Every founder has felt it — that gnawing uncertainty about whether the thing you're building actually matters to anyone. You've read The Lean Startup. You've sketched a canvas. Maybe you've even done a few interviews. But there's still this gap between "I have an idea" and "I have evidence."
That gap is where most startups die. Not because the founders lack talent or drive, but because they lack a system for turning assumptions into evidence before they run out of runway.
The Problem with "Just Talk to Users"
The advice is everywhere: talk to users. But nobody tells you how to structure those conversations so they actually produce actionable insights. Most founder interviews are confirmation bias machines — you ask leading questions, interpret polite interest as genuine demand, and walk away feeling validated without actually being validated.
We know because we've been there. Before building Foundry, we spent months in the trenches with early-stage teams. The pattern was always the same: smart people making smart-sounding decisions based on almost no real evidence.
The most dangerous thing in a startup isn't a bad idea. It's a good-sounding idea that nobody actually needs.
Introducing the Validation Stack
The Validation Stack is how we think about the journey from "I have a hunch" to "I have product-market fit." It's not a rigid process — it's a mental model with five layers, each building on the last.
Layer 1: Assumptions
Every startup is a bundle of untested assumptions. Your job isn't to be right — it's to figure out which assumptions are riskiest (high impact, low certainty) and test those first. This is where the Lean Canvas comes in. Not as a static document, but as a living map of your riskiest bets.
Layer 2: Hypotheses
"Early-stage founders struggle to validate ideas before building" is an assumption. "7 out of 10 first-time founders we interview will describe spending more than 3 months building before their first customer conversation" — that's a hypothesis you can actually test.
Layer 3: Evidence
This is where customer interviews, surveys, and market research come in. But the key is structured evidence collection. Every interview should map back to specific hypotheses. Every data point should move the needle on a specific assumption. Without this mapping, you end up with a pile of notes and no clear signal.
This is exactly what Foundry was built for. Every canvas card generates hypotheses. Every hypothesis generates interview questions. Every interview answer feeds back into your canvas. It's a closed loop — not a one-way street.
Layer 4: Patterns
Individual data points are noise. Patterns are signal. When you've done 10 interviews and 8 of them independently mention the same frustration, that's a pattern. When your hypothesis about pricing sensitivity is invalidated across every customer segment you test, that's a pattern too. The goal of this layer is synthesis — looking across your evidence to find the themes that should reshape your canvas.
Layer 5: Conviction
This is where you make the call. Do you have enough evidence to double down? To pivot? To kill the idea entirely? Conviction isn't certainty — it's a rational confidence level based on the evidence you've gathered. The Validation Stack helps you get there faster by making every step intentional.
Why This Matters Now
Building software has never been cheaper or faster. AI tools can generate MVPs in days. But this speed is a double-edged sword — it's now easier than ever to build something nobody wants, and build it really fast.
The founders who win in this era won't be the ones who ship fastest. They'll be the ones who validate fastest. The ones who can move through the Validation Stack in weeks instead of months, building conviction before they build product.
That's what Foundry is for. Not to replace your judgment, but to give it a foundation of evidence. Not to slow you down, but to make sure you're running in the right direction.
If you're an early-stage founder and this resonates, join the beta. We're building this with you, not just for you.