes with the first order. And also, you don't have to feel guilty like you're feeding your kid a bunch of sugar because these are good for you. They're made of like clean zero sugar, no gummy additives, just clean nutrition and real ingredients.

Also, bedtime battles and restless nights are a struggle in your home. They have a calming, nightly chewable design just for kids, and that could be exactly what you need, especially with the new school year coming around the corner.

High as new, kids' bedtime essentials is a total game changer. It's not a sleep aid. It's a non-habit-forming chewable with L-thian-B6, GABA, and camamil flour extract, plus magnesium and calcium

and calcium that helps kids relax before bed and wake up rejuvenated and refreshed. No synthetic hormones like melatonin, no sugar, just calm support to help them recharge for everything they've got going on. You can now find Haya in person at Target stores nationwide or order directly from their website with a special deal we've worked out just for you. Get 50% off your first order on any of their products. To claim this deal, you must go to Hayahealth.com slash crooked. This deal is not available on their regular website.

go to H-I-Y-A-H-E-A-L-T-H-H-C-C-C-C-Crook and get your kids the full-body support they deserve. Pod Save America is brought you by Wild Alaskan Company. The difference between low- and high-quality seafood is enormous, but it's also one of the things that's tough to accurately assess for yourself in the grocery store. Wild Alaskin Company offers the best way to get wild-caught, high-quality seafood delivered to your door. On your schedule, each Wild Alaskan box comes with individually portion filets, vacuum-sealed, easy to prep, and great for any meal, no matter how quick or elevated.

is quick frozen fresh from the Alaskan waters, which helps lock in freshness, texture, flavor, and key nutrients like omega-3s. It's fish you can trust with no GMOs, antibiotics, or other additives. Every order supports sustainable harvesting practices and Alaskan fishermen whose history is tied to the region and practice. I love Wildlasking Company. I get salmon from there and other fish, and it's just a really nice thing to know that it arrives, you can put in your freezer, and then if you're like, you know what, tomorrow

I'm going to cook, before you go to bed, you throw the fish in the fridge so it's frosts overnight and then you're good to go. You can also do in the morning before you go to work. Like it's just, it's an easy thing to know that when you, you don't have to buy fish at the supermarket and not being sure what you're getting sure what you're getting. And also it's like in your car and it's like a whole fucking thing. Instead, you can just have Wild Alaskin Company and it's great. Wild Alaskin company is so confident that their fish is the best. They offer a 100% satisfaction and money back guarantee. So you can try your first box risk free. Go to wildelaskin.com slash cricket for $35 off your first order of premium wildcaught seafood.

that's wildelaskin.com slash crooked for $35 off your first order. Thanks to Wildelaskin company for sponsoring this episode. Do you speak just a little more detail about the search data, like what you were looking for, how many people it was, and then we can get into what you found. Yeah, absolutely.

So anyone that's ever like talked to me in my life knows that I am like particularly obsessed with search behavior as like a proxy for understanding kind of where voters are at. I think it's like the single most interesting, thing that you can do to observe where someone's head is at. It is literally like knocking on someone's door and then being able to say to you, hey, I'm really interested in immigration. You can then have a conversation with them about immigration. They're like quite literally telling you what they're interested in. And so for a long time, I've just been really interested in how can we, you know,

do more to understand people's search behavior. How can we use it? How can we learn from it, etc. So we worked with a third-party data provider to acquire and pay for a set of people who basically opted in to share their behavioral data with this data provider. So basically, like, you could go to your YouTube account right now and say, download my history. You could do the same thing on Twitter. I actually highly recommend it. It's, like, very entertaining. You, like, download your history, upload it to Claude, tell it to give you some analysis. It is, like, better than therapy in a lot of ways.

Or at least maybe Claude will tell you to go to therapy after reading it. it. Yeah, potentially. I'll call it complimentary to therapy. But anyway, so we did that with a provider that was able to send us search behavior across, you know, the major kind of traditional search engines, Google being the most predominant one, but also, you know, Gemini, chat, GBT, BT, and then some of the major social media engines, YouTube, Instagram, TikTok, Facebook, etc. And so our team, it was about six million searches. We did it across two different time periods. The first was like the last couple months of 2024, and then the second time period was like a chunk of 2025. We wanted to be able to look at it into chunks. Our team spent a couple of weeks, like really, really analyzing the data.

It was like a huge, you know, undertaking to take a look at the trends. We definitely would not have been able to do this project if our team wasn't, you know, fully enabling AI to do a lot of the heavy lifting in the work. We are a very, let's lean into AI organization at priorities. And we walked away with, you know, unsurprisingly, a huge number of insights that our team is currently using to both run programs that we're running here in 26, but also looking ahead to 2028. So, yeah, happy to share more about kind of what we learned. Before we get to the actual findings, but the search, the context of search in this day and age is very interesting, because for most of the last 20 years, search meant one thing and one thing only. It meant Google. The people at,

The people of Bing would not. Yes, there was a small handful of Bing people out there who use it. And, you know, we had Netscape. We had some things earlier. Bing was actually a highly, highly effective fundraising platform on the two presidential campaigns that I worked on. Yes. Wow. Well, there you go.

I am learning something brand new today. But it meant via Google or Bing searching the web. That's something very, very different now, as particularly as the social media. And I'm going to get to AI later on this conversation. but as the social media platforms become more places where people sort of live, particularly TikTok and Instagram. Like, how are you thinking about, like, how does that, how many, like, are people, I assume people are searching for the same thing that people used to search Google for a few years

ago, are now searching on Instagram or TikTok or elsewhere? Or are there differences in those search? Like, what did you find? Yeah, I think, like, Google is still king. Like, let's just be really clear. Google is still huge market share, predominant market share, and Google makes a majority of their money off of search, not YouTube. Like, let's just name that. They were going to do everything they can to not lose that spot in the market.

But, you know, as people are spending more and more time online, I like to think about it. And this comes into play a lot when you're putting together a media plan. As people spend more time online, they're just doing more, right? Like you are both using Instagram and TikTok and search every day instead of, picking between them. And so I think what we are starting to see is that the behaviors between these different search engines are different, which is really helpful learning, right? Like, people are going to Google to really fact check and get basic information, but then they're going to chat DBT to really, like, put together robust plans and really go back and forth, whereas, you know,

Instagram, TikTok, it can become a little bit more entertaining in nature. So I like to think about it less of a binary choice that people are making between these different platforms. Obviously, Obviously, there are trends in that direction, but it's more about how are people using these different platforms for different reasons. The thing I will say that I think is a really important piece of the learning here is, like I said, Google is still top of the list. I think our party is very interested in understanding people's relationships to chat bots and AI as a search engine.

That is incredibly important and it's changing rapidly. But people using YouTube as a search engine is incredibly under-discussed. I think that YouTube is the number two search engine behind Google. Those datasets are obviously integrated under one umbrella. And it's a place that I think we definitely don't spend enough time thinking about. And we can actually do a lot more there than we can on chatbods. When you say do more because you can you can manipulate and run ads and produce video content that appears when someone searches for a topic that you want to appear on.

Right. Like we know for a fact that AEO is a hot topic right now. people really want to understand how they can influence a search result in a chat bot. But we, you know, YouTube search has been around for such a long time. It's a lot easier to manipulate what comes up when someone searches for something on YouTube in a way that I think the Democratic Party needs to be spending a lot more time producing video than they do, you know, manipulating chatbot results. It's not a choice.

It's not a, it's a false choice. But I think we've got to spend more time with it. I am of a certain generation because I'm old where you, YouTube is not the same to me. Like, I didn't grow up using YouTube, obviously. I don't really grow up with the internet. We're going to have to discuss that here. And but like, I think a lot of people who are in politics, many of them are my age and older and our gerentocracy. Don't fully understand just like how big YouTube is that it's the biggest video platform in the world. It's the biggest music platform in the world. It's bigger than Spotify on music. It's better than Apple podcast or Spotify as a podcast platform. That it is just this gigantic thing that dominates people's lives. I don't think people really get that. But let's let's get to. Let's get to what you found about, well, when you were analyzing this data, what are sort of, what are the major insights you got? Yeah. So I think the number one headline is that people are not searching for anything political. Of the six million searches that we acquired less than 100,000 were political or even politically adjacent. The reality is, and I think the thing that, you know, keeps the team at Priorities USA up at night, but also motivates our work is that people just aren't thinking about politics.

And we have to really, if we don't understand. that going into crafting a campaign, we're really going to miss the mark on how do we actually then reach someone. So, you know, I think in doing our analysis, we found a lot of really awesome patterns around when people do search for politics. It's, it's eye-opening for our team around, how soon out from the election are people starting to search for their polling location? When are they searching for information about their candidates? Like our team is quite literally right now. It's, you know, Friday afternoon. That is like literally what they're working on to craft our 2026 campaigns.

But, you know, more importantly, it's opening our eyes, I think, to the thing that I'm the most, you know, excited about and nerding out about is this category of what we're calling politically adjacent categories that are a really awesome entry point for us to be having relevant conversations with people when they aren't thinking about politics, but they're one step away from thinking about politics. And let me just give you an analogy on this really quick because I think it's helpful at kind of landing why this is important. I like to, I spend a lot of time thinking about how can we as political advertisers compare ourselves to corporate advertisers because it's a billion dollar industry and there's a lot that we can learn from from people that are quite literally

doing similar work to us. But at the end of the day, you know, I said this to Peter, but Nike knows that you're training for a marathon and they're using that information to sell you shoes at that particular moment. Target knows that you are pregnant probably before you do. There was that story that came out on that a couple years ago. My algorithm currently knows that I'm shopping for fall shoes. And it is quite helpful to me. And I actually really appreciate the algorithm for knowing exactly what I want. The reality is that advertising writ large is like very contextually relevant and like very genuinely helpful in a lot of ways. And that political advertising is not that. Political advertising is just interruptive. We like show up in the most random moments and want to like interrupt your day.

And we got to get closer to the corporate model. We got to create campaigns that are actually, you know, relevant to someone's day. And, you know, behaviorally, I think we can take a lot closer look at where people are actually interested and then create those connective points between what they're interested in and how we can shift the conversation to politics and have a million ideas of what that looks like that I'm happy to dive into. Yeah. So I want to Like you raised this very, and I was very struck by that quote from you. I got to get started. I got to get paid by Nike at some point.

That's right. No free ads is a rule we have. Everyone who spends any time in the internet understands that your phone is tracking everything you do. And then giving you, serving you information ads about that very thing, as you point out, right? Like, you have kids. Back to school comes. It's back to school stuff. It's they know you've done you've done you've searched Disneyland on the internet.

Now you're going to get a bunch of Disney content. You're going to get advertised in a million different ways through, you know, on Instagram, through TikTok on your, you know, in your Gmail, you're going to get Disneyland ads for right or from travel comes, whatever else. If the corporate model is behavioral based advertising, what is the political model been to date? What is that based on? It's the same conversation we just had.

It's based on your geography. and your voter file score. I think 90 plus percent of Democratic advertising investments to date have been based on where you live, not as a proxy for you being close to a data center or on a university campus, but just how likely am I if I were to throw a dart to reach a black voter or reach a college student? It is based on density and efficiency for the most part. That's like how people use geography to target. And then when it comes to a voter file, it's just, you know, the ease of someone is building you a model.

They're telling you which voter you should knock and which voter you shouldn't, which voter you should mail and which you shouldn't. And digital practitioners have been told for a really long time, hey, can you target this list of voters that we think is important? There's technology that exists to be able to upload an offline list and match it to the internet. But that's just completely, you know, ignoring every online signal that you can add to make sure that that, you know, becomes, becomes relevant to the, to the end user. So, yeah, status quo is,

I would say geography and voter file are like the primary drivers of current targeting. And we're out here screaming, there's more to it than that. Is identity part of that as well, at least like in a broad-based way, like, especially if you back in the days when you're advertising on Facebook, where you would be, you know, targeting people of a certain demographic group or certain age group that you would be reaching. But it's not tied to any specific knowledge about what those people care about, correct? Yeah, I would say, like, age and demographic come into play a lot, probably more rarely than you would think, right? Like, I think when I was on the Obama campaign in 2012, we definitely had age and gender tracks. We had demographic tracks layered on top of

kind of our broadcast track. But in a small congressional race, like, I think that's a lot more rare than you might expect. And I'm not out here saying that age demo geographic targeting isn't important. It's highly critical. I think we're just saying that the next step beyond that of really trying to understand how to reach someone based on things that we're not necessarily observing offline, but instead online, we should be pushing ourselves to think in that category because it's just going to make, it's going to make us better advertisers.

Is cost prohibitive here? Or is that a problem? Is it obviously much more expensive to market as someone based on, you know, specific targeted, not, you know, information about their online presence than just targeting a group who happen to live in a certain zip code? Not in a major way. I mean, you know, like I think Facebook, their CPMs get a lot lower when you kind of give them free will to pick out who you want them to reach.

So when you're running a campaign on Facebook and you're saying, as long as someone cares about fitness, go ahead and reach them, that gives them a lot more free will to target folks than if you were to upload a file and say, I only want to target these 100,000 people. So that does impact the CPM, but in a helpful way. We have a lot of questions and we're doing a lot of research at priorities around just how you can kind of validate that data that you get on places like Facebook, YouTube. But, no, I would say cost. If someone's bringing up costs as an excuse not to do a campaign like this, they're probably

not fully grasping what we're talking about. obviously I think what I am describing here does require incremental costs from a production standpoint. I think the model we are describing certainly requires you to make many, many more ads than a TV one-size-fits-all program. But I also think there's a ton more that we can be doing on the kind of AI automation side to make the production process faster, easier. That actually has nothing to do with, you know, AI slop. You can automate the back end without making the front end of it AI-driven.

But like, can we live in a world where using the voter file, we've identified a set of target voters in a swing state, and we now know that this swing voter has been searching on, this is creepy to a lot of people, has been searching on the internet about cyclospora and food recalls. And then could you then, could a well-funded, well-run, digitally forward presidential campaign, then target that voter with an ad in 2028 that said, You know, in the Trump Vance administration, presuming J.D. Vance is nominated in the Trump Vance administration, they cut all the funding, which led to these Bluebird recalls or let us recalls or whatever else. Like, is that something that can be done at scale? So Google doesn't allow you to use offline data to do targeting. They got rid of that. I think it was 2019. But everything else that you described, yes, would be doable. I think, though, the reason that I care a little bit less about the front end, like here's the

voters I particularly care about is think about how a model is built, right? You take a look at someone's vote history and then you ask them a bunch of questions in a survey to infer, you know, how they're going to, you know, how they're going to vote and you know, you apply that to a broader set of voters and you make predictions about how you think someone's going to vote based on them being male, going to church, reading books on the weekend, whatever it is. Like there's a bunch of really smart data people out there that are building really smart models. I think what we're saying, is in addition to asking someone a survey question to inform a model, what if you just observe what they do instead? How can we get to a point where we are not relying on offline data

to predict where someone is? And instead, just, you know, if someone's searching for whether their lettuce is contaminated, I don't really care how they voted in 2020. I think that ad is probably quite relevant to them. And you don't even pay for an ad until someone clicks on it. So why don't, why don't we just try and get rid of the kind of front-end marker to begin with. Does that make sense? Okay. Let's get into just sort of, all right, so we got a bunch of these voters who are less politically engaged.

You've looked at, you know, of the six million searches you had, you know, the overwhelming percentage of them were non-political. What are these theoretically non-political people doing on the internet? Where are they? What are they interested in? Yeah. So one of the most interesting, remarkable findings that we saw is just how much people are you know, searching for, you know, what we're calling kind of public benefit programs, people searching

for how to download a FAFSA form, how to apply for Medicaid. What is Medicaid called in my state? Those are all, I think, it's a very, like, large percentage of the searches that we are seeing in that kind of politically adjacent field. And it's a really important signal to us that someone who is searching for a FAFA form should probably fall into a particular bucket of communication that looks different than someone who's searching for whether their lettuce is contaminated. I think that was the most remarkable finding that I saw that I think we're, you know, eager to act on. Pod Save America is brought to you by Hymns.

ED is more common than guys talk about. The good news, getting real treatment is simple and through Hymns is 100% online. Hymns connects you with licensed healthcare providers online, giving you simple access to legitimate ED treatment options from home. No awkward appointments, no pharmacy lines. Just complete a simple online. intake and a provider will review your information to determine if treatment is right for you if prescribed your treatment ships directly to your door. That includes sildenafil,

also known as a generic for Viagra, available through HEMS at up to 95% less than the brand name version. And if that option isn't right for you, there are additional treatment options available so you can find what works best for your body. It's straightforward, transparent, and designed to make getting care feel easy. To get simple online access to personalized affordable care for ED and more, visit Hymns.com That's H-I-M-S.com slash crooked. Hymns.com slash crooked. Not available in all 50 states.

Prescription required. See website for details, restrictions, and important safety information. Sildenafil is the generic version of Viagra. Viagra is a registered trademark of Vietras, specialty LLC. Hymns is not affiliated with or endorsed by Viatris. The grand mystery, right, is like, I know where you get your information or you know where someone like me gets my information. How, but all these people who pay no attention to podcast. politics, like they don't actively pay attention to politics, are forming political opinions,

right? Or even if they're very broad, they have a general sense of how they feel about political figures, political issues, things that happen in the news. What did you learn about where those people are actually, like, where is that information coming from? How are they running into it in this crazy media ecosystem? For sure. So, you know, I think a stat that I like to throw out and remind myself of every day is that 54% of voters writ large are just not paying attention to news at all. They're trying to avoid it mostly. And so if they're not paying attention to news, how are they experiencing politics? I think we like to ask ourselves that question to really shape kind of what are the campaigns that we're going to put together to reach them. And, you know, I think a lot

of democratic investment over the last couple of years has focused on how do we make the news sexier, easier to reach, easier to access. There's nothing wrong with that endeavor, but I think it's going to have its limits if there's just a, if there's a swath of people that just really don't want a page into news at all. And so if we want to reach those voters, we have to do it in a different way. We have to have different cultural entry points. I think you hear Democrats say a lot, like, we need to be culture first. I'm frustrated that I don't actually feel like we have a next step to that question. I've never actually heard adequate, you know, ideas of what that looks like. And so I think part of the research that

priorities has been doing over the last two years is sparking ideas of what does that actually look like in practice, knowing that beauty and fashion, true crime, parenting, those are three categories that are reaching a really large set of voters who are both swinging and, you know, switch from party to party, but also don't pay attention to the news. What would it look like for us to create a campaign where we develop true crime content and then use that content to like talk about Epstein or whatever it makes sense. Like it's about really using, you know, using this data to uncover those other kind of non-news oriented entry points for us to then have conversations with people on their terms. Parenting is a category that I think is like highly interesting and super relevant to

what we're talking about. What would it look like for us to create a track of always on communication that targets people based on the fact that their parents that talks about issues that are, you know, highly relevant to parents. I think when we talk about, you know, the March to 2028 and the fact that we're getting going early, that doesn't mean that we need to be advertising all the time. We're trying to advertise in moments that are highly relevant. Back to school. Great opportunity to, you know, talk to parents about about how Trump's policies, advanced policies, reviews policies are personally impacting them. I think our goal here is to be relevant and not necessarily always on.

When people are, whether they're interested in true crime or gaming or fitness or whatever else, like, where are those voters hanging out, right? It's like we're not in a world where they're going to truecrime.com on the web to read about it. So is it they are in social media apps and they have both followed and are being fed that sort of content? You know, like just I want to try to like get a little granular about where these people are hanging out. And I think why it's particularly relevant is, you know, we have, we have. been absent in these spaces, obviously. That's why you're doing the work you're doing. But if you look at like health and beauty, for instance, or wellness as a space, like we were absent, but at some

course of time since, you know, 2016, 2012, that category, which would have seemed demographically to be very much ours, you know, a very, you know, young mom, younger women category became like the birthplace of Maha. So I'm just trying to understand, like, where these people are, like, where they actually hang out when they say they're like, they're into, fitness or true crime or whatever else. Yeah. I mean, a lot of this research is very baked in YouTube. Like, let's just name that.

I think there's priorities to spending a disproportionate amount of time and money really trying to understand YouTube behavior because we know how easy it is to just light money on fire on that platform. And there's a ton of upside and better understanding, you know, how people are using the platform. Let me give you an example of a program that we ran earlier this summer that I think will help answer the questions. So we, from our research, as we talked about,

young men super online, super swing up. They're swing voters. They, you know, swung from Trump, away from Trump, away from Biden in from 2020 to 24. From our research, we knew that young men are interested in podcasts, finance, gaming, fitness. I think those were the categories. What our team did is we ran this campaign in Ohio earlier this summer where all of the ads were about John Houston.

But each of the ads was highly. relevant to those categories that I just laid out. So you had like a guy lifting weights talking about John Houston running, you know, next to game to fitness content. You had someone who was an actual gaming creator talk about same topic in the context of, you know, what a normal kind of gaming piece of content on YouTube might look like. Part of the theory here, it's, it was actually twofold. It was one, our ads should natively look like the content that someone is otherwise watching on YouTube. Ads actually, you know, in today's world, it's actually better for an ad to blend in than to stand out. You kind of want that nativity. There's, there's benefit there. And I think

we had, we had some research from previous cycles that that proved to us that that was, that was a good theory. But then also, I think more importantly, anyone on YouTube is kind of, especially a young man, is kind of like two clicks away from right-leaning content at any given point. In part, because of how much the right has invested in just like having an always on content machine and being present in all the places that we are not, such that if you're watching fitness content and then you skip past the next two videos, you're probably going to be on a piece of content that's a little bit maha in nature. And so we had run a campaign a couple cycles ago where we just had this theory of what would it look like to just run ads on the most right-leaning content that exists and see what kind of

persuasion effect we can have, assuming that just because someone's on Breitbart, doesn't mean they probably chose to be there. They may have just ended up there because of the algorithms. And our theory proved out. Like we were able to keep, you know, retain these voters at a pretty high rate just by running positive pre-roll ads before someone, you know, was watching that bright bark content. And so the team was kind of like, how do we do this in a slightly more expanded way, not necessarily explicitly right wing, but these kind of right leaning places that you might not,

you know, expect on on the front end. And so those five categories are the ones that the team kind of proposed as let's interrupt these, let's interrupt this user's journey with ads and see what happens. And the theory proved out, right? Like we were able to effectively, you know, stop from one from, you know, getting to that right-leaning content. A lot of the times someone was like opening up a new tab on Google and searching for more information about John Houston, which, you know, it did the job of interrupting that person's

flow. So anyway, all of that to say, I think we are disproportionately. spending a lot of time thinking about how to find people on YouTube in particular, just because it is such a, it's a high, it's a high density, very efficient place to reach people, but it's a very easy place to burn money. And it's important to note that we talk about TikTok all the time, but at last night's act, you still cannot run political ads on TikTok, correct? You cannot run political ads on TikTok. That is true.

And so it's just out there destroying the world and we can't intervene. There's certainly interventions that you can do that are not, you know, explicitly renting attention. What is that I like my nerdy way to talk about advertising is you're renting someone's attention. We can earn attention on TikTok, but it's not something that we're going to be able to, you know, get in a short window. Let's talk a little bit about AI chatbots because in your survey, I saw that just about half of voters are using an AI chatbot once a week.

And, you know, obviously Google is still king. You pointed that out. But even Google is changing in the way a lot of people are getting their search results because they are coming in these AI summaries, which is interrupting the search ads that we've been spending gazillion dollars on for decades. But so as we think as AI chatbots grow as a search engine for people, how should Democrats be thinking about influencing those results? Like we had a playbook. We had a, both a, we had a advertising playbook for Google search, and we had an organic search SEO playbook for how to get our content near the top of the results. What the hell do you do when it's a, you know, a handful of very different, very independent AI chatbots?

Yeah. First of all, there's a lot of really smart people who are tackling this head on. I think the folks that run for something, Cox AI, are thinking really extensively about how do you, A, understand and then B, manipulate a chat bot result. It's really, you know, AEO is just SEO 2.0, right? It's just understanding what AEO is, just so people know. Yeah, it stands for answer engine optimization. And it's basically how do you manipulate what is going to show up in a chatbot result?

There's a lot of really smart people out there who are thinking about how do you, you know, understand what's showing up in a chat bot result and then B, how do you actually, you know, change it. You change it by updating your website. You change it by pitching stories to the AP or posting shit on Reddit because that's a place that's often getting picked up by chatbots. All of this is going to change incredibly fast. And anyone who is kind of going out there with best practices today, the best practices are going to change two months from now, six months from now. And so I think it does require a lot of urgency and a shift in conversation across our party that if we always had to invest in digital infrastructure,

Now is the time to like double, triple, quadruple down on that effort. For us to like really understand how quickly the internet is changing, like we have to be investing in innovation to make sure that we're able to stay on top of these trends. I think the work that that priorities is doing though in the kind of chat bot space beyond just, you know, understanding what a result is and understanding how you might manipulate that result is really understanding when and why are people turning to chat bots. What is the kind of driver of why someone's using it? And how does that change how we should be thinking about intervening and better understanding

their media journey? And so I think, you know, the research that we did over the summer, I think was highly, it was incredibly helpful at uncovering that people are not necessarily searching for information about a candidate as much as they are searching for information about policies and issues in a much more granular way. You know, I think as, you know, AI, I don't know if you feel this person, But I think as I'm, you know, scrolling TikTok and on YouTube and seeing content in my feed that feels a little like, is this real? Is this not? Like I am personally opening up Chad GBT, Claude and saying, hey, can you fact check this for me? There's a, I think a ton of, it almost feels like the Spider-Man meme where we're, because of the insurgents of AI, people are using AI to check if things are AI. It doesn't seem great for the world. But yes, that's where we are. Yeah.

Pod Save America is brought to you by Helix. I love my Helix mattress. I have a Dawn Lux. I actually took the quiz because I am a stomach sleeper. I'm trying to become a side sleeper and kind of rotating, but I love the bed. It's really comfortable. It helps make sure you get a good night's sleep. And you can take the quiz and see which mattress is perfect for you based on your personal preferences and sleep needs.

They have over 20 mattress models so you can find the perfect one for you. Helix delivers your mattress right to your door with free shipping in the U.S. The Happy with Helix Guarantee offers a risk-free customer-first experience designed to ensure you're completely satisfied with your new mattress so you can rest easy with seamless returns and exchanges. Plus, Helix offers a 120-night sleep trial and limited lifetime warranty. So as I said, I have it. I think it's great. It's so nice knowing that like at the end of the day, you know, right before bed, you get into super comfortable Helix mattress, throw in a streaming show, you know. I'm watching Furious.

There you go. It's awesome. It's a great crime drama. You know, this is not an ad for that, but you should check. that out too. Hey, everybody, get a Helix mattress, get comfortable. Watch Furious on Disney Plus, you know, but I just think it's a great mattress and you should get one. Helixleep.com slash cricket for 27% offsitewide. This offer is exclusive to Potsave America listeners. That's Helixleep.com slash crooked for 27% offsite wide. Helixleep.com slash crooked.

We're in the Village 202026 selection. You guys are focused on, among other things, trying to define the 28 potential nominees, most notably J.D. Vance and Marco Rubio. talk to me about why start that process now. Yeah, absolutely. So I've been working at a super PAC for almost a decade now, which is crazy. And I'm very much of the opinion that the way that our party uses super PACs is outdated and not to the maximum that we can and certainly not to the maximum that Republicans do. And the way I would describe that is quite simple.

I think as much as this role of a super PAC is to raise money that a campaign cannot raise, a.k. we are able to raise beyond the contribution limit that a campaign has. We also have time that a campaign doesn't have. And I think really leaning into that time element is a thing that I feel really important as the role that a super PAC can play in today's world. And so I think in the aftermath of 2024, when we're sitting around asking ourselves, why the hell did we just lose that election and what can we do differently next time? Our team got to work on a couple of things, including, you know, at a high level, how can we do a better job of listening to the internet to be able to drive decision-making?

We can talk about polling and all the reasons by polling is flawed, but, you know, priorities is very much of the opinion that with the, you know, availability of AI, we can observe qualitative data at such a larger scale in today's world that we might as well lean into it. It's exactly how the Republicans are making their decisions, right? Like we like to look at the Kamala is for they, them ad. I think it's a perfect example of how the Republicans said that ad has likes to it. It's sticky.

It's doing really well online. Let's put money behind it. Whereas on the Democratic side, we have to like let an ad go through an entire polling rigmarole in order to take it seriously. Anyway, all of that to say, priorities got to work after 2024, to build a tech stack and do research to really understand how can we be leaning into listening to the internet to understand what's going to resonate and what's going to break through. At the same time, you know, we recognize that the March to 2028 is going to be long and

it's going to include a lot of Democrats fighting each other. I think if this summer was indicative of anything, it's that primaries are challenging and they can be really helpful, but ultimately they're tough. And we're about to go, we're about to face that on a level that is going to be different than what we just experienced over the summer. And I'm really worried that our party has not been, you know, we have not gone through an election without Trump on the ballot in quite some time. I think our muscle memory there is quite weak.

We're going to be running against someone new no matter what happens. And the Democrats are going to be fighting each other. So the team at priorities was like, what can we do to be the most value ad heading into 28? Let's get going now. Let's actually start to think about what it looks like to define our eventual opposition so that by the time we have a nominee, they have a slightly less uphill climb to be able to go after and, you know,

compare themselves to the Republican. So that's the work of prologue. We're really excited about it. Our tech tech all feeds into kind of what we're doing. It's not going to look like a traditional advertising campaign. It's going to be always on, but that doesn't mean we're on all the time. And we have a million ideas that we're excited to dive into. So the, I am in after 2017, or after 2016, in 2017, one of my strong views was that we

should take a lot of the money that was sort of flowing. into democratic politics in the wake of Trump and be always on, right? And to like, like, not, like you say, not always on like we're going to run ads for the rest of time. But in moments, high leverage moments, you would run advertising to shape that impression, you'd shape that moment where people are paying attention in their minds, in our favor. And I was, I like that idea because I believed correctly, I think, that we were just so outgunned in the media space that we were competing with Trump, competing with Fox News,

screen with Facebook at the time. This is very powerful entity that was a sort of a conduit for MAGA adjacent messaging. Like we needed to, you know, put some like messaging thrust behind the things that happened. I went around. I sold this lots of people. And the thing that everyone came back to me to say was, and smart people, was there is no evidence that early ads work. And then you're basically recommending that we are lighting money on fire. I was never fully convinced of that, but I also was incapable of convincing the people who said that to do something different. What is your response to that? I mean, does that logically make sense? Well, I mean, they will throw a bunch of political science research at you that people have very short memories

or are not paying attention at the time. And so it doesn't, the impressions are not worth the money, I guess. I mean, research also didn't show that smoking caused cancer until very recently. Fair enough. Fair enough. No, I just, I think that I really appreciate the folks in the academic space for pushing us to be rigorous. But at the end of the day, like, I think it is really important to name that political advertisers and what we are trying to do when we are trying to win elections is fundamentally different than any other, you know, communication challenge, marketing challenge in the broader world, right? At the end of the day, not to bring Nike back into it, but like Nike spends millions of dollars a year to think they're

a really cool brand. They run ads up around every sporting event to make you like. them, but ultimately, they're trying to sell you shoes. And they get to use shoe sales as a proxy for whether they're just do it ads or working or not. We don't have that. We do not have proxies that we get to use to understand whether someone is mind is being changed, right? Like, that's why surveys are, you know, we invest so much in surveys because it's the, it's the closest we can get to be able to being able to ask someone, were you persuaded by this? Did we, did we sell you the shoe. And, you know, that I've spent a huge chunk of my career trying to get to a point where we can like zero in and understand how to, you know, properly measure a persuasion effect.

And it is, it is quite challenging and it is probably the hardest thing that I've had to, you know, face in my career. But the reality is, you know, the Republicans are investing close to a billion dollars over the last decade in always on infrastructure and communication. And I it is quite obvious that it has paid off, right? You look at, you know, the example I like to give is Trump, after the shooting, you know, within an hour, he had people on the internet talking about the ballroom and how the shooting necessitated the need to build a ballroom. We don't have that amplification, you know, program on our site. There's a lot of folks that are working really hard to build it, but it just shows the long-term payoff of what always on infrastructure

looks like. Meanwhile, you know, Democrats sit around and our amplification is like crossing your fingers in toe hoping for a retweet from Barack Obama. If you want to look at that and say that our situation is better, I just don't buy that. I think at the end of the day, like art and science have to have to come into play. And we have to be able to use our judgment and be able to say that, you know, always on communication, it has to matter, right? Like we can see it playing out in front of our eyes. But I think a critical reason that we are doing things differently. And I empathize with the criticism of Always On and that sometimes you can look at the way it plays out in a budget and just think, well, there's no way in how we can afford that. You can't afford to be at the impression levels

now that we would want to be at in the fall of 2028. But that's why the priorities approach of really using behavioral targeting as an entry point to have these conversations is I'm so excited about this program because it's the first time that I've really looked at a program through the lens of how do we reach people, not necessarily during a time when they're registering to vote or when news is breaking, kind of those natural political moments, but instead during the moments of, hey, I'm Googling whether my lettuce is safe. Like, that's the perfect time to have a political conversation. It doesn't mean we're going to be on all the time. In fact, we're not on right now, but like the next time something you know, politically relevant happens, you'll see us go up. All right. I want to look at some of your ads

because they are, they look nothing like a typical Democratic ad. I want to start with one about J.D. Vance. Let's play it. Hi, I'm J.D. Vance. As vice president, I cast a time-breaking vote to cut your health care. And I've been helping our great leader create tremendous wealth, which has been trickling down. There's a little more. To you, Rome wasn't built in a day, of course. But ultimately, you're going to feel the benefits. And your Medicaid? Don't worry.

That's just immaterial minutia. Anyway, get back to work. You're almost like for your second job. And President Trump and I have to get back to redecorate it. Okay. I believe that was Donald Trump taking a leak on J.D. Vance's head to make a point about Trickadoon economics. I love it.

Yeah, so go ahead. Yeah, talk to me. Just talk to me about the concept of the ad, who you showed the ad to and what the reaction to it was. Yeah. So, you know, my friends at home think being an executive director is like out there talking to rich people. It's actually just getting a slack from your CMO that says,

hey, how do you feel about an ad with some piss in it? That's quite literally the reality of this job. But, you know, I think this ad is about recognizing that if you can't actually get someone to watch your ad, it doesn't matter how well it tested, right? The number one thing that we need to be aware of is the skip button and the second screen that exists and the fact that people are just naturally going to tune out of what we're saying. So if we want to be able to actually get our message across about gas prices, we need to deliver it in a format that's going to capture people's attention. This is just one ad, right?

I think it's really, you know, the media responds really well to, hey, look at this one ad and cover it and get some media attention on it. The reality is like prolog is going to be thousands of ads over the next two years. We're going to throw shit at the wall, see what sticks, and then move on with our lives. This ad is one example of something that we tried. It actually did get a ton of attention. people watched it all the way through. I think we were really excited about the metrics that we saw. People want humor.

They want things. They want politics to be a little bit less serious. And so that was our attempt to kind of play around with those insights. So, you know, the typical way we have been measuring the efficacy of ads and messaging and over the last, you know, decade here is persuasion effects, right? How much does that based on a, we're testing the ad in front of tens of thousands of people, and to what, how many points did this take off of Trump's approval or how much did it move vote choice in our favor? And these are all in polarized environment, very small numbers that, you know,

we're like an ad could be three times as effective as another ad and it's because it's moving approval three times, that one, three points, not one point. Yeah. What is your metric of success here? Is it completion rate? Are you testing persuasion effects? How are you thinking about it? All of the above. I think we are of the opinion that there's no singular metric that's ever going to tell you your ad is working, you have to take a really wide lens and, like, take it all in. You know, we didn't do creative testing in 2016, I think in the aftermath. I was really excited when the idea of creative testing kind of came into the ecosystem because I was like, okay, like, we can finally like understand before we put money behind an ad, how is it going to do?

And I don't know if you, you guys remember, like, 2016, the biggest thing was like so many of our ads were causing backlash. They were actually, like, moving people in the wrong direction. And that was like such an eye-opening thing. And so I was always of the opinion that, you know, know, you can take a look at a creative test result and weed out the ads that are going to do harm and actually move people in the wrong direction. And then if you have five ads at the end of the day that move people in the right direction, put them all online and see how they do. Right. Like let's compare offline and online metrics against each other to really understand, you know, how they're going to both be complimentary and maybe disagree with each other. Not to shout out the

team on the Newsom campaign, but they did this really well during the Prop 50 campaign. I think their whole theory was, we're going to test our ads. We are going to see what works in a kind of lab environment. We're going to see what horse race looks like. But instead of, you know, picking the top ad that has the highest horse race effect, we're going to just going to put everything out there that doesn't do harm and then see, you know, on YouTube, which audience is which ad move. I just give them a lot of props for kind of taking that theory. It's in strong alignment with kind of the priorities theory of the case. And it obviously had a really awesome impact in that race. But yeah, I think part of our, our theory, and this is one of the other tools that we're actually in development on right now

is what does it look like to amass a much larger set of signals to understand how something might resonate online beyond just how something tests in a creative test and how something pulls. How is an issue being talked about on a podcast? How is it being talked about in a comment section, in a search trend, at the doors? There's a million different data sets that we have available to us that we can be listening to. And part of what priorities is doing is amassing all of those into a centralized view so that we can take everything in and make strategic judgments accordingly. All right. Let's look at a ad about Marco Rubio. And then let's discuss it. Okay. Think America can't get any more expensive. Did we can't afford another pointless war?

Well, the call isn't yours to make. It's kids with a little Marco invasion action playset. Boots that are too big promises he can't keep. going to be a pro on constantly. He starts the wars. We pay for it. Social security, not anymore. Medicare? Bye-Fi.

The toy that sends wheel boys off to war. The Little Marco invasion action play set. I love it. I love Little Marco. Similar vibe to, they all have a South Parking, South Parkian, I guess is their term, vibe to them, which I do like. Who was the target audience for this ad and how did it do? Yeah, so generally speaking, priorities is always kind of trying to go after the young voters

and voters of color as our primary audience. Looking ahead to 2028, our team has identified just over 10, just under 10 million voters across the battleground states that kind of fit into that category. So that is, you know, the primary audience for this. I think for this particular creative brief, we were also quite interested in having an effect with young men in particular, just because those are, that's a tough audience for us. You know, you, I heard you say to, you mentioned the YouTube comments earlier that you read them and that they're actually sort of influencing some of the

creative decisions you're making, which is a far cry from, which is probably causing data nerds within the party to lose their mind that we would look at YouTube comments. But one thing you said to Peter Handy was that the YouTube comments about J.D. Vance and Merca Rubio were different in one very notable respect. Could you say what that was? Yeah, for sure. So let me explain quickly why we care so much. about comment sections at parties because it does cause a lot of stress, but I do think it's super valid. So we did a research project last summer where we interviewed a bunch of young men on how they watch YouTube. And by far in a way, the biggest takeaway from that research project

was just how reliable these young men are on comment sections. They watch a video, they jump to the comment section. They really want to understand what their peers think. They really want to understand if there's agreement or disagreement. It is like peer validation that is like happening in front of our eyes. And it is like, you know, people like to say the internet is not real life. Comment sections are not real life. I'm not here to say that people who are commenting on videos are representative of voters, but it is an environment where conversations are happening and we need to take it seriously and learn from it. So that is why we care so much about observing what is going on in these comment sections. And what we saw when we were crafting, you know, our initial set

of polling on Vance and Rubio in particular, you know, our team was going about, you know, picking out the messages that you would, you know, pull out when you're pulling together a poll. But we also spent a lot of time kind of scouring the internet to understand what are the narratives that are kind of naturally occurring around these guys. What are the late night host saying about these guys? What are comment sections? What are podcasts saying about them so that we could, you know, have a little bit of an internet first lens to the narratives that are appearing in our polling. And we found that while comment sections about J.D. Vance were often, you know, quite negative. definitely poking fun at him.

People are not afraid to poke fun at JD Vance. Marco Rubio's comments were a lot more positive. There's just not a lot of criticism happening organically about him online. And it's actually why in rolling out this program, I think our team walked away feeling a lot more anxious about kind of what we have in front of us to define Marco Rubio just because, A, he's super undefined. B, people are not willing to, you know, go super negative on him online. And C, like, in rolling out this program to the press,

the only person anyone wants to talk about right now is J.D. Vance. People don't want to talk about Mark or Guibio, and that's a problem. The J.D. Vance thing is interesting because I think a lot about political messaging when you're trying to define a person is you kind of want to memeify them. And you can either be memeified in a cool way or in a bad way. Trump may be the, Trump may actually be the one person to resist meming because he kind of is a meme of himself, who sort of developed his own meme in the 80s and it's just like live with it since. But, you know, it is what it's like this is even before we even talked about me. It's like that's what happened to John Kerry with the the windsurfing and the you know the I was four before he was against it.

Like they he was nemoified. He lost Al Gore even before that. We know who a man who was chosen by Bill Clinton to be on the ticket because he was thought to be the most ethical and moral person in the Senate. Then by the end was turned into a giant liar who claimed that he invented the internet and was so boring. He got memeified right. And so like van. seems like I always think about when we like you can I'm sure if we do message testing and we ask the million voters media focus groups like like you would want to say they're bad on affordability but that's not like a it's not a culturally relevant message or culturally relevant message is that jd vance is a goober I actually looked at a creative brief right before we got on just I wanted to

remember um our our creative brief literally describes like what we think the meme of these guys should be um for jd vance it's the flannel shirt doesn't fit him anymore um right he he grew up seemingly middle class. He's now in the Trump administration. And like, unfortunately, the flannel shirt that he once wanted to wear no longer zips up. It's, he's grown out of it because he's spent the last couple of years catering to the Trump administration. So what you thought you knew about him before, it's not so true anymore.

So everything you're going to see us put out over the next couple of years, like for now, like I think the meme could evolve. And part of our research is going to be to try to understand, you know, poke holes in that theory. But that's kind of the story that we're trying to tell about Vance. And it also allows us to, you know, tell that story through a climate change lens or an abortion lens or an economic lens. It doesn't require us to, you know, pit issues against each other. I think every time as a party we're pitting issues against each other, we're losing.

The meme, I think that our team has picked out for Rubio is he's a chicken hawk. He's all talk. But when it actually comes to it, like he's never served in the military. He's actually nowhere to be found right now despite everything that's going on in Iran. So those, like I said, like our team in putting together this creative brief and really distilling the research, It was less about what are the top issues and more, what's the one sentence that you want someone to describe them as when they walk away? And I think that's really our theory on narrative and messaging at priorities.

Before I let you go, like, as we are sitting here today, there is somewhere between a dozen and 400 people who are planning to run for president very soon. In those campaign headquarters, there are already people, you and I have been these people in amids in the past who are charting out in memos and slide decks what the campaign should look like, what the strategy would be, what the campaign structure would be? What advice do you have for these campaigns to learn from what you guys have done and to avoid making some of the mistakes we've made in recent campaigns where I think some of our structure and strategies have been anachronistic at best. Yeah. I mean, I would be remiss not to say that like digital has to be at the

table. Like if you are not bringing a digital person. Should you digital be the table? Yes. Right. There's all the debate. A digital is a team versus a tactic. Yes. Like I hope we are past that, but I have to say it. Right. I mean, the other thing is just like, you know, I think that I get asked all the time, well, how are you guys actually acting on this research? It's really smart. It's really insightful. But like, what, how should it change my strategy? And I think my singular answer to that question is more is more and more video. You are never going to regret making more video for the way today's internet operates. I think Kyle Thart put this study out a couple weeks or last week around just how much

Cory Booker is just churning out video and constantly putting things out. And it's like earning him a ton of awesome, you know, goodwill on the internet. You are never going to regret doing more. But I think that requires a fundamental shift in we need to stop editing ourselves on the front end. We need to stop putting things through 20 rounds of legal compliance research review and just start putting things on the internet and seeing what sticks. The internet is a place where we have a lot to learn. We need to stop thinking of it as just a distribution channel and instead think about it as a

place where we can do a lot of good in listening and observing how people are responding to our content. Daniel Butterfield, thank you so much for joining us. Thank you for nerding out with me on my favorite topic in the world. has been so much fun. We will be watching what you guys are doing to the rest of this election and the next. Thank you so much. Thank you. Talk soon. Pod Save America is a crooked media production. Our show is produced by Austin Fisher, Saul Rubin, McKenna Roberts, and Ferris Safari with Reed Jirlin, Elijah Cohn, and Adrian Hill. Our team includes Matt DeGroote, Ben Hethkoe, Jordan Cantor, Charlotte Landis, David Tolls,

Mia Kellman, Ryan Young, and Naomi Single. Our staff is proudly unionized with the Writers Guild of America East.