Three-Question Market Validation Framework
Before you scale a marketplace, prove demand, supply economics, and willingness to pay — in that order.
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 75%
Before expanding beyond their Palo Alto pilot, DoorDash's founders systematically answered three sequential questions to validate their multi-sided marketplace: do consumers actually want on-demand delivery beyond pizza; can drivers be kept utilized and paid enough during peak and off-peak hours to behave like a real logistics workforce (solved via the gig/1099 model); and will restaurants pay a commission for the incremental orders being generated. Only after all three came back clearly positive did they commit real capital to geographic expansion. This is a lightweight, sequential way to de-risk any new two- or three-sided marketplace idea before scaling.
Origin
Developed by Tony Xu and team during the Palo Alto Delivery pilot period, informed directly by Tony working shifts as a Domino's driver to understand logistics economics.
Core principles
- 01Validate demand cheaply before building infrastructure.
- 02A marketplace is only as strong as its weakest side — test supply-side economics as rigorously as demand.
- 03Willingness to pay must be tested directly with the paying stakeholder before assuming it.
How to run it
- 1
Test raw demand cheaply
Launch a bare-bones MVP (a website with PDF menus and a phone number) to see if real customers will order.
Pro tip Blast a couple of targeted email lists instead of paid marketing to get a true signal.
- 2
Interview merchants for hidden pain points
Talk directly to business owners to uncover problems beyond your original idea (the macaron shop delivery insight).
- 3
Validate supply-side utilization economics
Determine whether the labor supply (drivers) can be kept busy and paid enough across both peak and off-peak hours, using a flexible/variable-cost labor model if needed.
Pro tip A 1099/gig structure lets you avoid paying for idle off-peak capacity.
- 4
Validate the paying stakeholder's willingness to pay
Directly ask the other side of the market (restaurants) whether they'll pay a commission for the incremental value you're creating.
- 5
Expand only after all three answers are yes
Use the validated local playbook as the template for entering new geographic markets.
Watch out Expanding before all three questions are answered risks scaling a broken unit economic model.
In the wild
Tony Xu drove pizza deliveries for Domino's for a week or two to learn how a world-class logistics operation actually worked.
→ He realized restaurants like Domino's ran on paper-based systems and that DoorDash's smartphone-enabled approach could out-execute them.
Common mistakes
Validating demand only
Confirming consumers want the service without testing driver-utilization economics can hide a fatal cost problem.
Ignoring off-peak utilization
Failing to test whether the labor pool can be kept economically utilized outside of meal rushes can quietly break unit economics.
Is it for you?
Best for
Founders building any multi-sided marketplace before committing to geographic or category expansion.
Not ideal for
Single-sided products with no supply-side labor or partner dependency.
From the transcript
“there were three questions that they wanted to answer. One of which was just simply like, do people want this?”
“Two was can they find a way to do this to pay the drivers enough and keep them utilized enough with making trips that they…”
“would the restaurants be happy to pay us for these incremental orders that we're generating for them?”
From the episode
DoorDash