You’ve seen the headlines. Google Cloud is surging. Revenue is up, AI workloads are flooding in, and analysts can’t stop cheering.
But if you’re a cloud customer — someone who has to bet your company’s infrastructure on a platform — those headlines should make your stomach drop.
Because Google doesn’t just kill it. Google kills things.
Remember Google Reader? Killed. Google+, Inbox, Allo, Stadia, Domains, Podcasts — killed, killed, killed, killed, killed. There’s an entire website — the Google Graveyard — dedicated to tracking the body count. Over 290 products buried and counting.
Google doesn’t have a product strategy. It has a product crematorium.
Now here’s the paradox that nobody in the earnings call is talking about: Google Cloud’s current growth isn’t the result of brilliant strategy. It’s the result of a lucky external event — an AI demand surge that happened to wash over Google’s infrastructure like a tidal wave. Google didn’t catch the wave. The wave caught Google.
And that distinction matters more than you think.
Because when the AI hype cycle cools — and it will — what’s left is Google’s internal culture. The same culture that looked at Stadia, a cloud gaming platform with real potential, and said “eh, kill it.” The same culture that built Google Domains, gained millions of users, and then sold it off to Squarespace like it was a garage sale item.
Lucky breaks don’t fix broken instincts. They just delay the consequences.
If you’re an enterprise architect right now, you’re looking at three clouds. AWS is boring but immortal — it will outlive your career. Microsoft Azure is entangled with everything your company already runs — it’s not going anywhere. Google Cloud is… the exciting one. The one with the best tech. The one that makes engineers genuinely happy.
And that’s what makes it terrifying.
Because being great has never been Google’s problem. Commitment has. Google builds beautiful things and then walks away from them like a bad first date. The engineering is world-class. The follow-through is nonexistent.
Think about what that means for cloud customers. Migrating to a cloud platform isn’t like switching email providers. It’s a multi-year, multi-million-dollar decision that reshapes your entire technology stack. You don’t just pick a cloud — you marry one. And Google’s track record suggests it’s the kind of partner that ghosting was invented for.
The top comment on the original analysis nailed it: “The AI demand is a lucky external event, but you only get exposed to these if you stick around. Lesson for Google not to kill products so quickly, maybe?”
Maybe. But maybe is the problem. “Maybe” is not a foundation you build infrastructure on.
When the company holding your data can’t commit to a messaging app for more than two years, what makes you think it’ll commit to your cloud workloads for ten?
Now, to be fair — Google Cloud is different from Google’s consumer experiments. The enterprise business model creates different incentives. Cloud generates real, recurring revenue. Thomas Kurian actually understands enterprise. The Google Cloud team has fought hard to earn trust, and they’ve made genuine progress.
But here’s the twist: that’s exactly what makes the risk invisible.
When a company is obviously failing, the risk is priced in. Everyone knows to stay away. But when a company is winning — when the numbers are going up, when the analysts are cheering, when the platform genuinely is the best technical option — that’s when the risk is most dangerous. Because you stop asking the hard questions. You start believing the momentum will protect you.
Momentum doesn’t protect you. Culture does.
And Google’s culture — the DNA-level instinct to build, get bored, and kill — doesn’t change just because the revenue line looks good this quarter. The AI surge is masking a structural vulnerability: Google Cloud’s long-term competitiveness depends on sustained commitment, and sustained commitment is the one thing Google has never proven it can deliver.
The market is betting on Google Cloud’s growth. It should be betting on Google Cloud’s stamina. Those are two very different things.
If you’re a cloud customer, you need to ask yourself one question before you sign that contract: In five years, will Google still care? Not “will the platform exist” — Google Cloud isn’t going to vanish tomorrow. But will Google still invest in it with the same intensity? Will it still push the roadmap? Will it still treat your workloads as a priority, or will some new shiny thing have distracted leadership by then?
Google’s answer, based on every piece of evidence we have, is a shrug.
And a shrug is not a strategy.
FAQ
Q: Google Cloud is enterprise — they can't just kill it like a consumer app, right?
A: True, and that's why it's still alive. But 'can't kill it' and 'won't starve it' are different things. Underinvestment is the slow-motion version of cancellation, and Google has a black belt in losing interest.
Q: What should cloud customers actually do with this information?
A: Hedge. Multi-cloud isn't just about redundancy — it's about not being held hostage by any one provider's attention span. Build portable architectures so you're never fully dependent on a company that might get bored.
Q: Isn't this just Google hating? They're clearly investing heavily in Cloud.
A: No. Google Cloud has arguably the best tech in the space. That's exactly why its commitment problem is tragic, not trivial. Great products deserve better than Google's graveyard — and recognizing that isn't hating, it's pattern recognition.