Don't take the cloud call - focus on your startup

Recent events reminded me of a lesson I learned the hard way: don’t take the cloud call. I’m not going to unpack them here. It lit up a part of my brain that’s all too familiar with what comes next.

I’ve seen this movie before.

Blinq. AutoGrab. OpenClub. Firelabs. I’ve been a founder twice, a founding engineer more times than I can count. And I used to take the call.

You know the one — cloud vendor shows up with a smile and a deck.

We love startups. We’re here to help. Let’s get you onto our platform.

They dangle some credits. Maybe offer to help you migrate your infra. Maybe they reach out to your DevOps person or Principal Engineer instead of going through you.

It’s subtle.

But by the time you realise what’s happening, someone’s halfway into building something with their SDK, their API, their service.

And then you’re stuck.

At first, I believed them. I thought the credits were worth it. I thought the support was real. But the truth is simple: Google, AWS, Azure — none of them care about your startup.

I know this won’t win me any friends in the cloud vendor community, but frankly, you’ve already taken enough of my time and money.

They care about beating each other. You’re just a metric in someone else’s OKR doc. A line item in a quarterly review.

Time is runway #

The worst part? These distractions cost you time. And time is runway.

That’s your real currency. Every engineer you hire shortens that runway. Every week spent on infra complexity instead of shipping product tightens the noose.

Startups die from indigestion more than starvation. You don’t need more infra. You need less of it.

You need simplicity. You need to be fast, boring and direct.

I’ve seen the alternative. Startups where “innovation” was code for “sprawl.” Where early engineers wanted to flex. Where sales engineers pushed tools that looked cool and solved nothing. (Tipping a hat to you, Nx)

And I’ve also seen what happens when you resist. When you say no to the shiny stuff. When you ignore the cloud vendors and stick to the basics. These are the teams that ship. That win.

So here’s what I’ve learned:

• The credits aren’t worth the complexity
• The “support” dries up the second you need it
• Vendor lock-in always comes back to haunt you
• Every distracted engineer is a longer time to product-market fit
• The more boring your infra, the faster you ship

(Closed a chunky Series C and need a press release about your “strategic partnership” with a cloud vendor? Knock yourself out. This rant isn’t for you. But if profitability actually matters and runway has a finish line, stop playing vendor chess and get back to building.)

Don’t take the call. Not anymore.

If you’re early-stage, stay agnostic. Keep infra simple. Ignore the sales engineers. You’re not in a fight between clouds. You’re in a fight to stay alive.