How to Decide — and Time — Whether to Cannibalize Your Own Product

Apple and Netflix arrived at the identical strategic verdict: the product responsible for most of the company's money was going to be destroyed by the future whether they acted on it or not, so each chose to build the destroyer itself. Apple aimed a new phone at its own iPod, then carrying roughly 40 percent of Apple's revenue.² Netflix aimed a new streaming habit at its own DVD-by-mail machine, the more profitable of its two businesses by a wide margin.⁵ Apple's version of this decision is now taught as the cleanest act of corporate discipline in a generation. Netflix's version is taught as a near-death experience. Both companies made the right call. What happened next is why one story reads as triumph and the other as a scare.

The method separates two questions: the decision that gets studied to death, and the execution that almost never does

Every self-disruption is actually two decisions stacked on top of each other, and the businesses that get hurt are the ones that only ever answer the first. Question one: is the thing you're protecting going to be disrupted regardless of what you do, and if so, are you the party best placed to build what replaces it? Get that right and the strategic case is usually not close. Question two is the one almost nobody schedules time for: how do you carry the customers who already trust you through the transition without making them pay, in money or hassle, for your foresight? Apple answered both questions well. Netflix answered the first one about as well as a company can, and the second one so badly it lost about 800,000 U.S. subscribers in a single quarter finding out the difference.⁹

Apple: name the threat, then make it disappear inside the answer: nothing for an existing customer to give up, log into, or pay extra for

By 2006 the iPod was not a side project. It was Apple's single largest product line, worth roughly 40 percent of the company's revenue that fiscal year — about $7.7 billion of just over $19 billion in total sales.² Steve Jobs's answer to what a rational competitor would eventually do to that business was to become the competitor himself: build a phone with an iPod already inside it, aimed squarely at his own cash cow. He framed the logic as a rule, not a one-off — if you don't cannibalize yourself, someone else will.¹ The decision was the hard part. The execution was almost effortless, because Apple never asked an existing iPod owner to give anything up or pay a penalty to keep what they had — the iPhone was simply a new product, sold at a new price, to whoever chose to buy it. Nobody's existing plan changed. Nobody logged into a second account. The old product didn't get worse; it just stopped being the newest thing Apple sold.

Netflix: the same decision, a business that argued back harder: the healthier the margin on the dying business, the easier it is to protect it for the wrong reasons

Netflix's version of the same choice was arguably harder, because the business it was choosing to shrink was not fading — it was still, by a wide margin, the better business on paper. When Netflix finally broke out the numbers years later, its domestic DVD-by-mail operation ran at a 47.3 percent contribution margin in 2012, against just 16.0 percent for domestic streaming — and Netflix said outright that the high-margin DVD business existed partly to fund the lower-margin one that was replacing it.⁵ Streaming, which had launched in January 2007 as a free add-on called 'Watch Now,'⁴ was the objectively worse business: thinner margins, a smaller catalog, and content costs that only rose as studios caught on. It had exactly one advantage. It was where the audience was headed, and Reed Hastings had already decided which side of that trade he'd rather be on.

Where the two companies stopped rhyming: folding the old product in, versus cutting it out and sending the customer the bill

Here is where the two companies stopped rhyming. Apple folded the iPod into the iPhone — one device, one purchase, the old thing quietly dissolving into the new one. Netflix tried to physically sever its two businesses, and it asked customers to feel every inch of the cut. In July 2011, Netflix broke its single $9.99 plan — streaming plus one DVD — into two separate $7.99 plans; anyone who wanted both now paid $15.98, a roughly 60 percent increase overnight.⁶ Two months later, instead of absorbing that lesson, Netflix doubled down: it announced it would spin the DVD business off entirely into a separately branded company called Qwikster, with its own website, its own login, and its own queue.⁷ Reed Hastings opened the announcement with an apology for the price increase, then used the same letter to confirm the split.⁷ The internal case for the move was coherent — different economics, different customer, different roadmap. None of that reasoning survived contact with a customer holding a red envelope, staring at a higher bill and a second website to manage.

The difference wasn't conviction — both companies were equally sure of where the market was going. It was what each made the customer absorb on the way there. Apple's transition had a built-in kindness: nobody had to adopt the iPhone, so nobody who preferred the old product was forced to feel the change. Netflix's transition had no such release valve, because DVD-by-mail and streaming shared the same customer base, the same billing relationship, and now, suddenly, a much higher combined price. A cannibalization decision made at the strategy level doesn't automatically arrive at the customer level pre-translated. Someone still has to design what the transition feels like to the person paying for it — and that design work is a separate job from the decision itself, done badly here even though the decision underneath it was sound.

The bill arrived fast. Netflix's October 2011 letter to shareholders reported its first-ever domestic subscriber decline — about 800,000 U.S. subscribers gone in a single quarter.⁹ The stock, which had touched an intraday high of $304.79 that July, fell to roughly $62.37 by late November: a decline of about 80 percent.¹⁰ None of that was a referendum on streaming. It was a referendum on being handed two logins, a higher bill, and no real warning, all in the same season. Netflix reversed the Qwikster split 23 days after announcing it,⁸ but the price increase stayed, and so, for months, did the damage.

The honest objection: a harder transition is a reason to rehearse it more, not an excuse for how it went

The honest objection is that this comparison is a little unfair. Apple was selling an optional new device; Netflix was changing the terms of a subscription millions of people already relied on, which is a structurally harder transition to make gently. That's true, and it doesn't let Netflix off the hook — it raises the bar on execution rather than lowering it. A harder transition is exactly the case where you stage the price change separately from the brand split, test the two-website idea with a slice of customers before rolling it out to millions of them, and make sure someone senior is allowed to say 'this will read badly' before it ships. Netflix's own later account of the episode locates the failure inside the building: colleagues who had doubts didn't push back hard enough to stop it.¹² A harder starting position is a reason to war-game the rollout more carefully, not a reason it was destined to go badly.

Two variables sharpen the first question before the checklist even starts. How much profit is actually at stake — is this a rounding error or the crown jewel, the way the iPod and DVD-by-mail both were? And is the window to act on your own terms still open, or has a rival already forced the question? Move too early, before the replacement is any good or the market is ready, and you've sacrificed a profitable business for nothing. Move too late, and someone else owns the future you gave away. Apple and Netflix both moved while the profit pool was still large and the window was still theirs to choose — which is exactly why both are worth studying: neither company waited for the market to make the choice instead.

Both companies eventually got the vindication their decision deserved. The iPod Apple chose to make obsolete was quietly retired in 2022, well over a decade after the iPhone that replaced it first shipped.³ The DVD business Netflix set out to shrink mailed its final disc in September 2023, more than a decade after the strategy that nearly sank the company was proven exactly right.¹¹ Neither ending was really in doubt by then. What decided how much each company paid to get there wasn't the quality of the strategic call — both were sound. It was whether anyone had done the separate, unglamorous work of designing what the transition would feel like to the people still holding the old product when the new one arrived.