The Revenue Model

The most expensive argument I have sat through was about one field on a price book. Product wanted to launch a new module priced by consumption, metered per job the customer ran. Sales wanted it sold as an annual seat add-on, because that is what their quota model and their pipeline math already understood. Finance wanted annual too, because annual is what the board hears as “recurring.” Four people, one product, and three different businesses on the table, and none of them realized they were arguing about the base layer of the entire revenue stack.

That base layer is the Revenue Model, and it is the first of the three static models Module 1 stacked up. In the models-and-loops chapter we said choices in a lower model propagate up and constrain every model above it (see /course/models-systems-loops/). This is the lowest model there is. Get it right and the metrics, the math, and the operating rhythm inherit something true. Get it wrong and you will build a very precise machine on top of a number that was never real.

The base layer everything else sits on

Revenue Architecture lays the monetization choices along an arc. On the left is Ownership: the customer pays up front to own the asset (on-prem hardware, then hardware plus support, then a perpetual software license). At the top is Subscription: the customer pays to use it, on a schedule that tightens from multi-year to annual to monthly. On the right is Consumption: the customer pays only for what they actually use (per unit, then per action, then per outcome, the “no cure, no pay” end). Past the right edge sits Freemium, where the product is free and monetization is indirect, through data, usage, or a later upgrade.

The full Revenue Model is three things stacked together: the monetization strategy (pay to own, pay to use, pay per use), the pricing and packaging (the tiers and the meters, usually a good-better-best set of three), and the business model (monetization plus packaging plus the go-to-market for one target market). In CPQ this stops being abstract in a hurry. When I stood up a product catalog in Nue, the first real decision was not the tier names or the discount tables. It was whether a SKU is priced per seat per year or metered per unit consumed, because that one attribute decides which downstream fields even make sense: a per-year seat SKU wants a term and a renewal date, a metered SKU wants a rate card and a usage feed. Pick the wrong one and you spend the next quarter bolting fields onto an object that was never shaped for them.

The risk flip

Module 1’s factory chapter planted the distinction this whole model turns on: Value is the promise you make to win the customer, and Impact is the fulfillment of that promise. The arc is really a map of who is on the hook to turn one into the other. In an ownership deal, the buyer takes the asset and owns the work of making it deliver: install it, adopt it, get the result. In a consumption deal, the seller owns almost all of that work, because if the customer never gets to a result, the meter never runs and no revenue is booked. This is First Principle 4 from the first-principles chapter made concrete: recurring and consumption revenue shift the delivery risk from buyer to seller.

The overall level of risk actually drops for both sides as you move right, because shared cloud infrastructure means neither party is building the thing from scratch. What does not drop is the ease of leaving. When a customer pays as they go, they can stop as they go, which is why “recurring revenue is the result of recurring impact” stops being a slogan and becomes an operating constraint. It also changes how acceleration works. On the acquisition side, speed is consecutive: a lead becomes an opportunity becomes a commitment, and you go faster by shortening each step in the chain. On the retention and expansion side, speed is concurrent: adoption, expansion, and account growth run in parallel from the moment of Mutual Commit. The common mistake is treating expansion as consecutive, waiting until a customer looks “happy” before you sell more, when the usage signal to expand often shows up in the first ninety days.

What moves when you move along the arc

Slide along the arc and a whole set of characteristics move together, in the same direction, every time. Revenue Architecture puts rough numbers on each one, and they are worth keeping in your head because they are the sanity check on every plan built above this layer.

Price falls by roughly an order of magnitude at each step. The counterintuitive one is win rate: lower buyer risk makes the close rate worse, not better, because a low-commitment offer floods the top of the funnel with people who were never going to buy. A perpetual deal closes about one in three; a freemium signup converts closer to one in a hundred. Velocity moves the other way, from year-long enterprise cycles down to signups that clear in minutes. Retention is the quiet one. A perpetual license is effectively kept forever, an annual contract holds above 90%, and a month-to-month plan lands around 78% a year, because monthly buyers have more flexibility, less commitment, more price sensitivity, and usually lower engagement. None of these numbers is a target to hit. They are the shape of the terrain: if your annual-subscription motion is forecasting a 1:3 win rate and 100% retention, you are not optimistic, you are using the wrong map.

Four ways to get the Revenue Model wrong

Revenue Architecture names four failure modes, and I have watched every one of them in the field.

1. Applying the wrong metrics. The classic tell is pipeline coverage. A perpetual motion closes about one in three, so “carry 3x pipeline” is sound advice for it. An annual-subscription motion closes about one in five, so the same rep needs 5x coverage to land the same number. Import the ownership rule of thumb and you will forecast with total confidence straight into a hole.
2. Changing the monetization strategy mid-flight. Moving along the arc, perpetual to SaaS or monthly to annual, changes how you price and how you sell at the same time. It is a phase shift, not a pricing campaign, and it takes years, not a quarter. Adobe’s move from boxed licenses to Creative Cloud was a multi-year rebuild of the whole go-to-market, not a new SKU.
3. Running several monetization strategies at once. Offer perpetual and subscription and consumption together and you confuse customers, sellers, and your own reporting all at once. Below roughly $100M ARR the guidance is blunt: pick one monetization strategy and run at most three GTM motions on top of it.
4. Mistaking re-occurring revenue for recurring revenue. The one that costs companies their valuation, and the one worth its own callout.

Where this hands off

The Revenue Model is a decision, not a dashboard. You make it deliberately, you make it rarely, and you make it knowing it sets the price, the win rate, the velocity, and the retention that everything else inherits. If you take one thing into the next chapter, take this: the monetization strategy you pick is the constraint the rest of the stack is built inside, so name it out loud before anyone argues about tactics on top of it.

The problem is that a strategy on an arc is not yet something you can measure. To instrument it, you need a common structure that turns “we run an annual-subscription motion” into stages, volumes, conversion rates, and times that every tool and every team reads the same way. That structure is the second static model, and it is the most operations-relevant chapter in the whole course. Next, we build the Data Model: the Bowtie, the standardized, tool-agnostic map of the entire customer lifecycle that the metrics spine and the math both run on.