The a16z Marketplace 100 series explores the startups and trends behind largest and fastest-growing marketplace companies. See the complete ranking at a16z.com/marketplace-100.
For marketplaces, “owning demand” is the surest path to sustainable growth. In practice, that means users come directly to you, rather than going through intermediaries like Google or Facebook, and that they exclusively (or almost exclusively) rely on your marketplace instead of comparison shopping with competitors.
It’s one of the key challenges of any marketplace. “Demand efforts” like SEO, SEM, CRO, and amazing UX are necessary, but not sufficient; ultimately, the key to owning demand is through supply strategy.
The right supply strategy varies based on the product being offered and those customers’ needs. If users value consistency and predictability—like, say, UberX—the path to long-term success generally lies in being both better and cheaper than the competition.
However, in most marketplaces, users value having a high variety of supply. For these marketplaces, there are three main strategies for supply differentiation: comprehensiveness, exclusivity, and curation.
A comprehensiveness strategy prioritizes building an extensive and diverse supply base, giving customers many different options for any single transaction. For many winning marketplaces, comprehensiveness is the core value proposition that allows them to own demand and provide a sticky product.
When I was the marketing lead at Grubhub back in 2008, for example, our initial value proposition for demand was comprehensiveness: users wanted to discover who delivered to them, even when the restaurant didn’t provide online ordering through Grubhub. Users often thought that only five or six restaurants would deliver to them, when the reality was at least 10 times that in most Chicago neighborhoods at the time.
The problem was that when entering a market, it would take time to build the number of restaurants that users could order from on Grubhub. Fulfilling the value prop of comprehensiveness required us to be scrappy. We would grab every takeout menu, scan them, and upload them to the website. We would also call every restaurant to map their delivery boundaries. That way, users could discover who delivered, whether or not they could order from them on Grubhub.
This comprehensive supply was Grubhub’s source of differentiation. As we scaled the business and asked users what they most valued about Grubhub, there was a 50/50 split between variety and convenience. If we had not built comprehensive supply, and instead only showed restaurants you could order from, we would have lost the 50 percent of users who mainly valued variety.
Similarly, TripAdvisor also found success by prioritizing comprehensiveness early on, which helped unlock product-channel fit for its SEO-driven strategy. By setting a goal of having a page for every hotel, restaurant, and attraction on the planet, TripAdvisor was able to dominate travel results available on Google. Establishing listing comprehensiveness was the first step toward building review comprehensiveness, solidifying their position as the leading source for travel reviews.
While the comprehensiveness strategy usually provides a compelling value proposition to users, it is difficult to execute. Many marketplaces that pursue it ultimately fail, due to operational complexities or a power imbalance with incumbents. Acquiring and retaining a large, diverse group of suppliers can be challenging for even the strongest operations teams. As the supplier base grows, marketplaces encounter more variation and edge cases, which make day-to-day operations less efficient. Potential challenges include:
Constantly changing inventory: At Grubhub, as we scaled our online ordering product, we found that it was hard to keep up with restaurants opening, closing, and changing delivery boundaries, in addition to keeping the restaurants that worked directly with us on online ordering accurate. As delivery services emerged to cover restaurants who didn’t do their own delivery, we’d have to partner with those services to understand who actually had delivery capability.
Data quality and cleanliness: Many companies attempt to scale through automation, but run into data quality issues that require extensive effort to clean up. Groupon, for example, learned this early on. In order to track the volume of potential customers and sell itself to businesses, the company created Groupon pages for a large number of North American merchants. While that ultimately proved to be a successful sales strategy, it was also incredibly complex, resulting in duplicate merchant pages drawn from various data sources.
Unscalable costs: A growing supplier base can also strain infrastructure and require expensive investments to maintain the service levels customers expect. When Netflix started as a DVD delivery service, it seemed to have every DVD ever made available for immediate delivery. To create the value prop of comprehensiveness, if Netflix did not have the DVD, it would still list it on the site and allow customers to add it to their queue in case it became available. As the company scaled into streaming, it found this value prop impossible to maintain. Streaming rights were too expensive to buy everything, some titles already had exclusive licensing deals elsewhere, and technically encoding that many titles would take forever and cost too much on its own.
Concentrated supplier power: In industries with a naturally low volume of suppliers, such as air travel, comprehensiveness is much easier to achieve and maintain; however, that is usually to the detriment of margins. In fact, it is often the case that the easier it is to become comprehensive, the lower the take rate will be because of supplier power.
In the early 2000s, airlines had so much supplier power that they were able to squeeze online travel marketplaces’ margins. Not offering flights from a major airline would put a marketplace at a huge competitive disadvantage, so the marketplaces were forced to accept the airlines’ terms. Orbitz, Expedia, Travelocity, Priceline, and Booking.com didn’t make any money from flights and made the vast majority of their profits from hotel bookings, where supplier power is spread across tens of thousands of hotel brands. Ultimately, the low margins on flights made it unsustainable for most of these marketplaces to survive as stand-alone businesses, driving a wave of consolidation in the space.
With the exception of marketplaces where supply is concentrated among a few players, most marketplaces will never be able to achieve 100 percent comprehensive supply. Instead, they face the Comprehensiveness Asymptote: as supply coverage increases, each additional unit of supply requires more effort to add. Furthermore, external factors can constantly cause supply coverage to shift, such as new restaurants opening and closing or new home-sharing hosts going on and offline.
Ultimately, achieving 100 percent comprehensiveness is impossible, except for a few verticals where supply is extremely concentrated (which then results in compressed margins). Instead of requiring 100 percent comprehensiveness, marketplaces that successfully execute this strategy pursue “comprehensive enough” supply.
Successful marketplaces determine what is comprehensive enough by understanding of what level of supply fits their particular users’ jobs-to-be-done. In other words, it’s comprehensive enough from the user’s perspective, not the company’s perspective. Marketplaces typically achieve this “comprehensive enough” level of supply by either 1) verticalizing—focusing the marketplace on an increasingly specific product category, such as GOAT and Stock X for sneakers or Reverb for music equipment; or 2) optimizing supply acquisition loops.
What is interesting about “comprehensive enough” strategies is that competitors can come along that redefine the market to expand product market fit again so that the company is no longer “comprehensive enough” to own demand.
Doordash and Postmates did this by offering delivery to restaurants that didn’t offer it—and even to those that didn’t ask for it. Similarly, Booking.com changed the definition of “comprehensive enough” by expanding into boutique hotels, forcing Expedia and Orbitz to widen their focus beyond large chains. To keep competition at bay, “comprehensive enough” companies need to continuously focus on adding selection, even once they have achieved the current target to satisfy demand.
While comprehensiveness has been a winning strategy for many marketplaces, it may be infeasible in some industries. When transitioning from DVDs to streaming, Netflix realized that comprehensiveness was not going to be strategically, technically, or operationally possible. Instead, it moved toward an exclusivity value proposition: being the only source for in-demand content.
Netflix locked up streaming rights for popular content, taking advantage of the fact that most studios were still thinking in terms of traditional cable distribution arrangements. Netflix purchased, and then created original shows, understanding that by building its own distribution platforms, it would own its audience. Netflix’s Chief Content Officer and co-CEO Ted Sarandos once famously said, “The goal is to become HBO faster than HBO can become us.”
There is some hidden truth in Ted’s statement. While the exclusivity approach has helped Netflix build a sustainable business, it has also driven multi-tenanting by the demand side of the marketplace. Multi-tenanting means that users participate in multiple marketplaces to accomplish the same (or very similar) job-to-be-done. By using exclusivity arrangements to make sure supply isn’t multi-tenanting, Netflix has almost guaranteed that there will be multi-tenanting on the demand side. Because a marketplace can’t own all supply, users will go elsewhere to seek out exclusive content, if the content is strong enough. That is why Netflix, Disney+, and HBO can all be successful over time.
It’s important to understand that demand side multi-tenanting does not mean demand-side retention will be poor. It does mean you cannot capture all of customers’ willingness to pay though. Generally, companies shouldn’t care about supply-side multi-tenanting if they own demand, but frequently supply exclusivity drives demand, as is the case with Netflix.
Whether to pursue an exclusivity or comprehensiveness strategy depends on the maturity of the marketplace, customer expectations, and competitive pressures. Ultimately, companies weighing these strategies must answer the question: does exclusivity meaningfully drive demand?
In the entertainment space, exclusive content has been a meaningful driver of demand for marketplaces like Netflix and HBO. For instance, “Star Trek” does not meet the needs of a user who wants to watch “Star Wars.” However, if supply exclusivity does not drive demand, it may not make sense to push for exclusivity, which makes it harder to sign up supply.
Similarly, early on Grubhub allowed restaurants to list themselves on other websites if it helped them become more successful, whereas some of its competitors did not. Today, most restaurants list on various platforms—Grubhub, Uber Eats, and Doordash among them. Thus, given multi-tenanting, isn’t Grubhub struggling to own demand? The answer, of course, is yes—but Uber and Doordash collectively spent multiple billions of dollars to create that competition. Ultimately, the comprehensiveness strategy has become undifferentiated as competitors have been willing to spend heavily to catch up. Now, all of the food delivery companies have a mix of demand they own as well as demand that multi-tenants. This will continue to drive consolidation in the space, as it did in travel and dating.
Even companies that start out with exclusive inventory sometimes move to a comprehensive strategy over time. Airbnb brought an entirely new type of inventory online with sharing people’s homes instead of hotels. More recently, it has been adding more hotel-like inventory over time to become a comprehensive travel brand, as other travel marketplaces onboarded its supply. Hipcamp brought new camping sites online, but still lists the national parks as options as well. These are both attempts to make sure users can solve their travel or camping needs on Airbnb or Hipcamp every time.
The third common strategy marketplaces pursue is curation: specifically, achieving differentiation by hand-picking the supply on the platform. It is important to note that curation is different from verification (e.g., StockX or the RealReal verifying that items are not knock-offs) and from algorithmic recommendations (which do not manually include or exclude supply from the marketplace).
If a comprehensiveness strategy is about reducing multi-tenanting on the demand side, and an exclusivity strategy is about reducing multi-tenanting on the supply side, what is the purpose of curation? In looking at examples of curated marketplaces, the strategy becomes apparent: differentiating by appealing to a niche audience.
HotelTonight curated a select list of hotels available in an area to optimize for speed and price. The thesis is that if variety doesn’t matter that much, beyond a baseline level of selection at a certain quality, speed and price can be a more important differentiator than selection. For certain niches, like business travelers, this can be true.
Caviar took a similar approach in food delivery. The company curated high-quality restaurants and tried to differentiate on service and food quality. But again, this reserves Caviar to niches that are willing to pay more for quality.
Criterion Collection is a parallel story in media. It came to prominence as a DVD company that restored classic films for DVD with extras like commentary and bonus footage, whereas the majority of the DVD industry focused on new releases. However, the company’s focus on prestige films kept it from appealing to the mass market; it eventually became a niche streaming player.
Curation can help provide differentiation in the early stages of a marketplace, but it is difficult to defend against well-resourced competitors. What curation promises can normally be better solved via algorithmic recommendations, but those recommendations require tremendous amounts of training data in the form of ratings, reviews, and many other supply attributes.
In many marketplaces, there aren’t objectively better suppliers than others, just trade-offs of price vs. quality vs. speed that different buyers have different preferences for. So while curation can solve for helping buyers find quality supply, as supply accumulates trust signals that can be leveraged for recommendations, and demand showcases preferences that can be leveraged for personalization, curation loses its strategic appeal. Ultimately, curation is often best used as a secondary strategy: a way for marketplaces to segment their customer bases, as Airbnb did with Airbnb Luxe or Uber Black did with Uber Select.
Being neither comprehensive nor exclusive in markets where variety of supply matters is almost always a death sentence. Oyster, the monthly subscription app that allowed users to read books on their smartphones well before the Kindle app existed, provides one example. The company offered a great reading experience, but it struggled to get licensing deals for all the great books in the world and didn’t have any exclusive content. The segments Oyster found product-market fit in were niche volume readers of specific genres who read so much it destroyed the company’s unit economics. It eventually shut down.
Tidal is a similar story in music streaming. While Spotify and Apple Music are comprehensive enough (though YouTube is better than both), Tidal doesn’t have enough content to own demand, and too few exclusives that quickly end up on Spotify, since Spotify owns the demand. Tidal recently was acquired by Square for nearly $300 million, whereas Spotify is worth over 100 times more.
In thinking through marketplace strategy, it’s important to understand whether you are supplying a commodity good or not. If not, having a plan for whether exclusivity or comprehensiveness will be the value prop to demand and how that shifts over time is important to having a sustainable strategy. This will also help the company think through the roles that features like curation and algorithmic recommendations should play over time. It’s also important to understand that as marketplaces sequence their loops over time, the strategy may change dramatically. In fact, it’s hard to think of marketplaces that haven’t made major adjustments to these strategies as they scale over the long term.
Thanks to Reforge Executives-In-Residence, including Crystal Widjaja (ex-SVP Growth at Gojek), Ravi Mehta (ex-CPO at Tinder), and Dor Levi (ex-EVP at Lyft) for their contributions.
* * *
Casey Winters Casey Winters is the Chief Product Officer at Eventbrite. He formerly worked on growth at Pinterest and Grubhub and has advised marketplace companies including Airbnb, Hipcamp, Ritual, and Thumbtack on growth strategies, metrics, org structure, product development, and more. Casey is a Reforge part...
Anne Lewandowski Anne Lewandowski is an Operator in Residence at Reforge. She was formerly the Growth Product Lead at Apartment List and a Growth Product Manager at Course Hero.