What Did We Learn From Kamala HQ?

I talk to Rob Flaherty, the Kamala Harris campaign’s deputy campaign manager, about why the account's viral posts weren't enough.

One of my most requested interviews is the team behind Kamala HQ. Even after the November loss, requests kept coming in. The strategy clearly left a lasting impact on those who work in the social media industry.

Today, almost six months after the final post went up on the account, I am speaking with Rob Flaherty. He was the deputy campaign manager on Kamala Harris’s 2024 presidential campaign and oversaw digital strategy. We talk about the role online trends play in politics, if candidates should be put through a three-hour podcast test, and why campaigns should invest more in digital strategists.

Our conversation is on the heels of a guest essay Flaherty wrote for the New York Times. In it he argues that Democrats lack the tools to communicate to the vast majority of the electorate who get their information through culture. He writes, “Success in communicating online most often has less to do with social media trends and tactics and more to do with doing the right thing offline.”

If it’s not obvious already, this might not be the “How did that Femininomenon post come together?” interview some might have hoped for—but I think it’s a necessary conversation for understanding the role social media strategy plays in elections.

Rachel Karten: It's been almost six months since the last post went up on @kamalahq. I'm sure you've done some reflecting in that time. What's one of the biggest learnings you have from overseeing the account?

Rob Flaherty: Well, it’s funny because the lesson I took from Kamala HQ was mostly about empowering young people. We set some pretty clear guidelines and then mostly let them cook without a ton of supervision. That was more than just, like, me saying “it shall be so”: the whole org bought into the success. Parker Butler and Lauren Kapp who really ran the account and drove it day to day had the support of a legal and communications team that let them run with stuff.

I think a second thing from Kamala HQ was the way that social-first messaging can drive outcomes. When the campaign started talking about Project 2025, it was because creators on the internet started talking about it. We knew it had viral heat, and the accounts started lifting it, and then it showed up in ads, and then it showed up in the President and Vice President’s speeches, and by the end of the campaign it was Trump’s top negative. I think it’s a lesson that messaging needs to reflect the environment that it lands in. We do a lot of stuff that goes the other way: this tests well, make it land. That was “this lands well, make it test well.”

And the third is a bit about how organic social fits in an ecosystem. I think Kamala HQ was great, and a critical messaging node. But you also need to remember the ways in which owned social properties talk to your political base: you’d see Kamala HQ because you liked what it had to say more often than not. And that’s an important part of a persuasion architecture: you start with your true believers and build from there. But…we also didn’t win, and I think the spaces we missed were in reaching the voters who weren’t as politically engaged.

RK: I think social media trends can be great for awareness but original content tends to be better for longevity. There are more ways for a candidate to be “online” than participating in the meme of the week—look at Chi Ossé or Bernie Sanders for example. Obviously Kamala HQ participated in a lot of trends and got a ton of reach from doing so. In retrospect, do you think they were effective?

RF: I generally take the view that you’re better off building a long-term connection with folks than you are of chasing the sugar high of a trend. But I also think you need to think about how the campaign used Kamala HQ: it was sort of the alt account to do stuff the candidate couldn’t do. We very rarely had Kamala or Biden participate in the meme-du-jour – only in moments when we thought it was worth doing. Kamala HQ started as a way to do negative, but the team behind it sort of grew it into a place that let us do a More Online voice so the candidate could stay the candidate. Every person who does digital runs around talking about authenticity, and for good reason! But authenticity means “this plausibly is a thing a candidate would do” not necessarily “this is a thing that would play amazingly online.” You’ve always got to balance those two things.

RK: I was impressed with your very loose approval process for Kamala HQ when I first read about it here. Can you talk to me a little more about it? I often say that lengthy approval processes ruin good social strategies.

RF: Yeah, look: talent was the most important thing behind Kamala HQ. You had a group of a dozen brilliant folks who were basically all under 30 who were constantly digging for good clips and making good content. The thing doesn’t happen without Parker Butler and Lauren Kapp and their team.

They were also able to basically exist unencumbered: there was a signal chain and a five minute warning where more or less anyone from comms, legal, research, or otherwise could object. I think we set a good culture of establishing that you only flagged if it was existential. And the teams fuck up rate was basically the same as if they had 19 lawyers and the candidate looking.

I think the overall thing about this was about A) hiring people with judgment you trust, noting that judgment and experience are not the same thing B) giving them clear parameters and C) giving them room to run.

RK: On social in general, we’ve seen follower count matter less and less. Social platforms are entertainment platforms, and what matters most is breaking through the algorithm. How has this shift impacted the way politicians should show up online? For example, I keep getting served videos of Pete Buttigieg on Andrew Schulz’s podcast—despite not actively following Pete or the podcast.

RF: Yeah, I just don’t view follower counts as a materially important metric anymore. I think reach and qualitative stuff about how it’s playing in comments is the best way to judge. I usually tell people “it's mostly about how you interpret the weather report.”

I also think this speaks to something pretty important: the internet has made culture a honeycomb of subcultures. They sometimes overlap, but not always. In the same way candidates troll around the country doing interviews in different media markets, you’ve got to think the same way about media appearances in different “areas” of the internet. In an era with no broadcasts, you’ve got to collect narrowcasts. I think a lesson from me from ‘24 is that volume matters: Trump did 30-something interviews to one audience. We’d do one or two to a bunch of different audiences. I think we were right about the width – our coalition is different than Trump’s – but we needed to focus just as much on depth.

RK: Speaking of, should all potential candidates be put through a three hour podcast test?

RF: Yes. Sort of. I am less long on the idea that every candidate forever is going to need to be able to do a big long podcast and sit for hours and hours. That’s a trade off of the candidates time and how valuable the podcast is. But the general principle — that you need to be able to present conversationally and not disciplined-ly (not a word, but lets go with it) – I think is very real. We train candidates to repeat points, and to pivot back to core message. You obviously still need to do that, but the bar of expectations around a candidate holding real, deeply-held beliefs means that you need to be able to deliver it conversationally. So I care less about time and more about your ability to say it like you mean it.

RK: From look-a-like contests to surprise Lorde concerts, it’s been interesting to see how much offline activities can fuel the online narrative right now. In your op-ed you say “Success in communicating online most often has less to do with social media trends and tactics and more to do with doing the right thing offline.” In the world of politics, I think about Senator Chris Van Hollen going to El Salvador to visit Kilmar Abrego Garcia and Zohran Mamdani yelling at Tom Homan about the deportation of Columbia graduate Mahmoud Khalil. Over the past few months we’ve watched this realization happen in real time for Democrats. How will this (or should this) impact digital strategies?

RF: I think there are a lot of lessons you can learn from what happened in 2024, but I would hate the lesson to be, like: you have to have a vertical video strategy. That is both true and a little too small-thinking. I don’t think a candidate doing TikTok trends is going to move the needle or make the connection as much as the sense that you’re doing something offline. Trump doing McDonalds seemed cringey at the time, but it was actually brutally effective because it showed him doing something offline that drove attention online. It’s sort of the same thing with Van Holland or Zohran or Cory Booker. These are big moments that you create that generate social coverage. I’d take a candidate with no TikTok presence but a knack for doing big, bold attention getting, story-advancing stuff over one who’s very online but mostly yapping. I think it’s important to remember that people still remember that our political leaders are political leaders. Yeah, they’re creators in their own way, but I think you have to actually do things in order for your content to work.

RK: Can you teach a candidate to care about social media and digital strategy?

RF: I think so? I think this obviously changes as more people who grew up online natively get elected. But there’s a fundamental thing here that everyone sort of thinks the media environment they exist in is the media environment everyone exists in. If the people making decisions all don’t use YouTube regularly, they’re going to have a harder time seeing its importance over, say, the Wall Street Journal. So part of the challenge in getting folks there who don’t live in that environment is showing the data, articulating the value, and I’ve often found showing them the qualitative feedback is helpful (eg: some senior staffer’s kid saying they liked it). But I also think there are folks like Mark Warner and Tina Smith and Tim Kaine who have shown some range in this space despite not being “digital native.” I think folks can get the hang of it, they just need to see the importance of it first. That’s sort of the educative work that digital folks need to do, which can often be hard because they’re often the youngest person in the room.

RK: Your op-ed argument has a lot of clarity—but you were the one in charge of a lot of this strategy in previous roles. What’s been the biggest challenge to making what you're proposing a reality?

RF: Yeah, this is a fair question. Look: we lost. I’m proud of a lot of what we did, but there’s no version where we can call it a success. If I could do it again, there’s a lot I’d approach differently.

That said, I think you have to look at the shift: nationally, there was a 7-point swing toward Trump; in the battlegrounds, it was closer to 2 points. That tells you that marketing and organizing had an impact — just not enough. And in this business, you don’t get a t-shirt for "close." We didn't win.

To me, the bigger lesson is about how campaigns organize and communicate. For decades, campaigns were built on mobilizing existing voters. But when the challenge is opt-out voters — people who don't even see politics as relevant to them — sending strangers to knock on their door isn’t intuitive outreach. We need to rethink the tactics, not just do them louder.

But even more than that, I’m focused on the 7-point cultural shift. That’s not something a campaign alone can fix. A campaign is a last-mile marketer. If the underlying culture is moving away from you, it’s already too late. Fixing that starts long before Election Day — it’s about building a new ecosystem that connects with how people actually experience the world now.

That means shifting how we invest — not just in ads, but in long-term culture work. It also means changing who leads campaigns. Digital and cultural strategists need to have a bigger seat at the table, not just be treated like special teams that come on for a trick play.

So if anything, working inside those campaigns made me more convinced: the old playbook doesn’t match the way voters live now. It’s not about better messaging. It’s about building something different — and starting much earlier.

Thank you for reading today’s bonus newsletter. We’ll be back tomorrow with a guest essay.