Level 1
Ship → Solution concept
Problem Discovery
Is the pain real?
Score the customer's relationship to the problem with the Pain Scale. Target zone: Hair on Fire and Incremental Improvement.
A four-level framework that pairs every stage of customer evidence with a build artifact — so founders never have to choose between doing discovery and making progress.
The customer discovery canon — Blank, Fitzpatrick, Constable, Christensen, Torres — was built when building was expensive. Discovery was insurance against burning six months on the wrong code. Talk to customers, the canon said, before you build.
AI broke that deal.
Building a working prototype now takes a weekend. The deterrent is gone. Founders skip discovery, ship to nobody, get a few likes, and call it iteration. The cost of being wrong has collapsed — and so has the discipline that used to come with the cost.
The hard part used to live between insight and build. Now it lives between discovery and insight. The bottleneck is no longer execution. It's whether the inputs to your judgment are real.
Four levels, anchored by a foundational Level 0. Each level has a discovery objective and a corresponding ship artifact. Each level earns the right to ship the next.
Is the pain real?
Score the customer's relationship to the problem with the Pain Scale. Target zone: Hair on Fire and Incremental Improvement.
Where exactly does their workflow break?
Map the customer's current way of getting the job done. Identify the wedge — where pain is real and current solutions are weak.
Will they fire their old way for your new way?
Run probes that test value, not usability. Skin in the Game is the bar.
Can you reach them, repeatably?
De-risk the path to customer. Test channel, motion, unit economics.
Every level on the Ladder produces a build artifact. The ship side is on the framework, not separate from it. Founders don't choose between doing discovery and making progress — discovery is progress, and the artifact is the evidence.
L2 doesn't replace L1. L3 doesn't replace L2. Every customer conversation at L3 opens with a reconfirmation of L1 and the top job from L2. Founders carry earlier levels forward as the substrate of every subsequent conversation. Levels stack; they don't expire.
The Discovery Ladder is only as auditable as the evidence behind it. The DRM is where that evidence lives. Verbatim quotes, not summaries. Behavior, not interpretation. Pain Scale classifications. Disqualification patterns. ICP tags.
Without a DRM, level claims are founder narration. With one, they're evidence.
The Discovery Ladder isn't a tutorial. It's a mirror.
If you've already done customer discovery — talked to humans, run interviews, maybe even shipped something — and you're not sure whether what you've learned is enough to keep building, the Ladder is calibrated for you. It tells you which rung you're actually on, what evidence you're missing, and what you need before the next build.
If you're stuck, the Ladder lets you descend — find the missing rung, fill the gap, and climb back up. Most teams don't need to start over. They need to find the level they skipped.
"Discovery and building are the same activity at different fidelities. Each level earns the right to ship the next."
The Discovery Ladder is a synthesis, not an invention. L1 builds on Talking to Humans (Constable & Rimalovski) and The Mom Test (Fitzpatrick). L2 builds on Christensen's Jobs to Be Done and Torres's continuous discovery. L3 is the spine of Testing with Humans (Constable). L4 borrows from Blank's customer development.
What's new is the structure — pairing each level with a build artifact, making the rungs additive, and calibrating the framework around AI-era founder judgment as the scarce resource.
Built and taught at the NYU Entrepreneurial Institute, in active development with the Startup Sprint cohort.
Download the framework overview — the four levels, the ship artifacts, the Pain Scale, and the DRM, in one PDF.