Instagram’s Existential Bet

Hiya! Before this week’s essay, two quick notes:

1. In case you missed it, I quietly announced the first ever Divinations workshop last week! It’s happening in early September and the goal is to help you conduct a market analysis of any industry you want—using my original fusion of the best ideas from strategy legends like Clayton Christensen, Michael Porter, and Hamilton Helmer. There are currently 17 reservations booked out of 30 total seats available, and registrations will close at midnight eastern on Friday. If you’re curious, sign up here for more info.
2. There’s a new episode of Ordinary Astronauts out! (New podcast with my co-founder Dan Shipper that explores tech, product building, and the psychology of work.) This new episode is a fun one: I interviewed Logan Bartlett! You might know him from his funny tweets, or from hosting the Cartoon Avatars podcast, which has recently been the home of the “web3 use-case” debate, where web3 enthusiasts line up to debate Zach Weinberg.

Alright, on with the post!

I unironically love it when Kylie Jenner weighs in on product strategy issues.

Every time she does it—first with Snap and now with Instagram—she creates a brief, shining moment where this topic I spend all of my energy thinking about is thrust into the spotlight. “Put me in, coach!!!” I shout into my lonely backyard writing shed.

Shoot, I even wrote about this exact thing two years ago!

Back then, the question was whether Reels (Instagram’s version of the TikTok video format) was a good idea at all. Now, the question is whether Instagram should replace their default view with a “full screen video experience” that would basically complete their transition into a wholesale TikTok clone.

On the “pro” side of the argument: Facebook’s data shows that short entertaining videos are drinking photos’ milkshake, and if Instagram wants to survive they need to jump off their burning platform. On the “con” side: the entire internet seems to be united against Instagram’s new product strategy, and it’s possible that this change could ruin what people like about the service. As Kylie Jenner put it: we just want to see cute photos of our friends!

This is a tricky situation for Zuckerberg and Mosseri. Both sides of the argument imply a rather weak position. But then again, the last time they copied a hot new rival it was Snapchat, and that went extremely well.

This week on Divinations, I attempt to get to the bottom of things:

• Is Reels even working?
• Why / why not?
• What’s with Instagram pushing it so hard?
• Will the whole thing blow up in their face?
• Or is this some weird, counter-intuitive genius strategy?

Let’s dive in!

Is Reels even working?

Yes and no. Reels is now two years old and, according to my reading of the data, it’s only working about half as well as Stories was two years after it launched.

Back in 2018, when Stories was as old as Reels is now, Kevin Systrom was asked about the proportion of time spent in Stories. He declined to give a number but claimed it was “almost just as important” as the main feed. So I’m assuming it was probably somewhere in the 40–50% range.

Reels, on the other hand, only accounts for 20% of time spent—despite the fact that Instagram is pushing it a lot harder. The only thing they did to get us to use Stories was put that little row of circles at the top of our feeds. But now they’re actually sprinkling Reels directly inside our feeds so they’re impossible to ignore. If it was just a little row you could quickly scroll past, I bet that 20% figure would be much lower, perhaps below 10%.

But still, to go from zero to 20% of all time spent in the app is nothing to sneeze at for a network as large as Instagram. I personally know several people—most of them boomers—that love Reels, for the same basic reasons so many people love TikTok: it can be quite entertaining! And if you’ve never tried TikTok (which is the case for lots of people, especially older people) then Reels is great and you don’t feel like you’re missing much, plus, you don’t have to download a new app. 

So it’s not completely dead on arrival, but it’s not going great, either.

Why is Reels neither a hit nor a total flop?

Of course it’s complicated and there are many reasons, but here are the main issues, in my opinion:

First, as I noted with co-author Neel Sharma two years ago, Reels isn’t just a different format for the same kind of content people already share on Instagram. Unlike Stories, which matched what people already wanted from Instagram (photos and videos from people you know), Reels/TikToks are created for and by randos. The whole point is that anyone might see it, so you should make it entertaining even if they don’t know you. Relationships are secondary. 

The reason Reels/Tiktok works this way has more to do with the network structure than the content format. With these types of videos, the whole point is the “for you” page which shows you content by anyone. This is the key design decision that made people create content differently, and allowed the TikTok algorithm to be so effective: it had a lot of broadly entertaining stuff to choose from. The one big format thing that matters is the music and reusable sounds, because when you have unfamiliar people in most of the videos you see, it helps to balance that with familiar audio. This is also why memes, challenges, viral dances, and trends of all sort flourish on TikTok: they create familiarity, which it pays to piggyback off of.    

Instagram’s network is structured completely differently, and the content people share on it is a function of that. Instead of seeing videos of random people, you see photos and videos of people you are already familiar with. This changes what kind of content people share and makes it far less reliant on memes, music, and trends. I think this is the thing people are bemoaning when they say they miss the old Instagram: by changing the network structure to prioritize entertaining content from randos, they make it harder for people to share the kind of content they used to, which assumed more familiarity. This content was already inherently interesting because of the social—or parasocial—relationship you had with the person posting it. 

This is why Instagram has had so much more trouble copying TikTok than they did Snapchat. Snapchat invented a new content format that still existed within the same basic network structure as Instagram. With TikTok, the fact that it’s a new network is actually a feature, not a bug. There is an inherent conflict in the goals of creators for the different formats, and this tension is I think keeping a lot of people from posting to Reels.

But besides the whole “network structure” thing, another reason the Stories clone worked so much better than Reels is the simple fact that TikTok is way bigger, and it’s much harder for Reels to get someone to switch than it is to get people interested in the format if they’ve never seen it before. Once TikTok pervades a community, Reels is seen within that group as low status, and has anti network effects.

Of course, this problem isn’t unique to Reels. When Instagram built Stories they had this problem with Snapchat users, too. And to this day lots of Gen Z who grew up on Snapchat still use it instead of Instagram Stories. But because Snapchat was growing much more slowly than TikTok, it wasn’t as big of a problem.

Source: FT

(Note: it is absolutely wild how fast TikTok is growing and how stuck Twitter is. It’s like they are from different planets.)

So those are the two reasons Reels isn’t working as well as Stories. But let’s not forget that hundreds of millions of people use Reels, and it accounts for 20% of time spent in Instagram. It raises the question: if what I said before is true, why didn’t it totally flop?

To me this is just an illustration of how these forces are not all-or-nothing, black-and-white things. People really do go to Instagram to be entertained by visual content, and Reels gives it to them. As weird as it is to think of, there are probably some people who only use Reels. For instance if you don’t have a lot of friends that are that active on Instagram, Reels is probably more entertaining than following celebrities or meme accounts. It’s certainly a lot less work. But that’s just one example, Instagram is a huge app and almost anything they put in front of people will work to some extent.

In this case, the answer we have so far is that Reels works about half as well as Stories. There are some forces working for it, and some forces working against it.

That being said, I think it’s important to note that probably the main force working for it is the sheer will and determination to make it work by Mark Zuckerberg and Adam Mosseri. It’s worth asking: what’s with that?

So why is Instagram pushing Reels so hard?

In Facebook’s most recent earnings call Mark Zuckerberg said a thing that I think is probably true, and also fairly stunning:

“Since I started Facebook 18 years ago, we've seen multiple shifts in the media types that people use. We started as a website primarily with text. Then people got phones with cameras, and the main format became images on mobile apps. In the last several years, mobile networks have gotten faster and now video is the main way that people experience content online.”

If this is your worldview, then of course you’re going to be scared shitless by TikTok! Facebook has tried and semi-failed to make video work countless times: the pivot to video in the Facebook News Feed, IGTV, the list goes on. They might be running out of chances.

Zuckerberg knows better than almost anyone else just how delicate an art it is to build a network around a content format. Just because you get the format right doesn’t mean you get to own the most important network. For example, 15 years ago the content format of Twitter and Facebook were nearly identical. But look what happened on the graph above, and you’ll see why the details matter so much, and why Facebook and Instagram feel they need to copy everything they can from TikTok.

The secular shift to video is scary on its own, but that’s actually easy compared to the other big trend Zuckerberg sees coming: algorithms showing you content from people you don’t follow. Facebook is massively behind TikTok and YouTube on this one.

In that same earning call Zuckerberg talked about how there’s a “major shift” towards discovery algorithms, because it “unlocks a large amount of interesting and useful videos and posts you might have otherwise missed.” But he doesn’t just see this as unique to Reels / TikTok. He wants to turn this into a major drive for all types of content across the Facebook family of apps, saying “I think about the AI that we're building not just as a recommendation system for short-form video, but as a Discovery Engine that can show you all of the most interesting content that people have shared across our systems.”

Will the whole thing blow up in their face?

This is, to put it bluntly, not going to go well for Facebook. All the content moderation and misinformation problems they had—even if you think they’re overblown—are going to get way worse. It’s a weird PR miracle of some sort that people aren’t more focused on YouTube and TikTok, and Facebook has become a sort of scapegoat that attracts all the attention in that conversation. I think it’s inevitable that we’ll see a big new wave of criticism if they keep marching in this direction, and a lot of it will probably be deserved.

Also, I find it kind of wild that just a few years ago they were all focused on “meaningful interactions with friends and family” and “building real communities”—then TikTok came along and they were like “actually nevermind.” It seems as though they are lost; trapped between reacting to perceived existential threats and overwhelming public criticism.

Or is this some weird, counter-intuitive genius strategy?

My guess is Mark Zuckerberg is tired of all of it and can’t wait for us to just transcend into the metaverse. But building it requires a cash cow, and Instagram’s monetizable impressions have to come from somewhere.

Also, to be clear, I am aware that Mark Zuckerberg is objectively smarter than me. He owns a company that was recently worth a trillion dollars, and I write an email newsletter. Also I heard he invented calculus techniques during tests at Harvard because he didn’t study and had to figure it out on the fly. 

So, like, I am definitely aware that in the previous section I came off as a know-it-all. I am open to the possibility that I am wrong. It will, at the very least, be exciting to watch this all unfold.

(From a safe distance.)

Also just a reminder that you have till Friday to register for the course on market analysis. Click here to register for more info.

Thanks, and see you next Wednesday!