OTP vs Scaling Up (Rockefeller Habits), where the framework ends and the operating layer begins

The question I keep getting asked is some version of, "We run Scaling Up. Do we throw it out for OTP?" The honest answer is no, and the question itself is the wrong shape. You do not pick one over the other. They are not competitors. They live on different layers.

Scaling Up decides what the company should do. OTP is where those decisions get executed, by a workforce that is now part human and part AI. You run Scaling Up on top of OTP, the same way you run a great strategy on top of an actual team. The reframe matters because every framework I am going to write about in this series was designed for a workforce made entirely of people. That assumption is no longer true, and it changes what the operating layer has to be.

What Scaling Up is

Scaling Up, built on Verne Harnish's Mastering the Rockefeller Habits 2.0, is one of the strongest growth frameworks out there. It organizes the whole company around the Four Decisions: People, Strategy, Execution, and Cash. The premise is that if you get those four right and keep them in tension, you can scale without the wheels coming off.

The core artifact is the One-Page Strategic Plan (OPSP), which forces your BHAG (Big Hairy Audacious Goal), your three to five year targets, your annual priorities, and your quarterly Critical Number onto a single page everyone can see. The Function Accountability Chart (FACS) names who owns each function and the numbers that prove they own it. The meeting rhythm runs daily huddle, weekly, monthly, quarterly, and annual, so information moves and decisions do not rot. KPIs sit underneath a single Critical Number, and Themes and MiniGames rally the team around it for a quarter at a time.

It is rigorous, it is proven, and it fits ambitious growth companies well, often ones a little larger than the typical EOS® shop. If you are running it and it is working, keep running it. Nothing here asks you to stop.

What OTP is

OTP stands for Organization Transport Protocol. It is not a strategy framework and not a goal-setting system. It is the operating layer underneath the framework, the place where the work actually gets assigned, executed, and reported.

The specific thing OTP does that nothing else does is give your AI agents a seat. Not a chatbot in the corner of a dashboard. A seat. An agent on OTP has an owner, a scorecard, and KPIs, exactly like a human on your team. It reports its own numbers. It has accountability you can read off a chart. We treat AI as Agent Employees, not as assistance bolted onto software you already had.

That is the wedge. Most tools in this category added an AI helper and called it innovation. The helper drafts, summarizes, and suggests, and a person still owns every outcome. OTP starts from a different place. If an agent is doing real work, it gets a real seat, and the seat carries the same obligations a human seat carries.

The real difference: a framework versus a layer

Here is the cleanest way to see it. Scaling Up assumes the doers are humans. Read the FACS, read the meeting rhythm, read the OPSP. Every box is implicitly a person. That was a safe assumption for the entire history of business, right up until the last couple of years.

OTP does not make that assumption. The accountability chart holds two kinds of worker. Some boxes are people with names, and some boxes are agents with seats. Both have an owner, both have a scorecard, both report KPIs. A framework tells you what good looks like. A layer is what good runs on. When your workforce changes shape, the framework can stay the same while the layer has to absorb the change. That is the whole argument in one sentence.

Where they fit together

This is the part that should feel concrete, because the two genuinely interlock.

The OPSP becomes the agent preamble. Your one page of priorities, Core Values, and the quarterly Critical Number is exactly the context an agent needs to act on your behalf. The same page that aligns your people aligns your agents.
The FACS gains agent seats. You do not redraw your chart. You add rows. The function that used to be one human can become a human plus two agents, each with its own line and its own number.
The Critical Number can be a number an agent reports. If your quarterly focus is speed to lead or proposal velocity, the agent that owns that work posts the number itself. No one transcribes it into a spreadsheet on Friday.
The daily huddle consumes agent output. Instead of waiting for a person to compile status, the huddle reads from agents that already wrote their numbers to shared state overnight. The rhythm Harnish prescribes gets faster because half the reporters never sleep.

None of this replaces Scaling Up. It gives Scaling Up somewhere to land now that some of the doers are not human.

The agent-employee dimension

Two things follow from giving an agent a real seat, and both are hard to get any other way.

First, scorecards populate themselves. A human seat on a FACS depends on someone remembering to update the number. An agent seat reports its own KPI as a byproduct of doing the work. The scorecard stops being a weekly chore and becomes a live readout. That alone changes the texture of the weekly meeting.

Second, and this is the part that compounds, learning moves between organizations. OTP carries something we call OOS, a cross-org learning layer. When an agent on the protocol learns something, that learning can travel to agents in other organizations on the same protocol. A single-company framework cannot do this by design. Scaling Up makes your company better at running Scaling Up. OTP can make your agents better using what other companies' agents already figured out. That is a network effect, and it is the long game. Frameworks are content. They ride on top of the transport layer. The protocol underneath is where the compounding happens.

A practitioner read

Let me be straight about where this stands. Scaling Up is proven over decades and across thousands of real companies. OTP is early. It is the layer we built because we needed it, and I am describing where it is going as much as where it is today. If you want a settled, battle-tested system to run your company, Scaling Up earns that trust and OTP has not earned it yet.

So who should actually care about this post? One person. The operator already running Scaling Up well, who has started hiring AI into real work, and who looks at the FACS and realizes there is nowhere to put it. The agent is doing the job. It has no box, no owner of record, no number on the chart. That gap is the thing OTP exists to close. If that is not you yet, file this and keep running your rhythm. If it is you, the chart in your head already has rows that your current tools cannot draw.

Close

Scaling Up tells you what the company should do. OTP is the layer where a part-human, part-AI workforce actually does it. You are not choosing between them. You are deciding whether your accountability chart is allowed to hold both kinds of worker. Mine has to. I suspect yours will too, sooner than the framework was ever built to expect.

More in this series

This post is part of a series comparing OTP to the operating frameworks companies actually run on. Start anywhere, each one stands alone.

Or read the full series index.

Looking for the head to head against named tools rather than frameworks? See OTP vs Ninety and EOS One.