Supply Demand Balance

"In terms of increasing selection, the biggest early unlock was geographic expansion to the suburbs with big enterprise merchants like Cheesecake Factory, Chick-fil-A, and Chipotle, as well as local heroes (like Torchy’s Tacos and Jon & Vinny’s)."

"The research points to two ways to constrain a marketplace: (1) by geography, and (2) by category. If the offering requires supply and demand to be in the same physical location, the constraint is always geographical (e.g. a limited set of markets). Otherwise, it’s category-based (e.g. handmade goods)."

"A rule of thumb was that if we could get to 50-100 concentrated restaurants in a city, we had enough for a consumer to land on the site, cast a wide net, and get a consideration set that was meaningful enough to not be disappointed."

"The lesson we took away from this was that all that matters is supply. So we decided to focus on that above all else. We postponed building brand and delightful product until we had liquidity. Dozens of other startups took another direction. Case after case, it proved a mistake."

"For us, the scarce resource was demand. I don’t think we were ever supply constrained. Supply was easy. It was the dynamics of the market -- if you are someone who loves dogs, you work from home, you are in a situation where an extra $50 is meaningful, why wouldn’t you sit on Rover? It’s a no-brainer."

"The median number of levers that the biggest marketplace companies relied on to kickstart supply growth was just TWO (and the average was 2.5). Though this isn’t necessarily true for everyone, the lesson here for most teams is to focus, focus, focus. Early on, you are more likely than not to find most of your success in a couple of levers."

"One fascinating meta-learning that emerged from this research is how few levers individual companies found success in early-on. The median number of levers that the biggest marketplace companies relied on to kickstart supply growth was just TWO (and the average was 2.5). . . Figure out what those are and double down."

"One of the most significant learnings (and surprises) for me in doing this research was how important one-on-one direct sales was to most early marketplaces. Sales ended up being a crucial lever for about 60% of the companies I talked to — twice as common as the next biggest lever (piggy-backing and referrals)."

"We subsidized leads in almost all of our marketplaces to get them going. We wanted to show new users the quality of our connections and give them a risk-free way to get started. We would then slowly turn on pricing as we proved the value. This helped us build supply in the early days of each marketplace."

"We decided to invest in the software side of the business -- we created a solution that could stand on its own without the demand side. What this meant was building a suite of tools for restaurants to replace their manual booking process (usually a notebook they passed around). Our sales pitch initially was 90% “use our software to better run your restaurant”, and 10% “we’ll help customers find your restaurants and book a table online.”"

"The main thing that I believe really worked was recruiting sellers in person at craft fairs and elsewhere. This was a small activity in terms of human effort, but it scaled beyond that since the sellers marketed themselves."

"It simply means that your biggest constraint to driving additional transactions is a lack of supply (e.g. Airbnb Homes, Uber drivers) or a lack of demand (e.g. Rover dog owners, TaskRabbit customers). In theory, you always want more of both, but in many cases adding more of one side doesn’t actually lead to growth. Your resources are better spent elsewhere."

"To address this, these companies developed models or heuristics to help them understand which side needed their focus. Though never perfect, and constantly evolving, these models were instrumental in helping them focus their resources."

"Amazing! With this loop, Cameo can focus most of its efforts into recruiting celebrities, knowing that much of the other side of the marketplace will come along for free."

"That led us to make it a “one-checkbox experience,” which allowed sellers to take items that were already being listed in groups and cross-post them into the Marketplace Tab as a way to seed the early supply."

"My advice is to track not just raw new supply but ‘activated supply’—supply that has reached some milestone that tells you it’s valuable. For example, a Lyft driver with at least one completed ride, a Patreon creator with over $x in earnings, or an Airbnb home with at least one booking. This additional constraint protects you from creating the wrong incentives (e.g. teams driving tons of useless supply)."