Ch4 01: The Four Layers of Business Logic#

What if your product is perfect—and your business still dies?

It happens constantly. The technology works. Users love the prototype. The team ships fast. But the venture collapses anyway—because the market logic beneath it was hollow from day one.

Here’s the mistake: most founders treat business logic as a single plane. Either “it works” or “it doesn’t.” They pitch one coherent story, and when every sentence sounds reasonable, they assume the logic holds.

It doesn’t. Business logic isn’t a flat surface. It’s a four-story building. And each floor must bear the weight of everything above it.

The Logic Stack#

Every venture rests on four distinct layers, stacked vertically:

Read them bottom to top—that’s the order most founders build in. Now read them top to bottom—that’s the order that exposes hidden assumptions.

A team that builds L1 first and thinks about L4 last will discover—eighteen months and two million dollars later—that their beautiful product sits inside a market structure incapable of generating venture-scale returns. The product works. The business doesn’t.

Where Connections Fracture#

The dangerous failures don’t happen within a single layer. They happen between layers. Three fracture patterns show up repeatedly:

The Skip. A founder jumps from L1 straight to L4. “We have a great product, and the market is worth $50 billion.” Both statements might be true. But L2 and L3—how you actually capture revenue from that market—remain unexamined. The skip creates an illusion of logic by connecting the endpoints while leaving the middle hollow.

The Contradiction. Two adjacent layers contain assumptions that can’t coexist. A freemium product (L1) targeting enterprise buyers (L2) with a self-serve revenue model (L3). Enterprise buyers don’t self-serve. The product decision and the revenue decision pull in opposite directions—but because each layer sounds reasonable in isolation, nobody notices until sales cycles stretch to nine months and acquisition costs devour the margin.

The Missing Floor. One entire layer simply doesn’t exist. This happens most often with L3. The team has a compelling product and a clear market, but the revenue model is “we’ll figure it out once we have scale.” That’s not a logic gap. That’s a missing floor. You’re asking investors to fund a three-story building where the second floor is made of hope.

Two Ways to Stress-Test#

You can examine your logic stack from either direction. Both matter.

Bottom-up (Builder’s view): Start at L1. Does the product work? Move to L2. Can you reach users? Move to L3. Will they pay? Move to L4. Does the math justify the capital?

This feels natural. It’s also dangerous. Bottom-up carries momentum bias—by the time you reach L3, you’ve invested so much mental energy in L1 and L2 that you rationalize weak revenue logic. “The product is so good, monetization will follow.” It won’t.

Top-down (Investor’s view): Start at L4. What does the exit look like? What revenue trajectory gets there? What market dynamics support that revenue? What product delivers that market position?

Top-down is uncomfortable because it starts with the question most founders want to answer last. But it exposes something bottom-up never does: the hidden assumptions buried in your foundation. When you know you need $30M ARR for a viable exit and work backward through required conversion rates, average contract values, and addressable market segments, you often discover that L1—your beloved product—needs to be fundamentally different from what you’ve been building.

The Structural Integrity Test#

What separates functional logic from decorative logic:

Most pitch decks pass the decorative test. They tell a coherent story. Each slide looks reasonable. The narrative flows. But pull any single layer and ask “what specific assumption connects this to the layer above it?"—and you get silence, or a different answer each time.

The Interrogation Protocol#

Run this against your own venture. Be ruthless.

For each layer, write one sentence completing: “This layer is true if and only if ___.”

Then check: Is that condition within your control? Is it verifiable? Have you verified it?

If your L1 statement reads “This layer is true if and only if users prefer our solution over existing alternatives”—that’s verifiable. Run a test.

If your L3 statement reads “This layer is true if and only if the market shifts toward subscription pricing”—that’s a bet on external forces. You’ve built a load-bearing wall out of a prediction.

Now connect adjacent layers. Write: “L1 supports L2 if and only if ___.” Then: “L2 supports L3 if and only if ___.”

Every “if and only if” containing the words “should,” “eventually,” or “once we scale” is a fracture line. Mark it. That’s where your building collapses.

Common Objections, Directly Addressed#

“But Amazon didn’t have L3 figured out for years.” Wrong. Amazon had L3 figured out precisely—they chose to reinvest revenue into growth. Choosing low margins is not the same as having no revenue model. Bezos could explain the unit economics of every book sold on day one. Can you?

“We’re pre-revenue, so L3 and L4 don’t apply yet.” They apply most. Pre-revenue is when L3 and L4 assumptions are cheapest to test and most expensive to ignore. The cost of discovering your revenue model doesn’t work goes up by an order of magnitude with every layer of infrastructure you build on top of untested assumptions.

“Our market is too new for traditional logic stack analysis.” New markets don’t suspend the laws of business logic. They just mean your L2 assumptions are harder to verify. That makes the stack analysis more important, not less.

Logic Pressure Test #1#

Draw your four-layer logic stack on a single page. For each layer, write one sentence answering: “What must be true for this layer to hold?”

Then look at your four sentences together.

How many of those conditions have you actually verified with evidence—not intuition, not analogy, not “the market is big”—but evidence from your own data, your own customers, your own transactions?

If the answer is zero or one, your logic stack is not a structure. It’s a sketch.

And sketches don’t bear weight.