vector research partners ( aka V4RP )
How to Build a Billion-Dollar Marketplace — Do’s and Don’ts from the Growth Expert Behind Grubhub, Pinterest, and More
Casey Winters, the growth brain behind Grubhub, Pinterest & Eventbrite, breaks down exactly how marketplaces go from zero to $1B.
Opening Scene: Marketplaces in Your Life
Think Friday night. You RSVP’d to a local concert via Eventbrite, grabbed a pre-show meal from DoorDash, Ubered to the venue, and threw on that new hat from Etsy. Four marketplaces in a few hours.
We take them for granted now, but their path to product-market fit was brutal. Just ask Casey Winters — the guy who built Grubhub’s growth engine from 40k users across 3 cities to 1M customers in 1,000+ cities. He then leveled up Pinterest’s SEO strategy and helped Eventbrite pivot to virtual events right before the pandemic hit.
Across these gigs, Winters learned the hard truth: marketplaces are black-diamond-slope tough. You’re building for two customer types at once — demand and supply — and misstep on either, and growth stalls.
Now, as an advisor and angel investor, Winters has boiled it down: here’s how to turn a promising marketplace into a multibillion-dollar machine.
Picking an Idea
Do: Go where there’s plenty of suppliers
Your marketplace needs choice. Users must see options. Grubhub has pages of restaurants, Rover has tons of walkers, Amazon has… everything.
“Marketplace founders must think about scalable growth loops from day one,” Winters says.
Narrow marketplaces don’t cut it. Few suppliers = low competition = low take rates = no real business. Expedia makes most money off hotels, not flights, because variety = opportunity.
Don’t: Chase one-trick customers
High-frequency, high-promiscuity buyers are your sweet spot. Haircuts? Low frequency, high loyalty — bad marketplace. Sushi, Thai, Italian? Perfect for repeat orders.
Even Zillow proves you can hack it: low-frequency market, but Zestimate keeps users coming back.
Finding Early Customers
Do: Build a scalable acquisition loop
Early Grubhub got restaurants on board risk-free: no orders, no pay. Low-cost acquisition, easy scale. Supply drives demand via SEO or other content loops — but copy-pasting old playbooks rarely works today.
Winters’ marketplace checklist: “What networks drive attention? How can you edge them to acquire users cheaper?”
Don’t: Skip non-scalable work
DoorDash’s Tony Xu delivered meals himself first year. You still need boots-on-the-ground to validate liquidity.
Don’t: Obsess over software early
Your product is your supply, not your code. Grubhub faxed orders for years. Focus on quality suppliers and service first.
Retaining Customers
Do: Nail the three-step acquisition
- Setup: Give the user the tools to experience value (Grubhub = address input).
- Aha moment: First real taste of value (finding that hidden burrito spot).
- Habit: Repeat engagement.
Winters mapped these to predict retention. Second orders in 30 days? Habit forming.
Don’t: Over-discount
Discounting shifts perceived value to “cheap” rather than “good selection.” Only use when it increases lifetime value, e.g., Grubhub’s $10 mobile app incentive.
Don’t: Expect supply to bring demand
Marketplaces live off owning demand. If suppliers drive their own sales, you’re just a SaaS network. Example: Square vs. Grubhub.
Expanding the Customer Base
Do: Raise supply standards
Bad suppliers = bad reviews = blame marketplace. Kick off underperformers, incentivize top performers (Airbnb Instant Book). Share insights to help suppliers improve.
Don’t: Be rigid
Uber vs. Lyft: different trust strategies early, then Uber doubled down on price — scales better.
Do: Grow with your users
Early adopters tolerate flaws; mainstream users won’t. Improve supply faster than new users get pickier. Grubhub scaled restaurants before acquiring less frequent orderers.
Don’t: Rush market expansion
Funko Pops marketplace = niche. Expand only after conquering the core. Whatnot moved into sports cards, sneakers, and more after validating its original market.
From Startup to Scaleup
Do: Level up your data game
Post-Series A/B, dashboards and subsegment cohorts are essential. Eventbrite: music festivals vs. improv shows — habits look different, data must too.
Don’t: Ignore supply churn
Even two orders/day can retain a restaurant. Track satisfaction and NPS.
Do: Be a fast follower
Rover added dog walking after Wag exploded. Competition can’t be ignored.
Don’t: Fail the marshmallow test
Pinterest invested in international markets, sacrificing short-term US gains. Eventbrite created a consumer team separate from SaaS. Patience pays 10x.
TL;DR: V4RP marketplaces don’t just build software — they orchestrate supply, demand, and habits. Nail the loops, raise the bar for suppliers, grow smart, and play the long game.
