The Starbucks signal: why custom systems are eating SaaS.

Starbucks spends $400 million a year on software. This month they announced they're moving off IBM and Microsoft to build their own custom systems in-house. IBM dropped 3% on the news. Salesforce dropped 4%.

This is, unequivocally, the biggest signal since OpenAI and Anthropic launched their consulting arms: the largest companies in the world are done paying for software that half fits how they work.

What actually changed

For twenty years the math favored SaaS: building custom software was slow and expensive, so you rented a dozen tools and bent your processes around them. AI flipped that math. A system that would have taken a 10-person team now ships in a quarter. When building costs less than renting-and-working-around, the dam breaks.

We saw it coming a year ago and moved everything we build off Airtable and low-code tools — fully custom. It's already compounding.

The playbook behind the headline

1. Get all of the company's data into one system.
2. Build a single operating system for the entire business.
3. Cut out bad, redundant processes.
4. Then layer AI on top — under the correct processes.

That's the core of it: helping companies actually operate better. Not another tool on the pile.

A warning about shortcuts

There are a lot of fly-by-night offerings circulating right now. "Second brains," for example. Throwing scattered data into a second brain while the processes underneath stay broken does nothing. The companies that will destroy their competition over the next five years are rebuilding how they work from the ground up.

Starbucks needs a nine-figure budget and years. A mid-market company can do this in a quarter.

Starbucks is showing you what every serious company does next. The only question is whether your operation gets there before your competitors do.