Snap Didn't Choose to Cannibalize Its Cash Cow. The Choice Was Made For It.

In February 2018 a single tweet from Kylie Jenner — asking whether anyone still opened Snapchat — knocked about $1.3 billion off Snap's market value in a day.⁵ It is one of the great parlor stories of the social era: a celebrity sneezes, a public company catches a cold. But the tidy version hides the real one. Earlier that same week, a Citi analyst had already cut the stock to Sell, citing a spike in negative app reviews since Snap had rebuilt its own product.⁵ The tweet didn't start the fire. It poured petrol on one Snap had lit itself.

The official story is that Snap is the rare company brave enough to disrupt its own cash cow — to mess with the very thing that worked. The truer story is colder: Snap was pushed into cannibalizing itself by a competitor, and then fumbled the self-disruption twice. The cash cow it most needed to break — its near-total dependence on advertising — it left intact for years, until that dependence became the whole vulnerability.

Stories didn't eat the core. A rival did.

Start with the myth that Snap cannibalized itself by inventing Stories. Snapchat launched Stories — first called 'My Story' — in October 2013, letting users stitch snaps into a 24-hour reel.⁷ Stories didn't kill ephemeral messaging; it extended it, gave the product a public broadcast lane on top of its private one, and gave advertisers somewhere to live. That was addition, not cannibalization. The knife came from outside. In August 2016, Instagram — owned by Facebook — shipped a Stories feature so close to Snap's that it barely changed the name.⁷ A format Snap had pioneered for almost three years was suddenly available to Instagram's far larger base, and the copy worked. Snap's own IPO paperwork, filed months later, conceded that mid-2016 product launches had caused technical issues that 'resulted in a reduction in growth' of daily users.² The disruption to the cash cow was real. It just wasn't a choice.

When you redesign the cash cow and it stops mooing

Cornered by the clone, Snap reached for the one lever it controlled: it rebuilt the app itself, hoping a cleaner separation of friends and creators would make the product easier for advertisers to monetize. The bet inverted. In Q2 2018, Snap posted its first-ever sequential decline in daily active users — the number a social company is structurally not supposed to be able to lose. Spiegel told analysts the drop was 'primarily driven by a slightly lower frequency of use among our user base due to the disruption caused by our redesign,' and later admitted in an internal memo that the company had 'moved too fast.'⁶ The changes were partly reversed. This is the heart of the cannibalization trap: the only honest way to know whether self-disruption works is to ship it to the people who already love what you're disrupting — and they are exactly the people you cannot afford to lose.

Notice what the redesign was actually trying to protect. Snap's entire revenue line was advertising, and advertising rewards predictable, legible, scalable attention — the thing Instagram was now better positioned to deliver at scale. Snap reshaped its product to look more like an ad inventory machine, and in doing so dented the very engagement that made the inventory worth buying. It was disrupting the wrong cash cow. The dangerous dependency was never the app's layout. It was the single revenue source underneath it.

The cow it should have disrupted was the revenue model

Here is the cannibalization Snap actually needed and avoided for years. At its 2017 IPO, the company reported $404 million in revenue and a $514 million net loss, with the business almost entirely advertising-funded.² Eight years later, in FY2024, revenue had grown to $5,361 million — and the 10-K still describes advertising as 'substantially all' of it.¹ A business that earns one kind of dollar and only that kind of dollar doesn't have a moat; it has a single point of failure. Every algorithm change at Apple, every macro ad pullback, every Instagram feature lands directly on the only pipe Snap has. The real self-disruption — building a revenue stream that doesn't depend on selling its users' attention — didn't arrive until June 2022, when Snapchat+ launched at $3.99 a month and crossed a million subscribers in roughly 60 days.⁸

And Snapchat+ has worked, on its own terms. By early 2026 the subscription community passed 25 million members and the direct-revenue business cleared a $1 billion annualized run rate.³ 'Other Revenue,' which is mostly those subscriptions, grew 131% year over year in 2024.⁴ A revenue stream growing that fast off a non-advertising base is exactly the kind of cannibalization a healthy company welcomes: it competes for the user's wallet instead of someone else's ad budget. But scale tells the sober truth — even at $1 billion, direct revenue sits beside a $5.36 billion total that is still 'substantially all' advertising.¹ Snap diversified at the edges. It did not transform the model. The cow is still mostly the same cow.

Isn't slow-but-survived the right call?

The fair objection is that Snap's caution was vindicated. It survived the Instagram onslaught, grew daily users to 453 million by the end of 2024 — up 9% year over year — and turned Adjusted EBITDA from $162 million in 2023 to $509 million in 2024.¹⁴ A company that recovers from a self-inflicted user decline and triples its profitability hasn't obviously failed at anything. There's force in that. But it confuses surviving the disruption with choosing it. The 2018 redesign was not a clever wager Snap won; it was an error its own CEO called moving 'too fast,' and the recovery came partly from undoing it.⁶ The deeper count is the clock. Snap saw the danger of single-source revenue at its IPO and didn't ship a serious alternative for nine years.²⁸ Survival bought time. It did not buy the second pillar — and a company that can still be moved $1.3 billion by a tweet is telling you how thin the floor under that single pillar really is.⁵

Snap's reputation is for nerve — the company that wasn't afraid to break its own toys. The record describes something more human and more useful: a company that had its best feature copied, panicked into redesigning the product its users loved, watched them leave for the first time in its history, and only much later got around to disrupting the thing that actually endangered it. The lesson isn't 'be braver.' It's that not all cannibalization is equal. Eat the right cow, and you build a second business. Eat the wrong one, and you just lose weight — while the herd keeps grazing somewhere you don't get paid.