wirefeed
Competitive briefs for AI-native B2B operators
The weekly competitive brief About·Wirefeed
About Wirefeed

Twenty years of watching how the best companies actually work.

I started writing these notes because too much was happening too fast and I needed to understand it. Not the headlines: the playbooks. The rooms. The decisions that looked obvious after the fact and were anything but in the moment.

§ 01

Why this exists

It started at 1am on January 23, 2025. I had just left a dinner at China Live in San Francisco and walked home: 3.16 miles to Mission Dolores. Mark Pincus had said something about control versus value that I needed to write down before it dissolved. So I wrote it down. Then I wrote the next one. Then the one after that.

I have been doing this work for two decades: at Zoho, building Layerpath, sitting in rooms with founders at Sequoia and a16z, watching how companies that look inevitable from the outside were actually made. The playbooks were not documented anywhere useful. The patterns existed. The decisions existed. The reasons those decisions worked were not written down in any form a founder could act on Monday morning.

"In a world of automated summarization and collapsing discovery, proximity is the only durable strategy."

Pathfinder ran for sixteen months as a LinkedIn newsletter. Twelve issues. 1,314 subscribers, zero ad spend. Wirefeed is that work, productized. The judgment is the product. If the brief is wrong, you know exactly who to blame.

§ 02

Four rooms that shaped how I read the category

Four rooms. Each one asked the same question: what actually changed, and what did the operator in the room do next?

Jan 23, 2025
China Live, SF — the walk home was the notebook.
Control is not the same as value. Most founders conflate them until it costs them the company.
Feb 20, 2025
Dylan Field × Andrew Chen, Speedrun.
Four years of silence, one WebGL bet, one engineer coding in his underwear. The quiet ones are usually winning something.
Oct 31, 2025
46th floor, Salesforce Tower.
The FDE is not support. The FDE is an embedded founder. Palantir figured this out early. Everyone else is still calling it "customer success."
Feb 5, 2026
Repo-dive, local machine, 1am.
Not knowing how to read code is a high-interest loan. You pay it on every decision, every hire, every estimate you can't verify.

These are four. There are sixteen months' worth. The pattern that shows up in all of them: the room is never the point. What matters is what someone decided before they walked in.

About · Wirefeed · · © 2026 One operator · wirefeed.ai