Acquiring Supply for a Marketplace

By Brian Tr · Updated Oct 6, 2026

Supply acquisition is the work of getting sellers, providers, drivers, or listings onto your platform before the platform has proven it can send them money. It is the first hard job in a new marketplace, and it's also a job that never really ends, because suppliers leave, go quiet, and sign up elsewhere.

This article is a reference list of the channels you can use, not a prescription. For each one we cover how it works, when it fits, how it scales, and what can go wrong. The bigger question of whether to chase suppliers or buyers first belongs to supply-first vs demand-first. What happens after a supplier signs up is covered in marketplace seller onboarding, and keeping the wrong suppliers out is the subject of marketplace supply quality. This article sits inside the chicken-and-egg problem section and zooms in on one thing: the channels.

If you want the short theory of the cold-start problem first, our guides to two-sided markets and the marketplace business model cover it. We won't repeat the generic list of tactics here.

Why supply is hard to win and easy to lose

Before the channels, it helps to understand what you're actually competing for. A supplier decides whether to list with you by asking a few plain questions.

• Will I get demand? Orders, bookings, or leads are the main reason anyone joins. A supplier who sees no buyers will leave, however good the tools are.
• Does this make my work easier? Tools for scheduling, invoicing, inventory, or payments can be a reason to sign up even before demand arrives.
• Will I get paid, and how reliably? Payout speed, dispute handling, and payment protection matter more to a small business than a polished brand does.
• What does it cost me to be here? Commission, listing fees, and the hidden cost of managing yet another channel all count.

Then there's multi-homing, which is the habit of listing on several platforms at once. Suppliers do it because it costs little and protects them from any one platform's pricing. You can see the legal fingerprint of this in Grab's own filing. In its annual report, Grab argues its drivers are independent contractors partly because they are able to provide services on its competitors' platforms. A supplier who can work for a competitor will do so whenever the competitor offers better terms that week.

That's why "acquire" is only half the job. A supplier list is made of three things: demand, tools, and payouts. Channels get suppliers to the door. Those three things decide whether they stay and whether they're active.

One finding to keep in mind: most winners relied on few levers

Lenny Rachitsky interviewed founders and early employees from 17 marketplaces, including Airbnb, Uber, Etsy, DoorDash, and OpenTable, and wrote up how each grew its first supply. He reports that the median number of levers a company relied on to kickstart supply was two, with an average of 2.5 (Part 3 of his series, published November 25, 2019). Direct sales, referrals, and piggybacking on an existing network all show up in his list.

Treat the channels below as a menu, not a checklist. You're looking for the one or two that fit your supplier type and your budget, and then doing those well.

The channels at a glance

The rest of the article takes these one at a time.

Channel 1: Direct and manual recruitment

This is the oldest method and the most common. Someone on your team finds suppliers, pitches them, and signs them up, often by phone, in person, or through a personal message.

Lenny's research is the best primary source for how often it shows up. His supply levers are sorted by how commonly companies used them, and direct sales leads the list. His quotes from early employees describe the pattern. A former Airbnb team member says direct sales was critical to get the first listings, and that local teams decided which supply to acquire, down to neighborhood, size, and price. An early OpenTable employee describes hiring a direct sales force that went door to door carrying the software and demoing it. And Casey Winters, ex-growth at GrubHub, describes supply growth as door to door sales, walking into restaurants during their downtime.

How it works. You pick a narrow segment, build a target list, and contact every supplier on it. You remove each objection by name: price, setup, risk, effort.

When it fits. The supply side is professional and reachable, such as restaurants, local trades, or small B2B vendors. It also fits when you need a specific mix rather than just volume. The Airbnb quote above makes that point: field teams could choose which homes to bring in.

Cost and scaling. Cost per supplier is high and scales with headcount. That's fine for a launch city or a first niche. It's a poor way to reach ten thousand suppliers.

Risks. Quality depends on the people recruiting. And suppliers who join because a salesperson asked may not be the ones who actively use the product.

Channel 2: Aggregating existing supply

Instead of persuading suppliers to create listings, you gather listings that already exist elsewhere, either by importing from a source with permission or by collecting public data. The result is a marketplace that doesn't look empty on day one.

Lenny's list has a related lever, piggybacking on an existing network. He notes that for Uber, the first thing a launch team did in a new market was go straight to Craigslist to find supply. That's the mild version: using another platform to find people and then inviting them. The aggressive version is copying their listings.

When it fits. Supply is already public and structured, such as business directories, job posts, or product catalogs, and a supplier would benefit from being found. It's least risky when you import from partners or open data, and most risky when you copy from a competitor.

Cost and scaling. Cheap per listing and fast. It's the one channel where volume can arrive before any supplier relationship exists.

Risks. Three stand out.

1. Legal exposure. The best-known US case is hiQ Labs v. LinkedIn. In its April 18, 2022 opinion, the Ninth Circuit affirmed a preliminary injunction that kept LinkedIn from blocking hiQ's access to publicly available member profiles. But read the limits. That ruling was at the preliminary injunction stage and turned on the federal Computer Fraud and Abuse Act. The same opinion notes that entities that see themselves as scraping victims are not without resort even if that law doesn't apply: state law trespass to chattels claims may still be available, and other claims such as copyright infringement, misappropriation, unjust enrichment, conversion, or breach of contract may also lie. So "it was public" is not a safe answer. Get legal advice for your jurisdiction before scraping at scale, and note that this is a US case, not a rule for Southeast Asia.
2. Ownerless listings. A listing nobody created has nobody to answer a buyer, update availability, or honor a price. Buyers meet stale data, and your first impression suffers.
3. No supplier relationship. Aggregated supply doesn't mean the supplier wants to be there. Some will ask to be removed, and some will treat you as a competitor.

A workable middle path: aggregate, then invite. Show the listing, then tell the supplier you've listed them and offer them control of it. Do this only where you hold the right to use the data.

Channel 3: Tools first

Chris Dixon's essay on this idea, Come for the tool, stay for the network (January 31, 2015), describes the pattern: attract users with a single-player tool that's useful on its own, then convert them into participants in a network. He uses Delicious and Instagram as examples of a tool that grew into a network.

For marketplaces, the tool serves the supplier. It could be scheduling software, inventory management, invoicing, or a booking widget. Lenny's research includes an example in the supply section: an OpenTable employee described building standalone restaurant management software first, which put the product in restaurants before the consumer network existed.

How it works. You build something a supplier would pay for or adopt even if you sent them no buyers. Once they use it daily, you add the marketplace layer.

When it fits. Suppliers have a workflow problem that software can fix, and the sale to them isn't dependent on demand. B2B supply fits this well, since businesses already pay for tools.

Cost and scaling. Expensive to build, and slow. But once it works, it's the lowest-cost way to get supply that stays, because the tool is the reason to log in.

Risks. The tool can succeed while the network never forms, leaving you with a software business and no marketplace. A tool that isn't good enough on its own is just a weak signup form. And you're now competing with dedicated tool vendors.

Channel 4: Incentives and guarantees

Incentives pay suppliers to join or to stay. The forms include earnings guarantees, bonuses for completed jobs, signup payments, and fee waivers.

Lenny's research gives two blunt accounts from ride-hailing. One early employee describes calling limo companies and promising a minimum level of income while they kept the app on. Another describes a guarantee of $40 an hour to drive, conditional on maintaining a 70 percent acceptance rate and keeping the app running. The conditions matter. A guarantee without conditions pays suppliers for doing nothing.

Public companies show the scale of the problem. Grab's filing says it has paid significant amounts of incentives to attract new driver and merchant partners and consumers, and that these incentives have in the past exceeded the commissions and fees it receives. For the years ended December 31, 2025, 2024 and 2023, Grab incurred incentives of $2.3 billion, $1.8 billion and $1.6 billion, of which partner incentives were $1.0 billion, $0.8 billion and $0.7 billion. Grab also states that revenue is generally reported net of these incentives, so the cost shows up directly in reported revenue.

When it fits. Supply is elastic, meaning suppliers can start and stop easily, and you need a market to come alive in a short window. It works best in a bounded city or niche where the guarantee has a clear end.

Cost and scaling. Cost grows with volume. Unlike a field team, a guarantee doesn't fade as you scale, because every new supplier costs you again.

Risks.

• Mercenary supply. Suppliers who come for the payment leave when it stops.
• Open-ended liability. Set a cap, a duration, and a performance condition before you announce anything.
• Fee waivers that never end. A "free for now" fee is hard to bring back. Decide the date up front.

Channel 5: Partnerships with associations and aggregators

Some suppliers are already organized: trade associations, franchise networks, wholesalers, distributors, or software vendors that hold their customer lists. A partnership gives you access to the whole group through one relationship.

How it works. You offer the partner something in return, such as a share of revenue, members-only terms, or a better outcome for their members. The partner promotes you to its base.

When it fits. Supply clusters inside a few bodies, which is common in B2B. A national association of contractors or a distributor with ten thousand retail accounts can be worth more than ten thousand cold calls.

Cost and scaling. Low cost per supplier once the deal exists, but each deal is slow to negotiate and the number of possible partners is limited.

Risks. You depend on the partner. If they decide to build their own marketplace, change terms, or lose members' trust, your supply goes with them. A partner might also push suppliers who don't meet your standards.

Channel 6: Supplier referral programs

A referral program rewards existing suppliers for bringing in new ones. Lenny's research reports several accounts. An early Lyft employee says referrals and ambassador programs brought in double-digit percentages of new supply early on, and that the company targeted students. An early Uber employee says referrals were about a third of first trips, and the referred drivers were also the best ones, with the other thirds split between word of mouth and paid. DoorDash and Caviar contacts also describe referrals as a major lever on the supply side.

When it fits. Suppliers know each other and trust each other's judgment, which is true in trades, delivery, and many professional communities. It needs a base of happy suppliers first.

Cost and scaling. You pay on result, so it's efficient. It grows with your existing base, which means it does little at the start.

Risks. Rewards can attract people who refer anyone for the payout, so tie payment to the new supplier's first completed job rather than their signup. And referral networks tend to be homogeneous, which can narrow your supply mix.

A related, unpaid version is community. Lenny quotes an early Etsy source who says offline community building was one of the most effective early levers: Etsy provided tools and money to leaders who organized "street teams" that promoted the brand at craft and vintage fairs, and the source says those evangelists were not compensated.

Channel 7: Paid acquisition aimed at suppliers

Search ads, social ads, and content aimed at suppliers: "sell on our platform," "earn with us," "list your business."

When it fits. A signup is cheap, suppliers actively search for new channels, and you can measure cost per activated supplier rather than per signup. Gig and delivery platforms commonly run this kind of recruitment.

Cost and scaling. Cost varies and scales with budget, which makes it the easiest channel to turn on and the easiest to waste money on.

Risks. Lenny's research is a caution here. An early Uber contact says paid was one third of first driver trips, and in a separate comment says the paid channel was half the signups but only a third of the first trips, "which tells you it was the least efficient channel." Count activated suppliers, not signups. Otherwise paid acquisition will look better than it is.

Channel 8: Be the supplier yourself

Sometimes the fastest way to get supply is to provide it. The founders or employees do the work, at least in the early days.

Lenny's research has several examples. A TaskRabbit contact says the CEO and early team were Taskers in the earliest days, dispatched from HQ when tasks came in, to make sure every customer had a good experience. A DoorDash contact says everyone in the company dashes once a month. A Rover contact says all early supply was employees.

When it fits. Volume is small, quality has to be guaranteed, and you need to learn the work. It's a good way to understand what a supplier actually goes through.

Cost and scaling. High operating cost, and it doesn't scale like software. It's an early-stage method, not a model.

Risks. You can learn the wrong lessons if your team does the work better than a typical supplier would. And if you keep doing it, you drift into being an operator. That's a valid business, but a different one. See managed marketplace for the version where the operator stays involved by design.

B2B supply versus consumer supply

The right channel depends heavily on who your suppliers are.

For more on how the buyer type changes the build, see B2B vs B2C marketplaces. If you're a B2B founder in Southeast Asia, expect to rely more on direct recruitment, tools, and partnerships than on mass paid acquisition, simply because business suppliers are fewer and decide more slowly. We don't have a sourced SEA benchmark for this, so treat it as a starting hypothesis to test.

Choosing and combining channels

A few practical rules fall out of the above.

1. Match the channel to the supplier, not the other way around. Ask how your suppliers currently find customers and who they already listen to.
2. Pick one or two levers. Lenny's median of two is a warning against spreading thin.
3. Measure activation, not signup. Every channel looks better when you count signups.
4. Set an end date on anything that costs money per supplier. Incentives, guarantees, and fee waivers need a stop.
5. Plan the handoff to onboarding. A channel that delivers suppliers who never finish setup isn't working. That's the next stage, and it's covered in seller onboarding.

Key Facts: Marketplace Supply Acquisition

• Lenny Rachitsky's 17-marketplace study found that the median company relied on just two levers to kickstart supply, and the average was 2.5 (published November 25, 2019).
• Grab incurred $2.3 billion of incentives in 2025, including $1.0 billion of partner incentives, and states that incentives have in the past exceeded its commissions and fees (Grab FY2025 20-F).
• Grab's filing lists the ability to provide services on competitors' platforms as a reason its drivers are independent contractors, which is multi-homing in a legal document.
• In hiQ Labs v. LinkedIn (9th Cir., April 18, 2022), the court affirmed an injunction about public profile access under the CFAA but noted other claims such as trespass to chattels and breach of contract may still apply.
• Chris Dixon's tools-first pattern (January 31, 2015) is to attract users with a useful single-player tool, then build the network on top.
• Common supply levers include direct sales, referrals, piggybacking on existing networks, subsidies, using employees as supply, and paid acquisition. In one Uber account, paid was half of signups but a third of first trips.

Related Reading

• The Chicken-and-Egg Problem in Marketplaces
• Supply-First vs Demand-First
• Marketplace Seller Onboarding
• Marketplace Supply Quality
• Managed Marketplace
• B2B vs B2C Marketplaces
• Two-Sided Markets
• Marketplace Business Model