In 2006, the iPod and its related services accounted for 50% of Apple’s total revenue. One year later, Steve Jobs launched the iPhone, a device he knew would make the iPod obsolete. Apple’s revenue at the time was $19.3 billion. By 2014, it was $182.8 billion. That growth happened because Jobs was willing to destroy his most successful product before a competitor could do it for him.
Product cannibalization is one of the most counterintuitive decisions a product manager will ever face: deliberately launching something that takes revenue from your own existing product. Every instinct says protect what’s working. The data says otherwise.
The Price of Protection
Kodak’s story is the definitive cautionary tale. In 1975, Kodak engineer Steve Sasson built the world’s first digital camera prototype. When he presented it to executives, their response (as Sasson later recounted in interviews) was essentially: “That’s cute, but don’t tell anyone about it.” They feared digital photography would cannibalize their film business, which held a 70% share of the global film market.
While Kodak sat on the technology it invented, Sony, Canon, and Nikon pushed digital photography forward. Those companies had no legacy film empire to protect, and that gave them a structural advantage. They weren’t burdened by decades of institutional momentum tied to chemicals, processing labs, and paper.
On January 19, 2012, Kodak filed for bankruptcy with $6.8 billion in debt.
The lesson isn’t that Kodak failed to innovate. They literally built the future. They failed to cannibalize. They chose short-term revenue protection over long-term market relevance, and the market made the decision for them.
Three Companies That Got It Right
Apple’s iPhone. Jobs understood something most executives resist: if disruption is coming to your category, you want to be the one delivering it. His reasoning was simple and widely quoted: “If you don’t cannibalize yourself, someone else will.” The iPod didn’t die because of a competitor. It died because Apple replaced it with something better. The revenue Apple lost on iPods was dwarfed by the revenue the iPhone generated.
Netflix’s streaming pivot. In 2006, Netflix had 6.3 million DVD subscribers generating $996.7 million in revenue. The DVD rental business was highly profitable. In 2007, Netflix launched streaming anyway, knowing it would eventually replace DVDs entirely. CEO Reed Hastings made the deliberate choice to phase out the profitable business before it was forced on them. Today, Netflix serves over 301 million subscribers worldwide and generated $39 billion in revenue in 2024. The company that mailed DVDs became the company that redefined how the world watches television.
Amazon’s AWS. When Jeff Bezos launched Amazon Web Services in 2006, it had nothing to do with the company’s retail business. But the strategic principle was the same: invest heavily in a new business that could eventually overshadow the original, even if it means competing with yourself for capital and focus. AWS now generates over $80 billion in annual revenue and contributes nearly 60% of Amazon’s total operating profit. The “bookstore” that built a cloud computing empire did so by refusing to define itself by its existing revenue streams.
What This Looks Like at a Smaller Scale
You don’t need to be Apple or Netflix for cannibalization decisions to matter. In my two decades leading operations teams, I watched this pattern play out repeatedly with internal tools and platforms.
One situation stands out. We had a monitoring system that the entire operations team relied on. It worked, people knew it, and it had years of customization baked in. But the architecture was increasingly brittle, and the vendor’s development had clearly plateaued. When a newer platform emerged that solved the same problems with better scalability, the question wasn’t technical. It was organizational: do we protect the thing everyone is comfortable with, or do we replace it while we still have the luxury of a controlled migration?
We replaced it. The transition was painful for about 90 days. Within six months, nobody wanted to go back. The teams that delayed the same decision at other organizations spent years maintaining a declining system and eventually had to make the switch under pressure, with less time and fewer options.
The principle scales down to individual features too. Every product manager has a feature that generates steady usage but is built on assumptions that no longer hold. The question is whether you replace it on your terms or wait until users find something better.
Four Questions Before You Pull the Trigger
Not every cannibalization decision is the right one. The companies that get this wrong typically skip the strategic analysis and act on either fear or enthusiasm. Before recommending that your team cannibalize an existing product or feature, work through these four questions.
1. If we don’t do this, will someone else?
This is the Jobs test. If the disruption is inevitable (the technology exists, the market is shifting, competitors are investing), then the only question is timing. Kodak’s executives believed they could delay digital photography indefinitely. They were wrong by about 15 years.
2. Does the new thing expand the total addressable market, or just redistribute existing revenue?
Apple’s iPhone didn’t just replace iPods. It created an entirely new category of computing. Netflix streaming didn’t just replace DVD rentals. It reached audiences who never would have rented DVDs. The strongest cannibalization decisions open new markets rather than simply transferring customers from one product to another. If your new product only captures existing customers at the same price point, you need a much stronger justification.
3. Can we manage the transition timeline?
Netflix didn’t shut down DVD service overnight. They ran both businesses in parallel for years, giving customers time to shift and giving the company time to build streaming into a profitable standalone business. A McKinsey study found that companies failing to manage product cannibalization effectively can lose up to 20% of their market share. The loss typically comes not from the cannibalization itself but from a poorly executed transition that confuses customers and disrupts revenue before the new product is ready to carry the load.
4. What is the cost of waiting one more year?
This is the question most product managers underweight. The cost of cannibalizing too early is usually recoverable: you can slow down, run both products in parallel, adjust pricing. The cost of cannibalizing too late is often permanent: by the time Kodak was ready to embrace digital, Canon and Sony owned the market. Intel turned down the chance to make processors for the original iPhone in 2007 because the margins didn’t justify cannibalizing their existing chip business. ARM-based processors now dominate the mobile market that Intel refused to enter.
The Uncomfortable Truth
Product strategy requires the willingness to sacrifice current revenue for future relevance. That’s easy to say in a case study and genuinely difficult to do when your quarterly targets depend on the product you’re considering replacing.
But the data is consistent across industries and decades. Companies that cannibalize proactively (Apple, Netflix, Amazon) tend to grow through the transition. Companies that protect existing revenue at all costs (Kodak, Blackberry, Blockbuster) tend to lose everything they were trying to protect.
The product manager’s job isn’t just to optimize what exists. It’s to know when what exists has become the obstacle to what’s next.