Cannibalization: launching the thing that eats your best business

The rule Apple wrote down

In fiscal 2006 the iPod brought in about 40% of Apple's revenue. In 2007 Apple launched a phone that played music, which meant every iPhone it sold was one less reason to buy an iPod.

Steve Jobs had a rule for exactly this moment.

“If you don't cannibalize yourself, someone else will.”Steve Jobs, as recorded in the Apple case

Apple didn't kill the iPod. It let it fade on its own schedule, and retired the last model in 2022, long after the iPhone had made it unnecessary.

That is the whole decision in one story. Put that way it sounds easy: of course you eat your own lunch before someone else does. Yet many of the companies in this guide said something like Jobs's sentence, and 5 of them ended up taking their move back. The rest of this guide is about the gap between the sentence and the doing.

What cannibalization is

Accidental cannibalization is a pricing problem: a cheaper size steals from the bigger one. Deliberate cannibalization is a bet on the future. It comes up when the product that pays the bills is about to be made obsolete by something the company can see coming, whether or not it is the one that builds it.

Why it is so hard

On paper the choice is lopsided. In a real company, three things tilt it the other way.

The ways companies did it

The decisions on record take four shapes. Each answers the same question with a different amount of commitment.

What 39 companies did

Here is every decision on record, placed by the year it was made and marked by how it turned out. Hover a mark, or open the table, for the company and the outcome.

Of the 38 decisions that can be judged, 32 were never undone, 5 were walked back and 1 was kept only in part.

Remember what "never undone" means. BlackBerry's refusal to open up BBM was never undone either. It didn't need to be.

The walk-backs

Put the walk-backs side by side and a pattern shows. Most of them never changed what customers bought. They changed who owned what, with a split or a set of listings, or they announced a pivot and never funded it. Those are the easiest moves to announce, and the easiest to take back.

The rest pushed too much on customers, too soon. Under Armour reached up to fashion and down to the discount racks at once, and in 2017 its North America revenue fell for the first time after years of 20%-plus growth. Marvel flooded the market with films and series, and The Marvels earned $206.1 million worldwide, the lowest in the franchise's history.

The decisions that stuck had something in common too: they changed what customers bought, and they spent money that couldn't be recovered. Apple shipped the phone. Philip Morris paid $16 billion. Walmart paid $3 billion in cash and then ran its online business at a loss for years before its first profit.

An announcement can always be taken back. A product in a customer's hand can't.

What the share price did

For many of the decisions there is a clean market record: the company's total return over the three years after the decision, against the S&P 500.

The chart is a warning against reading the market too quickly. Disney's 2019 bet on streaming was never undone, and its shares trailed the index by 67.8 points over the next three years. In fiscal 2024 Disney's streaming business made $143 million in operating income on $22.776 billion of revenue, while ESPN's cable reach had fallen from 74 million households to 71 million in a single year. A decision can be right and still be expensive to hold.

How to run this decision

If you are weighing a version of this choice, here is what the record suggests doing, in order.

1. Write down when the old business stops, as a date

   Not a trend, a year: the one in which the old revenue line falls below what it costs to run the organization around it. If nobody will put a date on it, the decision isn't ready to be made, and the rest of these steps will produce theater.

2. Take the new thing away from whoever owns the old number

   Whoever is measured on the old margin will protect it, sensibly, right up until it is gone. That isn't a character flaw. It's an org chart, and the fix is to give the new business an owner whose number goes up when the old one goes down.

   Seen in: BlackBerry, 2012
3. Price the delay, not just the move

   The cost of acting is immediate and specific; the cost of waiting is distant and vague, so waiting always looks cheaper on paper. Ask what one more year of the current position costs if the market keeps moving the way it is already moving.

   Seen in: BlackBerry, seventeen straight quarters of lost share
4. Change what customers buy before you change the org chart

   Splits and listings announced before customers had proven the new business are among the walk-backs on record. Prove the new business with customers first, then restructure around what works, rather than restructuring in the hope that the structure produces the business.

   Seen in: Netflix, 2011 · Alibaba, 2023
5. Set the old product's retirement date before you launch the new one

   The point is not to hit the date. It's to have the conversation about when the old thing ends once, at the start, instead of every quarter for a decade while two teams fight over the same customers.

   Seen in: Apple, iPod retired 2022
6. Name the number that would make you stop

   Write it down with the date you will check it. A decision with no stopping condition isn't a bet; it's a position, and positions get defended long after the evidence has turned against them.

To work the numbers for your own case, read Product cannibalization: when to launch the product that eats yours.

Warning signs

What it has looked like when this decision was going wrong.

• The plan is an announcement, and no capital is attached to it in the first year.BP, 2020 · Alibaba, 2023
• The new business reports to the executive whose pay depends on the old one.BlackBerry, 2012
• You are reaching in two directions at once, up for premium and down for volume.Under Armour, 2016
• The old business's decline is described internally as cyclical or temporary.BlackBerry, 2008

All 39 cases

Each case is one company at one decision: what was on the table, what it chose, why, and what followed, with every fact sourced.

Questions

What is a cannibalization strategy?

Launching a product that you know will take sales from one of your existing products, on purpose, so that you own the replacement instead of a rival. Apple's iPhone, which made the iPod unnecessary, is the best-known example.

Is cannibalization good or bad?

Neither by itself. Accidental cannibalization usually signals a pricing or positioning problem. Deliberate cannibalization is a bet that the old product is going to decline anyway, and that it is better to lose its sales to yourself than to a competitor.

What are examples of companies that cannibalized themselves?

Apple with the iPhone and the iPod, Philip Morris with IQOS and Marlboro, Meta with Reels inside Instagram, and Walmart with its online business and its stores. BlackBerry and Kodak, which shelved its own digital camera to protect film, are the best-known examples of companies that refused.

Why do companies avoid cannibalizing their own products?

Because the old business is certain and the new one is a guess, because the executive who owns the old product is usually the one asked to approve its replacement, and because the threat rarely shows up in the results until it is too late to act on your own schedule.