What Is Multi-Tenanting in Marketplaces?
Turn this article into takeaways for your work.
Each assistant summarizes the article only for you and suggests best practices for your work.
A driver opens the Uber app, sees a slow stretch, and opens a second app. A host lists the same apartment on two booking sites. A seller pushes the same inventory to three e-commerce marketplaces. None of them has left any platform. They're using several at once, and that habit is called multi-tenanting, or multi-homing in the economics literature.
It matters because a marketplace's strongest claim is that users have nowhere better to go. Multi-tenanting is what happens when they do. This article defines the term, explains why the two sides of a market behave differently, covers what the theory says about pricing power, and walks through the levers platforms use to respond, including the legal limits on some of them.
For the underlying idea of why more users makes a platform more valuable, see our reference on network effects. Here we're interested in what happens when those users are shared between rivals.
What multi-tenanting means
A user single-homes when they use one platform for a given activity. A user multi-homes when they use two or more at the same time. The two terms describe behavior, not loyalty. Someone who uses Platform A for 90 percent of their trips and Platform B for the rest is still multi-homing.
Two things are worth separating:
- Joining several platforms. The user has accounts or listings on more than one.
- Transacting on several. The user actually completes activity on more than one.
Plenty of people hold an account they rarely use. The economics bites when the second account is active, because that's when a platform loses the exclusive claim on someone's attention, inventory, or availability.
Rochet and Tirole's foundational paper gives familiar examples: merchants who accept both American Express and Visa, consumers who carry both cards, readers who subscribe to several newspapers, and members of business-to-business exchanges who buy or sell on several exchanges. They call that behavior multihoming and note that competitive prices on one side of a market depend on the extent of multihoming on the other side. That sentence is the core of this article.
Why the two sides behave differently
Multi-homing is rarely symmetric. In most marketplaces one side does it much more than the other, and which side matters for who holds the power.
| Factor | Pushes toward multi-homing | Pushes toward single-homing |
|---|---|---|
| Cost of joining | Free or near-free sign-up | Fees, vetting, onboarding, hardware |
| Cost of using several | Same listing or availability can be reused | Separate inventory, calendars, or training per platform |
| Exclusive benefits | None offered | Tiers, bonuses, or status that need concentration |
| Product differentiation | Platforms look alike | One platform has unique supply or features |
| Frequency of use | Occasional | Daily, habit-forming |
Illustrative table: these are general tendencies, not measured values.
A typical pattern in ride-hailing and delivery is that supply multi-homes and demand mostly doesn't. A driver can run two apps at once; a rider usually opens whichever app is on their home screen. In accommodation it's different: hosts list in several places while guests search across several sites. In B2B software marketplaces, buyers often compare several directories, and vendors list on all of them. Which side multi-homes also shapes liquidity on each platform, because a driver running two apps may show up as available supply on both.
Uber's FY2025 10-K shows how it thinks about the supply side. It says Drivers are independent contractors partly because they can provide services on competitors' platforms, and lists among the factors that could shrink its driver and merchant counts the low switching costs between competitor platforms or services. Grab's 20-F makes the same legal point, listing that drivers are able to provide services on its competitors' platforms. Both filings treat multi-homing as a feature of the business, not an accident.
DoorDash's FY2025 10-K shows the merchant version. It warns that if merchants choose to partner with competitors in a specific geographic market, or engage exclusively with competitors, it may lack the variety or the most popular merchants and its offering would become less appealing to consumers. Note the word "exclusively." Exclusivity is the opposite of multi-tenanting, and a platform worrying about a rival's exclusive deals is a platform that already knows what multi-tenanting is worth.
The economics: competitive bottlenecks
The standard theory comes from the two-sided markets literature of the early 2000s. Two papers anchor it.
Rochet and Tirole (2003) studied how platforms split their prices between the two sides when users can join one or several platforms. Their key finding, quoted above, is that multihoming on one side intensifies price competition on the other, because platforms use low prices to try to steer users toward an exclusive relationship. If a lot of sellers are on both platforms, a platform can't easily charge buyers much, because buyers can reach the same sellers on the rival. The paper also says that the extent of multihoming and the cost of multihoming are among the factors that shape the price allocation between sides.
Armstrong (2006), in the RAND Journal of Economics, built three models: a monopoly platform, competing platforms where agents join a single platform, and what he calls "competitive bottlenecks" where one group joins all platforms. The abstract says the determinants of equilibrium prices are the size of cross-group externalities, whether fees are lump-sum or per-transaction, and whether agents join one platform or several.
The competitive-bottleneck idea is easy to state in plain terms. Suppose buyers each pick one platform, and sellers join all of them. A seller who wants to reach a particular buyer has to go through whichever platform that buyer chose. So each platform becomes a bottleneck, a gatekeeper to its own buyers, even though the platforms compete fiercely for buyers in the first place. The usual prediction from this model is that platforms compete hard on the single-homing side (cheap or free for buyers, sometimes subsidized) and earn their money from the multi-homing side, because that's the side with no alternative way to reach those buyers.
You can see the practical echo in how many marketplaces price. Consumers often pay little or nothing, while the multi-homing side pays the commission. Our article on take rate covers the pricing mechanics; the point here is that who multi-homes helps explain which side carries the fee.
We should be careful about overreading the theory. These are stylized models with stated assumptions. They explain a direction of pressure, not a forecast of a specific company's margins.
Why multi-tenanting erodes the moat
A network-effect moat rests on the idea that the biggest platform is the best one, so users concentrate on it. Multi-tenanting breaks that in three ways.
- It lowers the cost of trying a rival. A supplier already active on two platforms has no switching cost to shift volume. Uber's own risk language cites the low switching costs between competitor platforms. It is also why multi-homing belongs in any discussion of supplier retention: a supplier can drift away by shifting volume without ever closing an account.
- It caps pricing power. Rochet and Tirole's logic applies: if the other side can reach the same counterparties elsewhere, a platform that raises its fee loses volume. A high commission is easy to route around.
- It weakens the data and liquidity advantage. A supplier splitting time between platforms gives each less of their availability, so neither platform reaches the reliability that a single-homed supply base would give it. Local liquidity gets thinner for everyone. Our article on marketplace density explains why that thin-market effect hurts so much in local categories.
None of this makes network effects worthless. A bigger platform still offers more choice and faster matching, which is why many multi-homers send most of their volume to the largest one. But the winner doesn't take all, and the runner-up survives on the volume the leader can't keep exclusive.
It also changes how you read competitive claims. A platform with 80 percent market share and heavy multi-tenanting has a softer position than one with 60 percent and mostly single-homed users. Share counts transactions. It doesn't count how easily they could move.
Levers platforms use
Platforms rarely try to eliminate multi-tenanting. Most try to make single-homing more attractive, or make multi-homing less useful.
Increase switching costs. The platform becomes more costly to leave: reviews and reputation that don't transfer, saved history, integrated payments, and tools that hold the supplier's operating data. Our article on marketplace disintermediation covers the related risk of users moving off-platform, which is a different escape route with the same defense: give users a reason to stay inside.
Reward concentration. Loyalty programs, tiers, and bonuses that depend on volume with one platform tilt users toward single-homing. On the demand side, DoorDash says its membership programs (DashPass, Wolt+, and Deliveroo Plus) aim to lower transactional friction by reducing delivery and service fees. Uber describes Uber One as a single cross-platform membership across rides, delivery, and grocery and retail, with discounts and priority service. Neither filing frames these as anti-multi-homing tools, so that reading is ours: a membership that pays off with volume gives a consumer a reason to put more orders in one place.
Offer services that make the platform the operating system of the supplier's business. The more of a seller's workflow lives on your platform (scheduling, invoicing, inventory, payroll), the more costly it is to run the same business through a rival. See value-added services for how these tools work and what they cost to build.
Win supply acquisition and retention separately. Getting a supplier to sign up is a different task from getting them to prioritize you. Our reference on acquiring supply notes that multi-homing is common and cheap, so the question is rarely "will they join?" but "will they give us their best hours and inventory?"
Use exclusivity, carefully. The most direct tool is to require or pay for exclusivity: a supplier agrees not to list elsewhere, or not to list at a lower price. This is where the law matters.
Exclusivity and its legal limits
Exclusivity and price-parity clauses attack multi-tenanting head on, and regulators have noticed. Three points, framed narrowly.
Price parity for designated gatekeepers in the EU. The Digital Markets Act (Regulation (EU) 2022/1925), Article 5(3), says a gatekeeper shall not prevent business users from offering the same products or services to end users through third-party intermediation services or their own direct channel at prices or conditions different from those offered through the gatekeeper. Article 5(4) adds that the gatekeeper must allow business users, free of charge, to communicate and promote offers, including on different conditions, to end users acquired through its platform. That applies to firms designated as gatekeepers under the regulation, not to every marketplace.
US treatment of anti-steering rules. In Ohio v. American Express (decided June 25, 2018), the US Supreme Court treated a credit-card network as a two-sided transaction platform and held that evidence of a price increase on one side can't by itself show an anticompetitive exercise of market power. Plaintiffs had to show that the anti-steering provisions raised the cost of transactions above a competitive level, reduced transactions, or otherwise stifled competition. The case concerned payment cards, so it isn't a rule for every marketplace, but it shows that US courts evaluate both sides of the platform together.
Worker and merchant classification. As the filings above show, the ability to work for competitors is one of the factors Uber and Grab cite in arguing drivers are independent contractors. A platform that imposes strict exclusivity on drivers may weaken that argument. Whether it does depends on the jurisdiction and the facts, so this is a question for counsel, not a rule of thumb.
The practical takeaway is that exclusivity is a tool with a legal perimeter that depends on your size, your market, and your region. Treat it as something to clear with a lawyer before building it into a growth plan.
Measuring multi-tenanting
You can't manage what you don't count. Common approaches, none with a universal benchmark:
| Measure | What it shows | Where the data comes from |
|---|---|---|
| Share of active suppliers who also appear on a rival | Breadth of multi-homing | Public listings, scraped or surveyed |
| Share of a supplier's volume you capture | Depth of multi-homing | Survey, or inferred from your own utilization |
| Utilization by supplier cohort | Whether your supply is being split | Your own data |
| Share of buyers who compare before booking | Demand-side multi-homing | Surveys, search behavior |
| Volume shift after a fee or incentive change | Switching sensitivity | Your own experiments |
Illustrative list of metrics; set targets from your own data.
Breadth and depth tell different stories. A supplier on three platforms who gives you two-thirds of their hours is a healthier relationship than one on two platforms who gives you a tenth.
Key Facts: Multi-Tenanting in Marketplaces
- Multi-tenanting (multi-homing) means using several competing platforms at the same time. Rochet and Tirole (2003) found that multihoming on one side intensifies price competition on the other.
- Armstrong (2006) models "competitive bottlenecks" where one group joins all platforms, with prices depending on whether agents join one platform or several.
- Uber's FY2025 10-K lists low switching costs between competitor platforms as a factor that could reduce its driver and merchant counts.
- Grab's 20-F cites drivers' ability to provide services on competitors' platforms as a reason they are independent contractors.
- DoorDash's 10-K warns that merchants engaging exclusively with competitors could leave it without enough merchant selection.
- The EU Digital Markets Act, Article 5(3), bars designated gatekeepers from preventing business users from offering different prices elsewhere.
Related Reading
