I spoke with Substack co-founder
yesterday, just before the Trump-Harris debate, about how Substack is doing its thing during the US elections.He talks in particular about how Substack’s focus on paid subscriptions rather than ads has made political debate on the platform calmer, simpler, deeper and more satisfying than in other social media, where the ad-driven need to maximise outrage, tribalism and performative pile-ons has contributed to a polarisation and toxicity in politics more generally.
Hamish details Substack’s growth since the last US Presidential elections and the key role of trustworthiness in the communities on the platform. We also have a chat the similarites and differences of last-minute candidates Kamala Harris and Jacinda Ardern, who were both elevated from roles as deputies to leaders, Joe Biden and Andrew Little, who stood down at the last minute because they believed they couldn’t win.
"When you change the business model, you change the thing that sits at the very root, the very heart of these media systems, you can get better and different results." Substack co-founder Hamish McKenzie
Here’s the full transcript for those who prefer text
We spoke three hours before the one and only TV debate between US Presidential candidates Kamala Harris and Donald Trump. I’ve edited it lightly for clarity and brevity.1
Bernard Hickey (00:01.836)
Kia ora and welcome to Hamish from Substack. It's wonderful to have you here on
talking about these US elections, which are a big deal for everyone, but increasingly a big deal for Substack. Could you tell us about how it's different this time than four years ago when Substack was a slightly smaller thing?Hamish McKenzie (00:29.437)
Kia ora Bernard and all your subscribers. Four years ago, Substack was just sort of getting going. The onset of the pandemic was a huge accelerant for Substack where we had some really strong early growth. But even then, Substack was probably only known in Silicon Valley and media circles and must have had a few million people in the ecosystem as subscribers. And today there many tens of millions of subscribers, active readers, active watches, listeners, and we've got comfortably more than 3 million paid subscriptions.
Some of the biggest names in the world in politics writing and politics coverage are on Substack and it's (politics) our biggest category. So the election is a huge deal for us and we're really leaning into it.
Bernard (01:28.588)
From down here in New Zealand, we watch what we can, but we're not really immersed in it. Could you give us a sense of, how much of the audience who are looking at the various substacks are thinking about the election, debating the election, and also the prominence and the quality of the writers and everything else you've got on?
Hamish (01:57.073)
Yeah, everyone's obsessed with it for obvious reasons. It's another carnival election in that it's spectacle and it's full of froth and vigor and color. But it's also a really important election. It's going to say something about America, like it's either going to go this way and become something, or go the other way and become something else.
There are lot, no one really knows. It's another toss up. It's looking very sort of 50, 50 -ish at the moment. No one really knows what's going to happen. So it's a, there's deep fascination, a deep interest. This election also has a dynamic that others haven't with Biden dropping out sort of halfway through his campaign and being replaced by Kamala.
And the people of Substack covering it with gusto and we have people from
, who's writing the projection models and editorial around that, to , who's sort of doing TV style coverage that he used to do on MSNBC and is now doing it here, to , who was former communications director for Donald Trump or a press secretary.And he's interviewing some of the big figures from the MAGA world. And there's a true diversity of political opinion. It's not like picking up the New York Times where you kind of know what you're to get. It's not like turning on Fox News where you definitely know what you're going to get. And it's not like social media like Twitter, where it's kind of like an all-in brawl. It's kind of like the Thunderdome where you score points, not for your thoughtfulness, but by how spectacularly you take down your opponent. It's something different.
And some of that's happening with posts, which tend to be long form and more considered, or some of it's happening in notes, which is our Twitter-like feed, but for Substack, where there is actually some productive discussion across communities on Substack. And there's also chat, which became really interesting and important on the first debate for Trump versus Biden.
Chat really popped off with commentary and chat is based around the writers. So Nate Silver might have a chat or
of might have a chat. , who's an MSNBC host still, just sent out a message a couple of minutes ago saying she's starting the chat around the debate.So we expect those chats to be extremely lively and fun today. And that's a good way to follow along. That's a little taste of the whole ecosystem view of Substack.
Bernard (04:37.004)
One of the things I was really attracted to with Substack and I've found very useful over the last three or four years is that the people who we can interact with on Substack are there for different reasons. The incentives are different within the network. And one of the things I've struggled with with various social media over the last four five years is that the algorithmic drivers have been for those networks where advertising is the main way that people have made money. Those algorithmic drivers have been really all about pushing my various buttons for that surge of dopamine or to be angry or whatever it is.
And that algorithmic driver has created a sort of a feedback loop which has separated people. You put out a piece a few days ago called ‘a less sterile derangement’, which I thought was a really clever and thoughtful way of thinking about the election and also the way that Substack operates. And you quote
at as saying, ‘we're no longer comprehensible to each other.’Could you talk about, you cause you're right in the middle of it and you must've thought quite deeply about this area of incentives and how algorithms work to drive people in certain directions. How does the pure business model of Substack create a different set of drivers and incentives, particularly around political debate?
Hamish (06:27.635)
Yeah, it's a huge topic. I think that people tend to say that the algorithms are the problem and the algorithms themselves are just equations, they're technology. It's the driver, the word driver that you're using there that's the important part.
When the algorithms are tuned to seek out the maximization of attention so that more ads can be served to people, so people spend more time in feeds and therefore have more chance of seeing ads, then you get a certain quality of discourse and content as a result. And so the kinds of things that do really well in those systems are performative outrage, polarized takes,
Scrapping and fighting on the streets and those things can be compelling. They can be fun. They can be interesting sometimes. They can be enlightening, but they don't really have that much to do with truth or trust or seeking to understand one another in fact. Oftentimes they're completely eroding those things. And so Substack does have algorithms. We have a Twitter type thing within Substack that is supposed to help people discover work and writers and journalists that they otherwise don't know.
And that is different because that algorithm seeks to maximize paid subscriptions, money for the writers. And the way that's important for us as a company, because that's the only way we make money. That's how you know that we are not going to not going to mess with that rule book. We're not going to mess with that system. The only way we make money is by taking a 10 % cut of the publisher's subscription revenue. So we want them to make as much money as possible. You want to maximize their paid subscriptions. But on the ideological level, it's also a better way to uphold and promote and support trustworthiness and truth seeking and good faith discussion with people seeking to understand each other.
Because the way you make money as the publisher, who is Substack's main customer, is to win and then keep the trust of your subscribers. And that is about more than performative outrage in a single moment, or in like a little Twitter bot on a feed. That's about consistently showing up over time, demonstrating that you're a thoughtful player.
In some cases, demonstrating that you're willing to speak up against your tribe, that you have some courage, that you have some conviction and values. And we think that when you switch, when you change the business model, you change the thing that sits at the very root, the very heart of these media systems, you can get better and different results. You switch to the good business model, the one that maximizes trust and defending yourself at length then you're going to get better results than the one that maximizes performative outrage.
Bernard (09:29.346)
It's fascinating to think how a lot of this is about human brain chemistry in a way and how the social media engines that are based on advertising have used those gamification techniques to hook us in and get us active and get us clicking and moving around and, of course, seeing the ads.
I wonder though, whether over time people become sort of burned out by the outrage. And you often hear about people who are, withdrawing from social media, having a break. I've had a few myself, where you have a sense of exhaustion and it's almost a chemistry thing. You rinse your head with dopamine enough, eventually you're exhausted. And I wonder if you're seeing this after a good 10, 15 years of Twitter, Facebook, Instagram and the likes, where there is a sense of exhaustion amongst internet users generally, and people who are weaning themselves off the dopamine and going to something a bit more, as you say, fulfilling, trustworthy, in a way, something a bit more about the protein than the sugar.
Hamish (10:57.927)
We see and hear a lot of anecdotally people saying Substack is a breath of fresh air. It helps me get back to what I really care about. And it's helping me rediscover the joy of writing or photography or making podcasts or even being a listener or a reader or a viewer of these things. And then I think we see it through behaviors generally like people are kind of fed up. I know TikTok is still doing really well. They still have a bunch of addicts who are spending all their time there.
But I think I also heard recently, this is not scientific yet, I don't have the source at hand to prove it, that the younger generations are a little bit hip to this kind of exhaustion this crazy state, crazy mind state that people try themselves into because they see the older generations. There's nothing more embarrassing than a GenX person who was addled by Twitter or addled by Instagram and they stop becoming like well -grounded individuals in public society. And so that's not like the younger people should rebel against that.
And I think we are seeing that. I think we're going to see more of it. I do think this is where the internet is still growing up. We've had, you and I have seen it from the start, we've had 30 years of this stuff where people are scrambling around trying to make the best stuff for the internet and find the biggest audiences and have the best conversations. And so far the attempts to commercialize it have revolved around one business model really.
There are some few exceptions, but the central ones undergirding the biggest media platforms that have ever existed — Instagram, TikTok, Facebook, Google, LinkedIn, Twitter, YouTube — is all advertising focused. And so we only have known kind of one thing for 30 years, three generations, the entire life of the internet, at least the commercial internet.
And I think what we moving into here with models like Substacks and like Patreons and some others that are out there is like the discovery of a new kind of model that can support culture and that can support creativity online and that can lead to different types of discussions and different kinds of outcomes.
It's still early days for that model but I think from a distant view, in the future looking back on this time, I think we'll think of the internet where people were still trying to figure out how this whole thing would work. And I think we'll look back with a little bit of embarrassment for how naive we were in this early time of the internet when everything was assumed to just have to revolve around aggregators and ad-based models.
Bernard Hickey (14:00.31)
It feels like we've just had a decade long experiment in real time with our brains. And in a way, this election is a sort of a test of it, whether the fits of madness that have emerged from our social media world maybe is about to end.
Hamish (14:08.413)
I think it's still going on.
Bernard Hickey (14:28.056)
I'm curious too about how you've thought at Substack about issues like hate speech, objectionable material,bad behavior, that sort of thing, because it's clear it's one of these issues that social media have had to deal with over the years. And you're starting to see now, you know, states pushing back, Brazil, Australia, various other places are going, okay, we're going to turn you off. So how does Substack approach all of this in an election year when, know, the stakes are high and emotions are high too?
Hamish (15:18.557)
I'm not sure that the states have the right solutions here, but I'm also not sure that the previous generation of internet companies had the right solutions. The way we think about it is that it's impossible to stop all bad things being said, bad things are gonna be said online, bad people are gonna say them and you have to have guardrails. So in our case, if there are incitements to violence, there's porn, there's spam, if there's doxing, that kind of stuff isn't allowed.
But beyond that, I don't think that the right approach is to come up with an ever more sophisticated content moderation apparatus. Facebook and Twitter and YouTube and the likes have spent billions of dollars on this over the years. They've hundreds of thousands of people. They've got policy documents that are 150 pages long with arcane stipulations like it's okay to show nipples as long as they're not female presenting nipples or is okay to show buttocks as long as the buttocks are like Photoshopped or something. You can fact check me. If it's not specifically that, but you understand the absurdity of some of the definitions that they're forced to make when they're down this path of trying to regulate speech through sophisticated content moderation apparatuses.
Our approach is not we’re not going to do that and be smarter than they are at like building the content moderation apparatus. Our approach is we don't actually think that the problem is that content moderation is not good enough, or that there's not enough resources being pumped into content moderation, or that the technology platforms are not responsive enough to demands from governments that have their own ideas about what kinds of speech should be allowed and what should not be allowed.
Hamish (17:08.079)
Our approach, the way we think about it is that the problem again is at the root level. It's the business model level. And that's the problems that we've seen from the last generation, the last decade in particular, are problems related to having a centralized aggregator who controls what gets rewarded and who gets paid, what gets seen the most. And instead of something like Substack, which is about a decentralized system where can have more decentralized moderation, where the publishers and the consumers can have direct relationships and the publishers can have the power to moderate their own communities and decide how they want the tenor of those communities to exist. And given that they are direct relationships, that Substack is facilitating, not mediating, then the bar for intervention and the bar for like coming in as the wise authority on what stands as far as what can be said and what cannot be said should be much, much higher.
And with this approach, I think that we're seeing that again, because you change the business model from this other thing that encourages bad things to this new thing that encourages good things, you get different results.
The substack ecosystem is not perfect. There are some nut jobs saying crazy things, and there are people that we have to kick off for breaching those content guidelines that I mentioned before. But overall, the tenor of the discourse, the tenor of the discussion, the productiveness of the works that are being produced and considered and shared is a hell of a lot better than what has come before.
Obviously, I'm extremely biased on this, but this is one of the reasons I wanted to build this company and that I continue to build this company and give everything to it , because it's not perfect, but it's about a hundred times better than what you see on Twitter or a hundred times better than what you get on Facebook or Instagram.
I'm confident that the reason for that is the business model is different. It's changed and different at the kernel of the platform.
Bernard Hickey (19:15.84)
Is there any way to sort of measure the pool of nastiness, ugliness, of, of horribleness that you have on Twitter X, YouTube comments, those sorts of things, where people feel completely relaxed about attacking, saying nasty things about each other, particularly when they're not paying for it and often it's anonymous.
I've seen it myself with our subscribers. I only allow paying subscribers or people on that top paying tier to comment. It's like if you've paid to go into a pub and you've paid for the beer, you really don't want other people tipping their beer into your beer, or you don't want the sound to be a squelch underneath your feet.
So no one wants to really mess in something that they're paying for. And I wonder if there's any data out there showing that that difference in incentives and the way that people tend not to destroy the nest that they build themselves.
Hamish (21:08.083)
I'll say a quick thing that backs up what you're saying. It's not only that, it's not just if you've paid, the quality of discussion is gonna be better probably. Even if you've just given your email address as a free subscriber, there's a moment of intention there that is different from being in the great wash where you can get like a pretty good reward if your goal is just to get someone's attention or just to piss someone off, just while you're doing a drive by comment. So
, one of the top revenue earners on Substack publishes exactly the same thing every day on Facebook than she does, that she publishes on Substack. On Facebook, she makes zero dollars for it. On Substack, she makes millions of dollars from it. The same content. And the discussion on Facebook is a total troll fest. And the discussion on Substack is a total love fest or a thought fest.We don't measure how much bad stuff is going. We actually don't pay much attention. We don't focus on it at least. We're more interested in measuring how many subscription dollars come in or how many subscriptions generally come in because that's a positive measure. That's the thing we're trying to encourage. And it's when you have a successful writer and a happy writer and that means their community is happy as well, then things are generally working well. So I don't have tip of the tongue sort of numbers to to share it, that would give us some indication of negativity. I could only do what you do, which is like, anecdotally, it seems a lot better. I have to deal with a lot fewer trolls.
Bernard Hickey (22:44.293)
One of the differences this time around versus 2020 is that you've got notes this time and also chat, which has again given us a discovery engine to find out other things. What are you seeing in terms of people discovering new substacks using notes as the launching off point to become subscribers, compared to four years ago when it was a bit more accidentally-on-purpose stumbling into things and discovery was the main problem.
Hamish (23:23.165)
Four years ago, I (don’t) think we really drove a significant percentage of subscriptions to any publications. There were definitely psychological network effects. If you came across something in the wild and you recognize it as a substack, maybe you'll be more likely to subscribe. Maybe your email address is already in system. Maybe your credit card is already in the system. You recognize the format and you know how that kind of game works. That kind of stuff is true.
Two and a half years ago, we introduced recommendations, which is a peer to peer recommendation thing. So when I sign up to
, I see three or five other writers who you really trust that you're recommending I subscribe to. And that goodwill growth loop is really meaningful because it's coming from you — a person. It's not Amazon saying, I think you might want to buy this next.And notes is an extension of that where people have a lot more power to set their own terms of engagement and the discovery is driven from me seeing something that you might have posted on notes like a more of a casual place that is global and reach is not just limited to your subscribers can reach other people's subscribers and me being seduced by your how interesting your mind is or the like the cleverness of how you describe something with a beautiful image you posted and later on and I might click on your face and say who's this guy and go down the rabbit hole and hopefully become a subscriber.
So that system now, which is a combination of the recommendations and notes and the app, which is becoming more and more of a destination, people open the Substack app every day to find good stuff to read and watch. That system now accounts for 50 % of all subscriptions across the platform. I think it's actually trending towards 60% and it's 30 % of all paid subscriptions across Substack. So probably The Kākā, I'm not 100 % sure, I haven't looked at your numbers and you started early, but if you looked in the last 90 days, probably 30 % of your paid subscriptions came as a direct result of the Substack network.2
Hamish (25:29.417)
And we charge 10%. So it's like 10 % of your revenue. That's what it costs you to use substack, or it’s otherwise totally free. But in return, you are going to get subscribers for free and paid subscribers especially.
Bernard Hickey (25:44.814)
I'm certainly seeing closer to 50%. Now for viewers of this from outside Aotearoa, you'll notice that we have a similar sort of accent going on here. As a former editor of a student newspaper in New Zealand3, one where the former Deputy Prime Minister and Finance Minister (Grant Robertson) is now the Vice -Chancellor. You've thought about New Zealand politics a lot, and although I know you've been in the States for well over 15 years, I'm sure you've got a vein of New Zealand political memory in your body somewhere.
Hamish (26:26.121) I think it's 14 years now.
Bernard Hickey (26:36.562)
You've got a pretty good idea of how the American system works for New Zealanders and for Americans. What could you say about this current situation we've got where Kamala Harris was at the last minute a vice president launched in as the party's leader pretty close to the election. That actually happened in New Zealand in 2017 with Jacinda Ardern. You would have kept an eye on that from a distance. What are the things that stand out for you as similar and different in that comparison? Because Jacinda Ardern did appear at the Democratic National Convention.
Hamish (27:26.825)
She did. They should have given her a main stage speaking slot. It seems similar on the dynamics on the kind of like momentum and hype and sort of cultural dynamic. I remember Andrew Little in 2017 was tanking, Andrew Little's Labour was tanking in the polls, he was not a popular leader and he made the self -sacrifice so that Jacinda could come in and inject new blood.
She was still a relatively unknown quantity when that first happened. They didn't know if she was going to be loved and successful. And it turns out like that changed, helped elevate her to the place where people decided that, they really did like her. And there's this massive momentum shift that ended up with Labour getting into power and Jacinda having two terms as prime minister. And that was still a pretty close to one thing, wasn't it?
The election, the government, was decided by (now-again-Deputy PM) Winston Peters's decision to ally with Labour back then, but it was in the balance. It could have gone the other way. And I think we've got a similar situation here. It's a 50 -50 race now, but it wasn't when Biden was the candidate. I think
's projection was that he had a 28 % chance of winning. The writing was on the wall.And the switch to Kamala has given the Democrats a totally new lease of life and totally new momentum and wiped out that polling difference. essentially put it into a 50 -50 tie again, which is a much better place to be than 28 -72.
The one difference here is that the election cycle is so much longer. The campaign cycle is so much longer in the US. And so the time from Kamala becoming the candidate, the nominee to now is probably longer than it was for the entire campaign cycle for that 2017 campaign.
So that strong hype and news cycle for her that is just starting to like tail off a bit. And so had the election been held today, I think she'd probably have a really strong chance. And now we've got this debate tonight that is a chance for her to sort of revive the momentum that is otherwise starting to flag. So it's a really high stakes debate. Everyone's going to be watching with intense interest.
Maybe she can kickstart that momentum again. But I think it's going to be similar to like, similar to what happened in 2017 without the benefit of MMP. So could go one way or the other. It's definitely not a done deal, but they're in a much stronger position than they were when Biden was a candidate.
And I think this similar switch is the reason for it. It's not actually necessarily because this is specifically a better candidate with better policies or anything like that. I don't think it's the politics of it, the policy of it. I think it's the psychological effect, the momentum, and it feels much better to have someone who's kind of younger being up there on stage and more dynamic.
Bernard Hickey (30:56.578)
The American system being presidential really depends on the person less so than the party.
Hamish (31:02.823)
So much of it is a show. It's just a great thing about America, and a shitty thing about America. It's like the greatest show on earth. America is the greatest show on earth. But if you look closely behind the curtain, you know you can see the cracks. The cracks are pretty ugly.
Bernard Hickey (31:17.696)
The difference too is the immense scale of the spending on advertising and on, you know, politics in America compared to here, even, you know, dollars per vote, we have increased our ad spending and the likes here. But how does that look and feel for a New Zealander looking at the American political system and the role of money compared to when we grew up?
Hamish (31:47.911)
I'm not going to have an intelligent thing to say, except, when you first get exposed to it and first come into it, it seems disgusting that that money could play such a huge role. And then, over time, I think maybe trying to control that might have some serious negative outcomes as well. And so maybe the best thing you can do is like set a playing field where the rules are the same for each party and each candidate and then let chaos reign. So it's definitely a different kind of game.
The government here as well is less central in the everyday lives of people. I'm always struck going back to New Zealand when I watch the nightly news, how almost every story has a government spokesperson or an MP or the Prime Minister quoted or featured. And that's just not the same in the US, where the President is this like important figurehead and Congress does important stuff. But the private sector has much more of a role to play and people are not defaulting to the government as the source of all the answers or the source of all support. That's about all I've got.
Bernard Hickey (33:12.958)
That's great Hamish. It's been wonderful to have you on talking about this. I'm fascinated by it, being part of it, and I just wanted to say thank you as well for all the enormous work you've done building a different engine. And it's great to see four years on from the last election, just how much progress is being made.
Hamish (33:31.657)
Can I say I'm really proud that The Kākā is on Substack and that you personally are driving it and seeing the success of The Kākā is a huge source of satisfaction to me. So thanks for showing the way for others.
Bernard Hickey (33:42.508)
Yeah, nah, it's fantastic. Just in the last couple of years, the amount of political, economic, cultural coverage on Substack about Aotearoa is just so much more, and the quality is so much higher. And yeah, it's been a real thrill to be part of it and I'm sure it's just the start. Ngā mihi nui and mā te wa.
Hamish (34:11.849)
Cool, thanks so much.
mostly for people who don’t get the New Zealand accent.
I checked after the interview. The Substack network effects generate 69% of The Kākā’ paid subscriptions in the last 90 days.
How Substack works to take (some) craziness out of America's elections