One of the perks – I guess you can call it that – of my new position at work is that there’s a lot of travel involved. While I don’t particularly enjoy the literal part of the equation, I do enjoy new cities and new people and new Thai food.
Being 6’5″ on a commercial airliner is tough, but I mitigate it two ways: one, I fly JetBlue almost exclusively, because they’re awesome and two, knowing that Internet will be spotty at best, I’ll pre-load my devices with a deep dive on various product-related topics, allowing me to really sink into a topic that I’m interested in. It passes the time and generally keeps my mind off of the long-term damage I’m doing to my lower back.
For this flight’s dive, I was inspired by the big news of Instagram’s founders stepping down, so I dug into the differing philosophies between Instagram and Facebook, between Kevin and Mark, between product-driven thinking and business-driven thinking. I could fill about ten more posts based on thoughts that came out of that, but for this one I instead would rather focus on a nominal competitor to both – to the extent which it’s a replacement good, which it’s not – of those two: Twitter.
I’ve long thought that Twitter was and is one of the most disappointing products in memory. Not that isn’t wonderful when it’s at it’s best doing the specific thing it’s designed for – it is – but man, it really is terrible when it’s at it’s worst. The delta between how good Twitter can be and how bad it usually is, well, it’s as big as it gets in the product world. And while there are five or six small reasons why, I posit there’s one big one: anonymity.
(As an aside, Twitter has 3,500 employees. I’m going to assume that at least a quarter is in product development. Now tell me why you can’t edit a tweet. Or name a single notable feature they’ve built recently.)
The problem with anonymity is simple: it causes people to be assholes. Full stop. I’ll grant that it’s a cynical point of view, but if you dig into the dregs of Twitter or Reddit or YikYak or any community built without real-world identity, you’ll see a lot of base human nature coming out in the worst of ways. Twitter for example has an obvious abuse problem that is complained about by nearly everyone in the public eye, and it’s a problem that is fundamentally impossible to stop. Doing so would cause two issues: first, as a public company, it would have a huge impact on the vanity metrics that prop up the valuation and second, it would expose the company to even more accusations of targeting and bias than they already have.
That’s assuming they even want to fix the problem, which is questionable in itself. Nothing spawns more engagement than showing things that cause a visceral emotional reaction; what gives you more of an emotional reaction than abuse and awful people attacking your core beliefs? (This is actually where Facebook is the worst, but that’s another topic.)
I know a few investors (mostly notably Fred, here) have covered this topic, and as someone who has invested in Twitter and other networks, I’m sure he’d likely disagree with my thoughts. That’s the difference between investor and consumer; as an investor, he’s incentivized to look for the kind of hyper-growth and potential that can often only come with a network that is like a giant sandbox, full of freedom and expression. As a consumer and over a long enough time frame, that’s exactly how you get communities like Coontown on Reddit, and the entirety of 4chan. By then the investors likely will have exited, bailing out of the wave before it crashes.
One of the only bosses (Jeff Leventhal, when we were at WorkMarket) I’ve ever had that I truly got along with – which is probably more of a me problem than a them problem, honestly – once told me something incredibly insightful about motivation. In his view, people only really do things for three reasons: to get made, to get paid, or to get laid. That’s it. Crass, but I agree wholeheartedly.
Without the connection between your platform identity and actual identity, you fail that first criterium, for the same reason why people don’t use racial slurs on LinkedIn. If you’re an asshole, it affects your reputation; without it, there’s no downside to engaging in your worst impulses. Troll away. You’re not sitting next to that person; their response is abstracted. So where are the consequences?
Of course, we all know that those consequences simply get pushed from the user to the platform. In aggregate, the negativity gets associated not with an individual user, but for the platform’s inability to create a healthy environment for users and a brand-safe one for commerce. That’s why Reddit, Twitter, and 4chan can’t monetize within a fraction of how Facebook can. And what’s worse, there’s no solution: you can’t get the cat back in the bag. You’re forced to rely on self-moderation, which is not only unscaleable but incredibly prone to bias, which in turn fuels the partisan accusations that we’re seeing play out. It’s an endless, fruitless loop.
(Here’s some back of the envelope math: Reddit supposedly does 14b page views a month. Let’s say that each view has a sponsor attached to it. Given that Reddit should be able to target both from a consumer side (what is this user interested in?) and from the business side (I only want to advertise on a sports sub!), I’ll give that ad unit a CPM of $3. That would put their ostensible revenue at $42mm a month, or about $500mm a year. I’m taking the under, hard.)
Furthermore, I don’t think for a second that Jack or Mark are naive to this. I think it’s clearly a very cold calculation: they’re not incentivized to deliver a positive experience to the user, because doing so both costs money (in terms of tools, monitoring, effect on valuation, etc.) and subverts the engagement flywheel that comes with emotion-inducing content. In a world where all you’re ultimately concerned about is engagement and growth, the means for how you get there is irrelevant; after all, people share for the reason of “Look at this bullshit! How can anyone think this?!” just as much, if not more, than they share because they agree.
I do think, however, that when these companies were first started, these second-order consequences were almost unthinkable outcomes; cynicism just isn’t in the founder DNA to that degree. I definitely don’t think the founders of Reddit expected a community like Coontown to pop up. Not even Mark, knee deep in his “dumb fucks, they trust for me some reason” callousness, likely did not set out to create this kind of toxicity. But they’d be incredibly foolish to ignore it now, because ignoring it is a time-bomb: one way or another, it resolves itself. Either the users get fed up with the degraded, abusive experience (YikYak), or commerce becomes impossible as the waters become too polluted (4Chan, likely Reddit). All it takes to set off on the hard road to change is for one board member to say enough, to prioritize user experience and long-term revenue.
I’m not holding my breath.
One of the defenses I see now and again is the re-definition of these companies as audience businesses – literally solely in the business of providing an audience to advertisers, agnostic of the means – but I think that’s an extremely short-sighted view of it. I think defining them as such tacitly creates an environment in which you become economically tied to the decision to let anything go, which creates an increasingly difficult environment to monetize as the environment gets more and more extreme. And maybe that’s just the difference between the way I think about it as a product person and someone else who might think about it when they have to answer to their Board/shareholders or get fired.
If you’re starting a community from day one, I think it’s an interesting dilemma: do you tie digital identity to the real world, slowing down growth in the short-term but maximizing revenue generation at the long-term should you survive to a scale in which advertising is interesting? Or do you go the anonymous route, grow like crazy, and try to deal with the consequences? It’s a paradox: once you get to scale, the community becomes impossible to police, making monetization progressively difficult and the user experience progressively worse. But without that initial freedom, you may never get to the scale that would require you to seriously think about monetization in the first place.
But for me, the answer is obvious. I’m not resigned to think that a good and safe identity-led product experience and true market fit are mutually exclusive. Or, to put it another way, think of how much better an actual identity-only Twitter would be! I might even argue that the long-term business prospects of that concept as something that people would be willing to pay a subscription for outweighs the advertising business model upside, particularly in the world of ad-blockers! But that’s another post for another day.