Marketplaces, Brokers and Storefronts

Follow the discussion on x.com

Over the last decade, we’ve seen tremendous growth in commerce transacted online. We’ve gone from buying books to buying houses online in a very short timeframe. We’ve also experienced tremendous growth in B2B commerce, with transaction sizes sometimes even larger than houses. With that, we’ve seen incredible variance in the outcomes attached to startups pursuing e-commerce, despite the tremendous growth (and promise) of this category. 

E-commerce represents one of the most impactful demonstrations that “all revenue is not created equal”. This is a reference to a famed post by Bill Gurley (the “Master of Marketplaces” and an investor in Uber, Grubhub, StitchFix, Nextdoor, Glassdoor, Rover, and Zillow) back in 2011, where he cited the conditions that he felt led to some forms of revenue being more valuable than others. We’ve already opined on the irrelevance of revenue multiples and this post (which I highly recommend reading) qualifies some of the factors that Gurley felt were important in identifying which company characteristics would lead to long-term cash flow potential, making the revenues associated with companies more valuable than others.

As we look toward e-commerce, we see certain characteristics of transaction dynamics that make some of these companies more valuable than others. Companies facilitating commerce online are in the business of matching supply with demand, but not all of these companies are marketplaces (despite many of them attempting to convince you that they are).

As we see it, there are 3 types of transaction facilitators - marketplaces, brokers and storefronts. We say facilitators deliberately as none of these models develop their own products or services, they are simply facilitating the transaction between supply and demand. 

Marketplaces are able to create economic leverage in the value chain, by bringing ALL buyers and sellers to a centralized hub. We deliberately callout ALL, as multi-homing (the concept of a market participant operating across multiple competing solutions) is strongly disincentivized on these platforms. Due to a variety of different factors, buyers/sellers experience high switching costs (the concept of experiencing significant losses in productivity when moving from one platform to another) and/or increasing marginal utility (ie. deriving incrementally higher benefit from each marginal interaction) that strongly disincentivizes multi-homing behavior. For that reason, marketplaces tend to have supply that is both unique and exclusive - if a buyer wants that item, the marketplace is the only solution. This leads to incredibly high organic acquisition and retention of marketplace participants, resulting in higher levels of profitability than other models. Furthermore, the captivity that the marketplace has on each side limits competition, enabling the marketplace to charge whatever it wants. Marketplaces are deliberately low-touch (enabling a high-level of scale) and tend to experience a high degree of fragmentation across buyers and sellers. While marketplaces enforce behaviors and rules for the marketplace (the trust of the marketplace enables transactions to happen more efficiently than they would otherwise), they often have little/no recourse when transactions go awry. They are beautiful engines for long-term profits given all of these dynamics.

A great example of marketplaces is eBay.  I collect baseball cards with my kids, while there are many places to buy retail products (i.e. a box of baseball cards from Topps), there are very few locations to transact on individual cards better than eBay, given the high visibility, liquidity and enforceability of the platform. Users can attempt to scrounge garage sales, collectible stores or facebook groups to find their favorite player, risking fraudulent merchandise or sending money to a buyer who never ships the product, or they can go to eBay and find exactly what they want with ease and have pure visibility over the seller's prior history. Make no mistake, eBay is a terrible product (at least from a UX perspective), but its captivity over the collectibles market results in it selling over $10b in collectibles each year despite exorbitant fees). I often joke that you can assess the strength of a company’s captivity over the market based on the spread between its NPS and its Net Income. eBay has a NPS of -4 and did nearly $4b of pre-tax income last year…that spread shows their strength. 

Brokers differ from marketplaces in their lack of captivity. While there are incremental advantages to consolidating business behind a single transaction facilitator (i.e. moderate switching costs), buyers and sellers freely multi-home across providers. The products/services being facilitated by brokers are often unique, but non-exclusive as sellers frequently list the same product across multiple providers (whereas this practice is punishable on marketplaces like eBay). Brokers largely try to differentiate themselves via the quality of their services (resulting in higher operating expenses, but perhaps higher willingness-to-pay) and finding long-tail buyers and sellers where the transactions are more difficult to discover and curate. These are often the only places where they can yield margin as the more liquid supply in the market gets transacted incredibly efficiently (as in commodity markets). Unlike marketplaces, brokers may take exposure to the goods/services they transact, staking their reputation as a means to distinguish themselves from the pack. While this can lead to variability in experience (some brokers are undoubtedly better than others), it also limits the scale and profitability of these players. 

This isn’t to say that brokers can’t be great businesses. We see incredible examples of brokers in the insurance and freight industry. There are dozens of $10b brokers in the insurance business and dozens more worth >$1b in the freight industry. The challenge, however, is the competition amidst these brokers. There are >400k insurance agencies in the US and >80k freight brokers. The vast majority of these are small mom and pop shops, struggling to earn a profit, while revenue (and margin) is dominated by a handful of players who have leverage their collective balance sheet, marketing power and scale to extract leverage over the supply side (insurance carriers and trucking companies), serve profitable specialty segments unaddressable by smaller players and/or cross-sell ancillary services (ex. many insurance brokerages have leveraged their brand to cross-sell wealth management products to customers and their scale to access new insurance lines from carriers). Brokers can be great businesses, but they are businesses of tactical execution given the constant pressure of the market. This generally results in lower persistency of profitability and lower margins in the market. 

Storefronts are the final form of transaction facilitators. Storefronts provide customers with curated products/services for them to purchase from 3rd parties (we distinguish this from a branded retailer that sells its own products/services). Supply and demand freely multi-home across multiple storefront will little switching cost between competitors. Products are often not exclusive or unique and winners often distinguish themselves on either 1) price (a function of their economies of scale) or 2) ability to curate/discover products. Interaction with supply and demand sides can be highly variable (everything from zero interaction to white-glove recommendations) as exposure to the transaction inventory (some storefront carry inventory and some drop-ship). 

Storefronts can be great businesses as well, but they are very different than brokers and marketplaces. Walmart is a storefront, facilitating commerce of 3rd party products between buyers and sellers and it’s a pretty amazing business despite having inexclusive, ubiquitous products that customers could seemingly purchase anywhere. Walmart’s dominance laid in its ability to establish captivity in local geographies, leveraging scale to price out the competition to the point where they were the only game in town. At that point, customers had no choice but to buy from Walmart, enabling the company to entice customers with limited marketing or discounts. Walmart leveraged the scale they achieved through that strategy to extract leverage over their suppliers, negotiating discounts others couldn’t, resulting in them achieving unprecedented levels of profitability within their customer segment. As you can see, storefronts aren’t necessarily bad businesses, but they often need to find another means of establishing captivity over the network as they lack the customer network effects of marketplaces and have to fight against the forces of multi-homing (Walmart.com lacks the same captivity of its brick and mortar offering) 

Below is a table highlighting some of the key criteria for defining the differences between these 3 business models. 

Again, while some of these business models may be more attractive than others, it’s possible to build a great business in each of these 3 categories. The methods for success, however, are extremely different for each of them. Attempting to be a marketplace (generally a winner-take-all market given the increasing returns to scale) when your market’s competitive dynamics and business model lend itself to be a broker or storefront is a recipe for disaster. Recognizing which category you fall into is essential to determining the capabilities you need to build to develop competitive advantage and crafting your strategic plan toward success.