Big tech companies have killed 295+ major products since 2000. The average product lifespan is 7.6 years. Google leads with 290+ kills at 7.2 years average. Amazon kills fastest at 1.6 years. Apple keeps products longest at 13.8 years. 31% of products die within 3 years of launch.
Thirty-one percent. That is how many major tech products get killed within three years of launch. We tracked 295+ product shutdowns across Google, Microsoft, Apple, Amazon, and Meta. The average tech product lives 7.6 years before its parent company pulls the plug.
Some never had a chance. Others had millions of users who begged for a reprieve. The company killed them anyway.
Google Kills More Products Than Most Companies Ship
Google has killed over 290 products since its founding. That is not a typo. KilledByGoogle.com, a community-maintained tracker, catalogs every single one. The list includes social networks, messaging apps, hardware lines, developer tools, and entire gaming platforms.
The average Google product lives 7.2 years. But that average hides the real pattern. Google launches fast, measures ruthlessly, and kills without sentiment. Google Wave lasted 1 year. Google Buzz lasted 1 year. Google Allo lasted 3 years. Google+ lasted 8 years but was dead on arrival — it just took Google that long to admit it.
DropThe Data: Across 20,000+ companies in the DropThe database, Google’s parent Alphabet generates $1.8 million in revenue per employee — partly because it does not carry dead products. The company’s willingness to kill is a feature, not a bug. Unprofitable products drain engineering talent that could ship the next billion-dollar line.
How Long Products Actually Last, by Company
Apple keeps products alive the longest. The iPod ran for 21 years. iTunes ran for 18. AirPort routers ran for 18. Apple kills slowly and reluctantly, usually only when the product’s function has been absorbed into something else.
Amazon kills the fastest. The Fire Phone lasted 1 year. Amazon Destinations lasted 6 months. Amazon Halo lasted 3 years. When something does not work, Amazon does not iterate. It walks away.
Microsoft sits in the middle. It gave Windows Phone 9 years. Cortana got 9 years. Internet Explorer got 27 years — longer than most Google employees have been alive. Microsoft’s problem is not killing too fast. It is killing too slow. It spent a decade and billions on a mobile OS that never cracked 3% market share.
The Three-Year Rule
If a product survives its first three years, its odds improve dramatically. 31% of major product kills happen within the first three years. After year three, the product either has enough users to justify itself or gets embedded into workflows that make shutdown painful.
But “painful” does not mean “impossible.” Google killed Google Reader in 2013 after 8 years. It had millions of loyal users. Google killed it anyway because the usage numbers were declining and the team wanted to redirect engineering resources. Reader became the canonical example of why users do not trust Google to maintain products long-term.
That trust deficit is measurable. Every new Google product launches into a headwind of “yeah, but will they kill it in two years?” Google Chat, Google Meet, Google One — every launch carries the ghost of every killed predecessor.
The Money They Burned
Google spent an estimated $300-400 million on Stadia before pulling the plug. Microsoft spent over $2 billion acquiring and operating Mixer before shutting it down and sending its top streamers to Facebook Gaming (which also shut down). Amazon wrote off $170 million on the Fire Phone after selling fewer than 35,000 units in its first month.
Meta spent roughly $10 billion per year on metaverse hardware through its Reality Labs division — which includes the killed Portal line and a series of VR headsets with adoption rates that underwhelmed Wall Street every quarter.
The total cost of killed tech products across these five companies likely exceeds $50 billion over the past decade. That number is impossible to verify precisely because companies bury product-specific costs inside divisional reporting. But the individual write-offs we do know about — Stadia, Mixer, Fire Phone, Windows Phone, Quibi — add up fast.
Why They Keep Launching and Killing
Because the math works. Google generates $350 billion in annual revenue. Spending $400 million on a bet that does not pay off is a rounding error. The cost of NOT trying — missing the next big platform — is existential. Google watched Facebook eat social because Google was too slow. It does not plan to watch anyone eat the next shift.
The launch-and-kill model is not dysfunction. It is portfolio theory applied to product development. Launch 20 things. Kill 18. Keep the 2 that work. Gmail was an internal experiment. Chrome was a side project. Android was an acquisition. The winners pay for all the losers.
The question is whether users should trust any new product from a serial killer. The answer, increasingly, is no — at least not with data or workflows that cannot be exported.
Related DropThe Analysis
This pattern extends beyond product kills. We found that 90% of original S&P 500 companies have been replaced since 1957 — the corporate version of the same cycle. And when companies do survive, their efficiency varies wildly: Nvidia generates $9.5 million per employee while Accenture generates $89,000. The companies that kill fast tend to generate more per person.
Sources
- KilledByGoogle.com — Community-maintained database of 290+ killed Google products
- DropThe.org company database — 20,000+ companies with revenue, employee, and sector data
- Google Q4 2025 earnings report — Revenue and headcount figures
- Microsoft fiscal year 2025 annual report — Mixer write-down disclosures
- Amazon Q1 2015 earnings call — Fire Phone write-off confirmation
- Innosight corporate longevity forecast (2021) — S&P 500 tenure projections