The product operating model needs rebuilding

In 1931, Neil McElroy wrote a memo at Procter & Gamble that did something unusual. It didn't propose a new product. It proposed a new system: who owns what, who decides, and how knowledge moves across the organization. That memo created brand management, an operating model that shaped how companies worked for nearly a century.

We're watching another reorganization happen right now, and it's incomplete.

What changed, and what didn't

The factory itself transformed. Code ships faster than any team can process it. Design, content, research, prototyping all rebuilt around speed. You can ship almost anything now.

But the system that decides what gets built, for whom, and in what order never got rebuilt. That work still runs on scattered feedback, undocumented decisions, and context that lives in someone's head. The same setup from five years ago, before agents, before any of this acceleration.

So a faster factory ships more of the wrong things faster. That's not a tools problem. That's a systems problem.

Why every solution so far has failed

We've seen waves of tools promising to fix this: docs, spreadsheets, analytics platforms, roadmapping software, feedback boards, research repositories, issue trackers. Each iteration a bit better than the last. But they all made the same mistake. They organized around artifacts, not context. Feedback in one place, specs in another, decisions somewhere else. A lot of tools. Never a system.

Cursor showed what's possible when you collapse a workflow instead of improving individual tools. Planning, writing, reviewing, debugging, deploying all happened in one environment with agents that understand your codebase. One surface area for the full loop, not locked to one model or vendor.

Product operations never got that moment. Nobody collapsed collecting, understanding, aligning, and shipping into one space. Instead, you became the integration layer, copying and pasting context between tools that don't talk to each other.

The incumbents see this gap too. They're acquiring features, bolting on intelligence, trying to keep you locked into one platform. It's the Salesforce playbook: buy tools, stitch them together, call it a suite. Twenty years later, that strategy is quietly failing because you can't buy your way into a new operating model.

What the market needs is a different kind of environment. Not another silo. Not another vendor trying to own the full stack. A thinking layer that works across your entire ecosystem, no matter which tools you use today or swap out tomorrow.

The shape of what's coming

The early signal is visible. Sophisticated teams are stitching together Claude Code, Cursor, and custom MCP workflows to handle product thinking. But it's single-player right now. Built for one person, not an organization. No shared memory, no compounded learning, and it only runs when you remember to ask.

That era is ending.

Propane collapses the entire product operating stack into one system of context. Not another single-purpose tool owning one step. The layer that owns the entire loop from signal to outcome, regardless of which tools feed into it.

When you start with Propane, the system indexes your company the way Cursor reads your codebase on day one. Your website, features, value proposition, market, competitive landscape. Connect your existing tools and the picture deepens automatically.

What we're shipping now

Context is your global foundation. Mission, strategy, ideal customer profile, the language your team actually uses to describe your product. You set it once. Every person and every agent works from the same base, not a static document buried in a folder somewhere. It's a living system that connects everything downstream.

Signals flow from connectors across your entire stack: feedback, usage data, market signals. Always current, always monitored. You don't have to remember to check. The system watches for you.

Profiles unify every company and person in your world. You start from profiles, not feedback. You're building this feature for these specific people based on what you actually know about them: their behavior, their cohort, their needs. The reasoning flows from people first, not from disconnected feedback.

Canvas is a living document for product thinking where you and your agents work side by side. Signals, reasoning, specs, decisions, all on one surface. You can build the first iteration of an idea in minutes and push it directly into Cursor for development or into Lovable as a prototype.

Spaces give every initiative its own home. Each Space carries its own context, its own signals, its own history. Teams can work in parallel without losing connection to the bigger picture.

All of this works as connected tissue. Every layer feeds the next. The longer your team uses it, the sharper it gets. Better context leads to better decisions, which leads to better products.

Why this matters now

McElroy wasn't describing a product a hundred years ago. He was describing an operating model. The companies that understand their customers best and fastest are the ones that win in their markets.

Engineers got their AI moment. Designers got theirs. Product operating models never did. That's changing now. We're building the full environment for product thinking the way Cursor built the full environment for code.

Product management isn't dead. It's becoming a new operating model that requires new infrastructure to run on. We're building it.