How do 2-sided marketplaces actually grow?
Marketplaces grow on paid, not product: 82.1% of the 140 marketplace-archetype companies run paid performance marketing, versus just 12.9% PLG and 13.6% content-led [1]. Looked at by revenue model, Marketplace/Transaction-Fee companies are even starker — 100% paid, 0% PLG, and 0% word-of-mouth [1]. Marketplaces buy both sides of liquidity; self-serve and organic referral barely move the needle for them.
82.1% of 2-sided marketplace apps run paid performance, versus 12.9% PLG — July 2026.
The finding: liquidity is bought, not grown organically
The 2-sided marketplace archetype has the highest paid dependence of any archetype (82.1%) and one of the lowest PLG shares (12.9%) [1]. The transaction-fee revenue model makes it even clearer: those companies are 100% paid and 0% both PLG and word-of-mouth [1]. Marketplaces face a cold-start problem on two sides at once, and the tagged base says the answer is paid acquisition to bootstrap liquidity, not a self-serve funnel or referral loop.
The breakdown
Growth-lever mix for marketplaces, two ways [1]:
| Lens | N | Paid % | PLG % | Content % |
|---|---|---|---|---|
| 2-sided Marketplace archetype | 140 | 82.1% | 12.9% | 13.6% |
| Marketplace / Transaction-Fee model | 63 | 100.0% | 0.0% | 4.8% |
For the transaction-fee model, word-of-mouth and network effects are also both 0.0% — paid does essentially all the work [1].
How to apply it
If you're building a marketplace, plan your growth budget around paid acquisition on both supply and demand — 82% of archetype peers and 100% of transaction-fee peers run paid [1]. Don't count on PLG or organic word-of-mouth to solve cold-start; they're near-zero for transaction-fee marketplaces, so a 'build it and they'll refer' plan fights the data [1]. Sales-led has a minor role (11.1% of transaction-fee companies) for seeding large supply-side accounts.
Caveats
Two denominators: the archetype lens uses the 140 marketplace-archetype companies with a growth_engine (from the 757 product_archetype base), and the revenue lens uses the 63 Marketplace/Transaction-Fee companies with a growth_engine (from the 686 business_model base) [1]. Both fields are multi-select. A 0% PLG figure is 'none in this curated sample', not a proof no marketplace can ever run PLG.
The numbers
| Stat | Computed from |
|---|---|
| 82.1% of 140 | selfServeVsSalesByArchetype: 2-sided Marketplace paid_pct 82.1, n 140 |
| 12.9% of 140 | selfServeVsSalesByArchetype: 2-sided Marketplace plg_pct 12.9, n 140 |
| 100.0% of 63 | businessModelXGrowthEngine: Marketplace/Transaction Fees paid_pct 100.0, n 63 |
| 0.0% of 63 | businessModelXGrowthEngine: Marketplace/Transaction Fees plg_pct 0.0, n 63 |
| 0.0% of 63 | businessModelXGrowthEngine: Marketplace/Transaction Fees wom_pct 0.0, n 63 |
Sources & citations
- [1] Lazyweb Research analysis of 757 companies, July 2026. 2-sided Marketplace archetype growth mix from selfServeVsSalesByArchetype (n=140); Marketplace/Transaction-Fee model mix from businessModelXGrowthEngine (n=63). ↩
Source: Lazyweb Research — proprietary analysis of real, in-market app screens. Cite as Lazyweb Research, 2026-07-09.