Grab Ditches the Cloud, Saves Millions: A Reality Check
The Great Cloud Reversal?
Grab, the Southeast Asian super-app, just pulled a Basecamp. They yanked 200 Mac Minis out of the cloud and stuffed them into a Malaysian datacenter. The headline? A projected $2.4 million savings over three years. Sounds impressive, right? It is, but let’s dissect this a bit.
Grab's rationale is pretty straightforward. They needed Macs to build iOS apps (a necessity when you're a rideshare and food delivery giant). Initially, renting cloud-based Macs made sense. Scaling from one Mac Pro (the old cylindrical kind, the post notes) to 20 was easy enough with cloud providers. But then they ballooned to over 200 machines, and the costs became, shall we say, noticeable.
The problem, as Grab sees it, boils down to two key factors: expensive macOS build minutes on GitHub Actions (ten times pricier than Linux) and Apple's 24-hour minimum utilization rule for cloudy Macs. Their CI/CD pipeline has daily peaks and weekend lulls. Paying for a full day of cloud Mac time when you're only using a fraction of it? That's a recipe for wasted budget.
They considered virtualized macOS, but weren't thrilled. Performance and stability issues in virtualized environments outweighed the potential gains from better hardware utilization. "We have observed trade-offs in performance or stability," Grab noted. Which is analyst speak for "it was a laggy mess."
So, they went old school. Racks of Mac Minis in a datacenter. Four 42RU racks, to be precise, housing over 200 machines. They even use Jamf for zero-touch provisioning, which is a nice touch. (Makes you wonder why they didn't optimize their cloud setup in the first place.)
The stated benefits are a 20-40% improvement in CI/CD pipeline performance and, of course, the $2.4 million in projected savings. "This project proves that taking ownership of our core infrastructure can be a major competitive advantage," Grab claims.
Digging into the Numbers
Let’s put that $2.4 million into perspective. We don't know Grab's total IT budget, but we can make some educated guesses. Grab is a massive company, processing millions of transactions daily. A reasonable estimate for their annual IT spend would be in the tens, if not hundreds, of millions of dollars. So, $2.4 million over three years? It's not nothing, but it's also not a game-changer. It's a rounding error (relatively speaking).
But let’s not dismiss this so quickly. This move highlights a crucial question that every company should ask: Are you really leveraging the cloud, or are you just throwing money at it?
Companies often default to cloud solutions without fully understanding their own usage patterns. Grab's experience shows that a careful analysis of your CI/CD pipeline, build times, and resource utilization can reveal significant cost savings.
Here's where I think the real lesson lies: This isn't just about cloud versus on-premise. It's about understanding your infrastructure needs and optimizing accordingly. If you have predictable workloads and can effectively utilize dedicated hardware, then owning your kit might make more sense. If your needs are spiky and unpredictable, the cloud's flexibility might be worth the premium.
The question is, how did Grab let this get so out of hand in the first place? Did they not have proper monitoring in place? Were they simply too focused on growth to pay attention to the bottom line? I've looked at hundreds of these kinds of transitions, and the consistent theme is that many orgs treat the cloud as some kind of magical cost-saving button, which it isn't.
And this is the part of the report that I find genuinely puzzling. Why weren't these inefficiencies caught earlier? Was there a lack of communication between the development and finance teams? Or was it simply a case of "cloud-first" dogma overriding common sense?
The Cloud Isn't a Religion
Grab's move isn't a death knell for cloud computing. Not even close. But it is a wake-up call. It's a reminder that the cloud isn't a religion. It's a tool. And like any tool, it needs to be used correctly. Sometimes, that means ditching it altogether. Rideshare giant dumps 200 cloudy Macs, saves $2.4 million
A $2.4 Million Reality Check
Grab's cloud exit isn't a revolution, but a smart course correction. It's a reminder that data trumps dogma. The cloud is a tool, not a miracle cure.
