fig 0. meta_sandwich.layout() — top bread, middle, bottom bread. Tomi from Latitud calls it the sama sando.
**Big tech goes after the two slices of bread. The ham, cheese, and tomato in the middle is yours.**
Years ago, when we were building Sirena, we got into the WhatsApp Business API early — 800 lines, in a market still oriented to giant accounts. In a meeting at Meta's offices, an executive told me something that's shaped how I've looked at every market since.
*"The market is a sandwich,"* he said. *"Depending on how big you are, you're relevant to a different team here."* Then he got specific. We weren't relevant to the bottom-bread team unless we had at least 15,000 lines on the platform — a brutal number at the time. And we weren't relevant to the top-bread team without doing many millions of messages a month.
Two doors closed in one meeting. I left demoralized. Then on the cab ride out it flipped: the middle of the sandwich is the juicy part. The ham, the cheese, the tomato. And nobody's account team was flying in to take it.
## $ the two breads
The two pieces of bread are the top and the bottom. The bottom bread is consumers and very small businesses — millions of them, low ARPU, high volume. The top bread is the Fortune 100/500 — a few hundred giant accounts, each worth tens of millions in contract value. The middle of the sandwich — ham, cheese, tomato — is everyone else.
(In São Paulo this is a misto quente. Same principle. Also tastier.)
## $ why big tech can't help the middle
Big tech companies build product for the two pieces of bread. They have to. Bottom bread is solved with simple, mass-market apps. Top bread is solved with custom enterprise teams that fly to São Paulo to close one bank. WhatsApp had this exactly: WhatsApp Business app for individuals, WhatsApp Cloud API for the giants. Two teams reaching for each other. Eventually they squeeze the middle.
OpenAI is doing the same now. ChatGPT for everyone (bottom bread). Direct enterprise contracts with the largest companies in the world (top bread). They told us at an event with their head of revenue: above a certain ACV, they take it directly. Below, they go through partners. Big bread, small bread, and the middle is open.
**The middle is where founders live.** Companies with 50, 500, 5,000 employees that aren't big enough to get the OpenAI account team's attention, and aren't small enough to use ChatGPT off the shelf. There are a lot of them. Their pain is real. Big tech isn't reaching them yet, and won't for a few years.
## $ three corollaries
The enterprise account team in San Francisco hasn't even heard of your customer. That's not a bug. That's the moat.
1. **Don't build for the very top of the bread.** If your customer talks directly to OpenAI or Anthropic, you'll get squeezed. They have the scale; you don't. Tomi from Latitud says watch out especially for the *crust* — the bottom-of-top-bread companies that look middle but already have a hyperscaler account team flying in.
2. **Be careful at the very bottom.** Consumer AI is gorgeous, but big players are pouring billions into it. Go up a notch — instead of micro-restaurants, mid-size restaurant chains. Instead of micro-influencers, the next tier up.
3. **The bottom of the top bread is wide open in LatAm.** A regional insurer or a mid-market bank in São Paulo or Mexico City would be top-bread in the US — they'd have an OpenAI or Anthropic account team flying in. Here, they don't. They're top-bread companies that the top-bread sales motion never reaches, because those teams literally can't keep up with global demand. Tomás from Latitud has been pointing at this for a while — there's an opportunity in the chaos.
## $ the squeeze cycle
When the bread teams eventually squeeze toward the middle, the next disruption (robotics, something else) will force them to retreat to the bread again. By then you've built something that survives the squeeze.
## $ the rule
Pick the middle. Pick the ham.
*(Be a breadwinner, not the bread.)*