The Marketplace Monetization Map: Complexity and Asymmetry
Marketplace monetization depends on two factors — Transaction complexity and the asymmetry in value to the demand and supply sides
Like other types of startups built on network effects, marketplaces create value by connecting participants. Specifically, they connect demand with supply to enable transactions. This gives them an obvious way to monetize — take a cut of every transaction. However, this cannot be blindly applied to all marketplaces as there are constraints involved. Depending on these constraints, marketplaces can choose between five out of the six possible monetization models.
In my last post, I explained that monetization on interaction networks was a function of network structure, i.e. the nature of relationships between network participants. This is true for marketplaces as well. However, these relationships are governed by different factors on marketplaces. Specifically, they are influenced by the following:
- Transaction complexity: This is a measure of how complex it is for the demand and supply sides to find a “match” on the marketplace and then complete a transaction. This can depend on a range of factors including the nature of supply and the density of supply by region or category. For example, both Upwork and Preply are labor marketplaces with differentiated supply. However, Preply is focused on language tutoring while Upwork is “vertical agnostic”. This means that the demand side on Upwork has more filters to consider — the type of job, required skillsets, project experience, etc. This makes matching on Upwork more complex than on Preply. In addition to this, offline, procedural, or regulatory requirements can increase the complexity of a transaction even after finding a match. For example, completing a transaction after discovering a supplier on a B2B marketplace like Anvyl is much more complex than on Amazon.
- Asymmetry in value between demand and supply: Typically, the value created by a marketplace is asymmetrically distributed across participants. In other words, marketplaces are often more valuable for one type of participant— either the supply or demand side. However, the level of asymmetry is not consistent across marketplaces. For example, supply is more valuable than demand for both Uber and Airbnb. But the degree of this asymmetry is much higher for Airbnb — each additional Airbnb host is more difficult to acquire and brings more value to the marketplace than each additional Uber driver.
Let’s take a look at how these factors influence each potential monetization model.
1. Commissions (or Interaction Taxes)
Commissions (or interaction taxes) are viewed as the “obvious” way to monetize a marketplace. The mechanics of marketplace commissions are simple. The marketplace enables transactions between demand and supply. In exchange for enabling this, the marketplace keeps a percentage of the transaction (often dubbed a “take rate”) as revenue. The commission is applied at transaction completion, which aligns the incentives of all participants — demand, supply, and the marketplace. However, as I mentioned earlier, commissions do not work for all marketplaces.
The effectiveness of this model depends on the combination of complexity and asymmetry. Transactions need to be fairly straightforward (not overly complex) or participants should have limited options to circumvent the marketplace (not overly asymmetric). Complex transactions can be addressed by direct support from the marketplace, i.e. a “managed” marketplace model. And the disintermediation risk caused by high asymmetry can be addressed with SaaS tools for the more valuable side. However, marketplaces with both high complexity and high asymmetry are a bridge too far. In this case, the value that the marketplace brings to “enable” the transaction is not clear enough to validate a commission — this can increase the incentive for participants to go around the marketplace. These marketplaces should ideally avoid commissions.
Uber is the quintessential example of a marketplace with low transaction complexity and asymmetry. Matching demand and supply is simply a function of availability and pickup time — commissions are the obvious fit here. On the other hand, marketplaces like Classpass and Scoutbee avoid commissions entirely.
2. Paywalled Marketplace
Paywalls are the primary alternative for startups that cannot monetize with a commission. A paywall charges one side — either supply or demand — for access to the marketplace. Charging for marketplace access is different from charging for a single-player SaaS tool that complements the marketplace (e.g OpenTable’s reservation software). I’ll come to that later in this post.
For paywalls to be effective, marketplaces need to onboard the more valuable side for free and charge the other to “enable” interactions. As a result, it is most effective when one side is exceptionally more valuable than the other, i.e. it requires high asymmetry. In addition, a paywall creates barriers to entry (and liquidity) on the marketplace. As a result, it is best used when having sufficient demand and supply (liquidity) does not automatically unlock transactions, i.e. marketplaces with very high transaction complexity. In other words, paywalls need both high complexity and high asymmetry. This makes them a very good alternative for marketplaces who cannot monetize with commissions.
Classpass is a good example of a paywalled marketplace. Classpass attempts to match consumers with open spots in fitness classes. The scarcest resource in this marketplace is the availability of open spots, not the availability of demand. Since supply is much more valuable than demand, it becomes easy to charge demand for access. Next, matching is deceptively complex. There are many different types of fitness classes — from yoga to pilates, spin, weight training, etc. So Classpass needs to take consumer interest in a specific type of fitness class (category density) into account. In addition, the distance between consumers and classes (regional density) is also a key factor. To top it all off, it needs to fill these open spots before the scheduled time of the class — especially problematic when bookings are unlikely to be made well in advance. This combination of factors makes it exceptionally complex to match supply and demand in time. To deal with this complexity, Classpass gets consumers (the less valuable side) to commit to a monthly subscription which gives them a specific number of credits. This reduces uncertainty as Classpass knows the distribution of credits by region, and their proximity to classes. This allows it to match credits with classes, based on interest and available time.
B2B supplier discovery marketplaces like Scoutbee also fall into this bucket. For OEMs, the considerations involved in picking a supplier are complex and varied, with long lead times. The availability of high-quality suppliers is by far the most important factor here, making supply far more valuable than demand. As a result, Scoutbee is able to charge customers (OEMs) to access their supplier discovery tool. However, it allows suppliers to create a profile and access OEM tenders for free. On the other hand, Anvyl charges suppliers to list on their marketplace before they can receive inquiries from D2C brands.
3. Premium Network Tier
Unlike a paywall, a premium network tier does not charge participants at the point of entry. Instead, all participants are onboarded for free (with some limitations). This allows the marketplace to connect demand and supply, enable transactions, and collect a commission. On top of this, one side is offered an optional, premium variant of this marketplace with enhanced matching opportunities (not just premium tools). As a result, this is only suited for marketplaces that have high transaction complexity. If matching is easy, there is no pain to drive upgrades. And if matching is complex, it needs to have lower asymmetry in order to maintain a commission-based, “free to access” variant of the marketplace.
Upwork is a great example here. Upwork runs a marketplace that matches freelancers with companies. Companies can post projects or jobs, set pricing, search for freelancers, invite them to apply, hire, manage the project, and make payments. Upwork monetizes this free marketplace with a 10% service fee and a 3% payment processing fee. However, companies are limited to 3 freelancer invites per job post in this free tier. Given the diversity of skillsets and project specifications, this cap can have a meaningful impact on the quality of potential matches. Companies can pay for the premium tier (a monthly subscription) to raise or remove this cap and increase their odds of finding a high-quality freelancer. TheRealReal is another example — its First Look service gives customers early access to listings for a monthly fee.
Note that subscription offerings like Deliveroo Plus and Uber Ride Pass are not premium network tiers. This is because they do not offer enhanced or unique matching opportunities — all users still have access to the same drivers or restaurants without any limitations. Instead, they simply offer discounted prices in exchange for guaranteed monthly spend. As a result, they are better viewed as loyalty programs for high-frequency users, and not a unique way to monetize.
Paid advertising can be a tricky monetization model for marketplaces. On one hand, it holds promise as a high margin revenue stream. However, it can also interfere with supply curation. Commission-based marketplaces aim to highlight the best suppliers to maximize the odds of a match. Allowing suppliers to promote themselves with paid advertising conflicts with this goal. As a result, advertising is most effective when transaction complexity and asymmetry are high enough to make it challenging to use commissions exclusively, but not high enough to move to a true paywall (at least until the marketplace has reached steady-state).
Paywalls have some conflicts with advertising as well. The more valuable side (onboarded for free) has no incentive to advertise and the less valuable side is already paying a fee to reach the more valuable side. At best, advertising can only be a minor revenue stream here. One way to get around this problem is to use paywalled advertising packages — standardized advertising packages that require a recurring subscription to access the marketplace.
Take AutoTrader as an example. It is a used car marketplace that connects both car dealerships and private sellers to prospective buyers. Matching is quite complex here — each car is different and buyer preferences can vary widely. Most importantly, buyers often want to inspect cars in person (and take a test drive) before making a purchase decision. Asymmetry is also reasonably high as dealerships compete with each other for buyers. Because of these factors, AutoTrader primarily makes money in two ways — 1) ad hoc advertisements from private sellers and 2) paywalled advertising packages for dealers.
Shpock is another example. It primarily made money from advertising before introducing a commission-based model. Later, it also added paywalled advertising packages like Shpock+ Shops and Shpock+ Motors for specific types of retailers. This added complexity to the marketplace (and layered new network effects) but also opened up new revenue streams.
5. Complementary Products
Unlike the other monetization models I have explained so far, a complementary product does not generate revenue from access to or interactions on the marketplace. Instead, it involves selling a standalone product to enhance the marketplace’s value proposition. This is typically implemented in the form of a “single-player” SaaS product or optional financial services for one side of the marketplace.
Monetization is not the only benefit of single-player SaaS. It also allows marketplaces to bootstrap one side of the marketplace before opening up the other. This is called the “come for the tool, stay for the network” approach. It can also be monetized when asymmetry is high and the more valuable side of the marketplace faces unsolved, high-value problems. The standalone product is then sold to one side — typically, the more valuable side — to build critical mass and generate revenue at the same time. Note that this is the only form of monetization that can charge the more valuable side of a marketplace directly. The SaaS product locks in the more valuable side and makes it easy to attract the less valuable side which can then be monetized with commissions or a paywall (depending on complexity).
High asymmetry can also cause various forms of financial disadvantages for the weaker side of the marketplace. This creates an opening for the marketplace to offer financial services that ease the imbalance.
Docplanner is a good example of leveraging complementary products to both build liquidity and monetize at the same time. Docplanner sells a SaaS product to doctors to help them digitize patient records and scheduling. This approach helped it attract a critical mass of doctors (the more valuable side) and allow patients to book appointments on its healthcare marketplace. OpenTable, Treatwell, and Lantum also monetize similar complementary products.
On the other hand, Faire and Leaflink monetize by providing complementary financial services. Their suppliers typically have to wait for long periods to receive payments from buyers who have strong bargaining power. This allows them to monetize via accelerated payments and trade finance.
6. Derived Products
In my post on monetizing interaction networks, I described derived products as follows:
Derived products leverage network engagement and interactions to generate an asset that can be productized and monetized directly. These usually take the form of market intelligence or aggregated data products.
Derived products are rarely seen in marketplaces because this data is better used internally to improve matching on commission-based marketplaces or lead generation in paywalled marketplaces. In rare cases, marketplaces targeting complex and opaque markets, like Leaflink, can provide aggregated and anonymized market data as a part of a paid tier. More often, this is used as a public relations or branding tool (e.g. Zillow and Uber).
Marketplace Monetization Map
The relationship between potential marketplace monetization models and the constraints we discussed is summarized below:
Marketplaces with low complexity and asymmetry have just one option to monetize — commissions. This includes marketplaces like Uber and Convoy that rely on automated matching. Marketplaces that do not exhibit both of these traits typically have more flexibility to combine monetization models.
When both asymmetry and complexity are high, commissions are swapped out for paywalls. Marketplaces with just high asymmetry can monetize complementary SaaS or financial products — in combination with commissions at lower complexity (Faire, Docplanner) or with paywalls at higher complexity (Anvyl, Leaflink). Of course, some marketplaces like Choco may fall in a “dead zone” between commissions and paywalls — leaving complementary products as their only option.
Advertising comes into play at more moderate levels of complexity and asymmetry — this can be used by itself (classifieds) or in combination with a form of paywall (Autotrader) or simple commissions (Shpock). And finally, marketplaces with low asymmetry and high complexity (Upwork, TheRealReal) combine an optional, premium tier with commissions.
Keep in mind that the constraints mentioned here aren’t set in stone. Marketplaces can always layer new forms of network effects to change the nature of their product (e.g. Shpock+ Motors). This allows them to inherit the properties of the new network type and change the constraints that shape monetization.