The Design Sprint Scorecard
Break your founding hypothesis into rows and grade each one red/yellow/green after head-to-head customer tests
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 88%
A testing instrument that decomposes the founding hypothesis into its component claims — right customer, real problem, right approach, chosen over competition, differentiation valued — and grades each after prototype interviews with real customers. Run week over week, it turns subjective 'is this working' into a trackable red-to-green signal of approaching product-market fit.
Origin
A new addition by Jake Knapp and John Zeratsky, described as 'hot off the presses' and not in the original Design Sprint, layered onto the design sprints that follow a Foundation Sprint.
Core principles
- 01Test the hypothesis's variables one at a time, not the product as a vague whole
- 02Show prototypes head-to-head against each other and against real competitors so customers compare out loud
- 03A product 'clicks' with a person or it doesn't — clicking again and again is a strong PMF signal
- 04Interviews are a simulation, not the real world, but a helpful directional signal
How to run it
- 1
Decompose the hypothesis into scorecard rows
Turn each part of the founding hypothesis into a testable row: was this the right customer, do they have the problem, was this the right approach, did they choose it over competitors, was the differentiation valued and motivating.
- 2
Prototype and test head-to-head
Build multiple prototypes (e.g. three with fake brands) and test them against each other and against real competitors like Etsy and Shopify, asking customers to compare aloud.
Pro tip Using AI to generate realistic prototypes is like having an on-standby prototyping team — but don't outsource the thinking about copy, positioning and differentiation.
Watch out Vibe-coded prototypes look believable but tend to be generic and fail to describe what makes the product different, because the model is trained on existing products.
- 3
Grade each row and write a conclusion
Mark each row red/yellow/green per customer and write a conclusion column. A first scorecard full of red is normal.
Pro tip Watch for red flipping to yellow to green across successive weekly sprints — that trajectory is the real signal.
- 4
Revise the hypothesis and sprint again
Feed the learnings back into a lightly-edited founding hypothesis, recruit a fresh slate of customers, and run the next design sprint.
Pro tip Adjust product and marketing/positioning together — they move as one as the prototype becomes more robust.
In the wild
Latchet's first scorecard was 'bleeding' red — right customer and problem, but approach and differentiation failing. Week two showed a few reds flipping to yellow as differentiation dialed in. In a real all-green example shown, a team's entire scorecard eventually turned green.
→ Founders report compressing three to four months of learning into three to four weeks of back-to-back sprints, and customer conversations became far more fruitful once anchored to a hypothesis and prototypes.
Common mistakes
Outsourcing the thinking to AI prototypes
Teams that jumped straight to vibe-coded prototypes found them generic — believable-looking but failing to describe what the product was or how it differed — forcing them to step back and do the thinking they'd skipped.
Reading a red first scorecard as failure
A first scorecard full of red is expected; the value is in knowing precisely what isn't working after just one week, so you can sprint again with a revised hypothesis.
Is it for you?
Best for
Teams that have a founding hypothesis and want a rigorous, trackable way to test it with real customers week over week
Not ideal for
Products already validated with strong PMF signal, where continued simulated interviews add little
From the transcript
“So this scorecard is going to break down the founding hypothesis”
“while you're outsourcing that prototyping work, you don't outsource the thinking”
“you can see when a product clicks with one person and that's a that's a helpful signal”
From the episode
Rapidly test and validate any startup idea with the 2-day Foundation Sprint (from the creators of the Design Sprint)
Jake Knapp & John Zeratsky (Character Capital)