Friction Was a Forcing Function (3/6)

This newsletter is a six-part sequence built on one anchor: AI is compressing the prototyping cycle — but faster technical iteration without parallel commercial assumption-testing doesn't reduce risk, it accelerates the gap. Perspective 3 takes that gap from one specific angle: friction was a forcing function. Each edition puts one assumption under the knife. None of them end with a fix. The terrain is the point — you locate yourself in it.

The Assumption

The Assumption

More prototyping cycles, made possible by falling cost, automatically produce more validated learning.

The Reality Check

Booking.com runs roughly 25,000 controlled tests a year, and treats each one as a hypothesis to be proven or killed. A funded prototype now costs a founder nothing and takes an afternoon — and increasingly, nothing about it is treated as a hypothesis at all.

The Forensic Analysis

Both organizations operate in a world where the cost of testing has collapsed. Only one of them turns that volume into learning. Stefan Thomke's research on business experimentation, most directly summarized in "Building a Culture of Experimentation" (Harvard Business Review, March–April 2020), makes the mechanism explicit: when the cost of running a test falls, the volume of testing rises automatically — but the discipline required to make that volume mean anything does not rise with it. Companies that scale testing without scaling discipline get what Thomke calls undisciplined experimentation: many disconnected ideas thrown at the wall, generating activity without generating learning. Booking.com avoided that outcome because it built the discipline deliberately — a stated hypothesis, a decision rule, before the test runs, not after.

Expensive hardware iteration never manufactured that discipline on purpose. It manufactured it as a side effect: a $40,000 build required someone to justify the spend, and justification is discipline's cheapest form — you have to be able to say, out loud, in a budget meeting, why this test and not another. Remove the cost, and the side effect disappears with it. What's left is the same amount of discipline, spread across far more attempts — which is a way of saying less discipline per attempt, not less testing. A disciplined test that returns a clear no is not a failure; that's the prototype doing its job. An undisciplined test that returns a result nobody can interpret either way is the actual loss. The venture isn't worse off because it tested cheaply. It's less informed, because it tested without the discipline that makes a result mean anything.

Open Questions

Mechanistic

• What used to make a customer conversation happen before a build started — and what, if anything, triggers that conversation today?
• What's a legitimate reason your organization would build a prototype before any customer conversation happens?

Existence

• Was the minimum viable version written down before anyone touched a tool, or reconstructed afterward to explain what got built?
• Does your process include a decision rule — what result would make you stop — set before the build starts, or

Destruction Desk
We perform autopsies on innovation’s failed assumptions.

This newsletter was edited by Manfred Lueth.

You received this email because you signed up for this newsletter from DestructionDesk.com.
To stop receiving this newsletter, unsubscribe or manage your email preferences