Models, Systems, and Loops

A board member once asked our founder a five-word question in the middle of a QBR: “Why is growth slowing down?” We had every dashboard a Series C company could buy. Pipeline by stage, win rate by segment, ARR by region, NPS, cohort retention, a warehouse two engineers babysat full time. Forty tabs of truth, and not one of them answered the question. We could show what had happened in exhausting detail. We could not explain why, and we could not say what would happen next. That gap, between having the data and being able to reason from it, is the gap this chapter closes.

In chapter 1.2 we broke ARR into three engines and learned to reason from first principles instead of by analogy. That gave us the true statements. This chapter gives us the tools to see the business those statements describe: models, systems, and loops. Get these three ideas and the rest of the course stops being a list of tactics and becomes one connected machine.

Data is the readout; a model is the theory

Revenue Architecture draws the line cleanly, borrowing from Scott Page’s The Model Thinker: data is the raw record of what happened in the business, and a model is a simplified representation of how the business works, the thing that lets you interpret, predict, and explain. Data is the readout on the gauge. A model is the theory that makes the readout mean something. No single model is complete or correct, which is why Page’s advice, and the book’s, is to hold several and combine them. Our forty dashboards were data. What the room lacked was a model.

Every model earns its place by improving one of three decisions, what Revenue Architecture calls the 3 Es. Effectiveness asks whether the motion produces at all: velocity, output, scalable growth. Efficiency asks at what cost: is the price of acquiring and serving a dollar of revenue trending down, the test of sustainable growth. Economy asks whether it lasts: quality, productivity, and recurring impact, the test of durable growth. Those are the same three goals the factory chased in chapter 1.1 (achieve growth, lower cost, deliver recurring impact) and the same three maturity phases (scalability, sustainability, durability). The 3 Es are the decision lens. When you weigh any model or any change, you are always asking which E it moves.

Six models, stacked like a building

Revenue Architecture organizes the whole business into six models, and it stacks them like a building because the order carries weight. The bottom three are static. You set them deliberately and rarely, because a choice made low in the stack constrains everything above it. The top three are dynamic. They change with the market, and only the very top layer gets touched every day.

The Revenue Model (how you monetize) sets the price, the win rate, and who carries the risk. The Data Model (the Bowtie) sets what you measure and where. The Mathematical Model sets how those measurements combine into a growth number. Above them, the Operating Model, the Growth Model, and the GTM Model shift with conditions, and the GTM Model is where marketers and sellers actually work.

Read the stack top-down and you see why so many go-to-market fixes fail. A team reworks the GTM layer (a new sequence, a new methodology) while a wrong Revenue Model sits three floors down applying ownership-era math to a subscription business. The repair never holds, because the constraint lives below the layer being touched. This course is built on that stack. Module 2 designs the static three. Module 3 builds the dynamic three. The rest of this chapter is the vocabulary you need before we pour the foundation.

A system is not a process

Two words get used loosely on revenue teams, and Revenue Architecture insists on the difference. A system is a set of interconnected functions working toward one goal. A process is a proven sequence of actions that produces the same outcome every time, which is what makes it possible to document, automate, and scale. Acquisition is a system; the discovery call is a process inside it.

The distinction is not academic. A factory runs on systems and processes, not on staffing. When the answer to “how do we grow faster” is “hire more reps,” you are running the business on people, and that scales in a straight line until it snaps. When the answer is “raise the conversion at this step and cut the wait at that one,” you are running it on a system, and that compounds. Which puts the shape of the system in question.

The funnel was never the whole system

For a century the go-to-market system was drawn as a funnel: a wide mouth of prospects narrowing through awareness, education, and selection down to a closed deal. AIDA, its 1898 ancestor, ended at the purchase. As a picture of acquisition the funnel is fine. As a picture of a recurring-revenue business it is dangerous, because of what it leaves out.

The fix is the Bowtie: take the funnel and mirror it. The narrow point (the won deal) becomes the knot in the center, Mutual Commit, and a second widening structure opens to the right for onboarding, adoption, and expansion. The left side is acquisition, the domain of value and the promise. The right side is retention and expansion, the domain of impact and the fulfillment of that promise, which is the value-versus-impact distinction we planted in chapter 1.1. The Data Model chapter in Module 2 builds the full Bowtie, all eight sub-systems and their metrics. For now, hold the shape: the business is not a funnel that ends at the sale, it is a bowtie that weighs keeping and growing a customer as heavily as winning one.

Open loop, closed loop

The reason the shape matters is feedback. An open loop takes an input, runs a process, produces an output, and stops. A closed loop feeds the output back in as new input so the system corrects itself, the way a thermostat reads the room it just heated and adjusts. Acquisition on its own is an open loop, which is why chapter 1.2’s warning about nonlinearity cuts both ways: small declines across the funnel compound into large ARR swings with nothing to catch them. Retention and expansion form a closed loop: a customer who gets impact renews, expands, refers, and generates the next lead, and that output becomes the next input. Revenue Architecture turns the open funnel into a closed system in three moves.

Do all three and the customer stops being the end of the line. Product-Led Growth is this closed loop in its purest form: users experience the product, tell others, and generate the leads, which is why PLG teams track active users rather than raw signups.

A closed system runs several revenue loops at once. Revenue Architecture names six, each an output fed back as input:

None of these run themselves. Each needs a shared cross-functional goal, and the only goal that unifies marketing, sales, and customer success is impact. That thread is what turns six separate tactics into one self-correcting system.

People problems are usually process problems

Which leaves the question the whole chapter has been circling: who runs the system. Revenue Architecture draws two org shapes. The people-centric org is an inverted triangle balanced on its people. When something breaks, leadership asks who did it and fixes the person (blame, replace, rehire). It feels natural and it destabilizes as it grows, because every new hire is another wobble on the point. The process-centric org is an upright triangle resting on systems and processes. When something breaks, leadership asks what caused it and fixes the process. It steadies as it grows. The framing borrows from James Reason’s work on human error: in a well-run system, most failures are system failures wearing a person’s name.

The payoff is measurable. In a 2023 study of 50,000 opportunities over six months, cited in Revenue Architecture, reps who followed the proven process out-produced those who did not by 1.53x on the same opportunity type and the same volume. Same market, same leads, same product: a 53% production difference that came from following the system instead of improvising around it.

The handoff

You now have the working vocabulary for the rest of the course. Data is the readout; a model is the theory that explains it. The business is six stacked models, three static and three dynamic, and a choice low in the stack constrains everything above it. It runs as a system of processes, not a roster of people. Its shape is a bowtie, not a funnel, and its power comes from closing the loop so customers generate the next customers.

That closes Module 1. We start pouring the foundation in the next chapter with the model at the base of the stack, the one every other model inherits from: the Revenue Model, where you choose how the business makes money and, without meaning to, decide its price, its win rate, its retention, and who carries the risk. Continue to The Revenue Model.