More than 120 P2P sharing platforms have shut down since 2010 — documented in a single academic database researchers have called the “Sharing Economy Graveyard.” Most of them had convincing pitch decks, genuine user traction, and in several cases tens of millions in venture capital. They failed anyway.
Most P2P goods rental platforms don’t die from lack of demand. They die because the unit economics are structurally harder than Airbnb or Uber — and most founders only figure this out once the runway is gone.
This is not an academic essay. It’s an autopsy, written by someone who navigates these same structural problems every day.
Quick navigation: 1. Geographic chicken-and-egg · 2. CAC/LTV trap · 3. Trust · 4. Frequency · 5. Everything everywhere · Autopsies · Next generation
Researching this topic? We publish data from a live P2P rental marketplace. Open for interviews, data requests, and expert quotes — hi@mietzekater.de
1. The Geographic Chicken-and-Egg Problem
Every marketplace has a chicken-and-egg problem. P2P goods rental has a particularly vicious version of it.
The standard version: lenders won’t join without renters, renters won’t join without supply. This is solvable — subsidise one side, build supply, and wait for demand to follow. Airbnb, Etsy, and Uber all did this.
The goods rental version is spatial. A drill in Brooklyn is useless to someone searching in Chicago. A tent rental in London doesn’t help anyone in Manchester. Which means building a P2P goods rental platform isn’t solving the chicken-and-egg problem once — it’s solving it for every relevant city, every relevant neighbourhood, every relevant category, separately.
The operational consequence: you don’t need a million users. You need roughly 50 active lenders and 200–400 renters within 5 kilometres for a given category to work reliably in a given area. That sounds achievable. The problem is that there are dozens of such areas simultaneously, and each one needs its own supply-building effort.
Platforms that underestimated this launched national marketing campaigns — “Now available across the UK!” — and then found their fulfillment rate below 30% because supply and demand didn’t geographically overlap. Users came once, found nothing, and never came back.
2. The CAC/LTV Trap: Why the Math Kills You
This is the calculation most founders don’t run before they start. Here it is, explicitly.
LTV of a typical lender:
| Variable | Assumption | Reasoning |
|---|---|---|
| Take rate | 15% | Market standard |
| Avg. booking value | €40 | Typical goods rental (tools, cameras, party equipment) |
| Bookings / year | 8 | Realistic for active casual lenders |
| Avg. user lifetime | 2.5 years | Churns after life change (moving, job change, etc.) |
| LTV | €120 |
Wait — that €120 assumes the platform keeps 15%. It doesn’t. Every booking must flow through a payment processor to enable escrow, and that processor takes its cut first:
| Step | Amount (on a €40 booking) |
|---|---|
| Platform commission (15%) | €6.00 |
| Payment processing (~2.9% + €0.25 on €40) | −€1.41 |
| Net platform revenue | €4.59 |
| Effective net take rate | ~11.5% |
That’s 20% of commission revenue going to the payment processor before a single other cost is paid. The adjusted LTV: 0.115 × €40 × 8 × 2.5 = **€92** instead of €120. Across a cohort of 1,000 lenders, that’s €28,000 less contribution margin per year — from payment fees alone.
Against this adjusted LTV of ~€92, here’s realistic paid CAC:
- Facebook / Instagram Ads (lender acquisition): €30–70 per conversion
- Google Ads on branded keywords: €15–40
- Influencer campaigns: often €80–150+ per lender
At a CAC of €50 and a net LTV of €92, you have €42 left for customer support, insurance, infrastructure, and team. This only works on a slide deck.
The only solution: organic CAC close to zero.
When a listing ranks on Google for “pressure washer rental Berlin” and a renter lands on it organically — without the platform having paid for that click — the equation changes completely. The CAC of that booking is approximately €0. The €120 LTV becomes real contribution margin.
This isn’t a marketing strategy. It’s a survival requirement. Platforms that relied primarily on paid acquisition burned their runways before reaching viable unit economics.
What “success” actually looks like: the €1M revenue reality check
Let’s grant the best-case scenario: organic growth worked, you cracked supply density, and you’ve scaled to €1 million in platform revenue (commission income, excl. VAT). Here’s the full cost waterfall.
At 15% take rate and €40 average booking value:
| GMV required | €1,000,000 ÷ 15% = €6,667,000 |
| Transactions | €6,667,000 ÷ €40 avg = ~166,700 bookings/year |
| Monthly | ~13,900 bookings/month |
Now strip away what the platform doesn’t actually keep:
| Cost line | What’s happening | Amount | % of revenue |
|---|---|---|---|
| Payment processing | Stripe charges 2.9% + €0.25 on the full GMV — not just your commission | −€235,000 | −23.5% |
| Damage insurance | ~5% of GMV to an insurance partner for lender protection | −€333,000 | −33.3% |
| Identity verification | €1.50/new user KYC for ~8,000 new users/year | −€12,000 | −1.2% |
| Customer support & disputes | ~2% dispute rate, manual resolution overhead | −€80,000 | −8.0% |
| Infrastructure | Cloud, DB, CDN, email, SMS at 13,900 bookings/month | −€48,000 | −4.8% |
| Gross margin | €292,000 | 29.2% |
After the minimum viable team to run the business:
| Engineering (2 FTE) | −€160,000 |
| Operations & support staff (2 FTE) | −€100,000 |
| Founder salary | −€60,000 |
| Marketing (minimal, mostly organic) | −€80,000 |
| Legal, compliance, GDPR | −€30,000 |
| Accounting & advisory | −€20,000 |
| Tools & G&A | −€15,000 |
| EBITDA | −€173,000 (−17% margin) |
Loss-making — even at €1M revenue, at a standard 15% take rate, with a skeleton team and zero growth budget.
Now the context that makes this number devastating: reaching €6.7M in annual GMV in P2P goods rental likely required €5–15M in external funding. Fat Llama raised ~$12–16M. Peerby pivoted in 2017 when the free model proved untenable, relaunched in 2019, and raised €2.3M in 2022 after returning to profitability. Most category entrants raised similar rounds. That capital subsidised early lender acquisition, absorbed disputes, covered the years of sub-unit-economics operation — and the endgame is a business doing €1M revenue that loses €173,000 per year.
One lever changes the picture significantly: take rate. At 20% instead of 15%, the same €1M revenue needs only €5M in GMV — and because payment processing and insurance are charged as a percentage of GMV (not revenue), lower GMV means lower absolute costs:
| 15% take rate | 20% take rate | |
|---|---|---|
| GMV | €6,667,000 | €5,000,000 |
| Stripe fees | −€235,000 | −€176,000 |
| Insurance (5% GMV) | −€333,000 | −€250,000 |
| Gross margin | €292,000 | €437,000 |
| EBITDA | −€173,000 | −€28,000 |
The €145,000 difference — the gap between burning cash and approaching breakeven — comes from a single variable: how much of each transaction the platform keeps.
3. Trust Is Not a Feature — It Is the Product
There’s a moment in every potential lender’s decision journey that I think of as the “camera moment”: the instant when someone considers handing their £800 mirrorless camera to a stranger they know only from a profile picture.
Platforms die at this moment — quietly, with no data point, no feedback, no churn report. The user puts their phone down and never comes back.
The instinctive founder response: better UX, more reviews, a reassuring FAQ. That’s the wrong diagnosis. What resolves this moment is infrastructure, not interface:
- Identity verification — the renter is a real, identifiable person
- Digital rental contract — rights and obligations are documented
- Payment escrow — funds are secured before the handover happens
- Damage insurance — if something goes wrong, the lender is covered
Without all four of these, every rental is a leap of faith that most people won’t take. The “sharing economy” vision of strangers freely trusting each other is a TED talk abstraction. The operational reality is that trust has to be manufactured by the platform — systematically, before the first listing goes live.
What happens without a trust stack: users route around the platform to informal channels — WhatsApp groups, Facebook Marketplace, personal networks. This isn’t evidence that P2P rental doesn’t work. It’s evidence that the platform failed to solve the trust problem, so users solved it themselves.
Fat Llama understood this early. They built up to £30,000 damage coverage per booking into their product and led with it in every lender communication — not as a footnote in the FAQ, but as the primary reason to list. The result: lenders listing £5,000 cameras and £20,000 film production kits.
4. The Frequency Problem: Why 8 Bookings a Year Kills the Platform
Frequency is the most underrated metric in the sharing economy.
| Platform type | Supply-side transaction frequency |
|---|---|
| Uber driver | 5–15 trips / day |
| Active Airbnb host | 30–100 nights / year |
| P2P goods lender | 2–12 bookings / year |
Why this is fundamentally different: with Uber and Airbnb, LTV compounds through high frequency even at low per-transaction values. An Uber driver doing 10 trips per day has generated 300 transactions in a month — network effects and platform reputation grow quickly. A P2P goods lender doing 8 bookings per year takes over 3 years to reach the same transaction count.
This has two direct consequences:
For LTV: Low frequency × low transaction value = structurally low LTV. That’s the CAC trap from the previous section, compounded.
For network effects: Network effects compound through repeated positive experiences, word-of-mouth, and growing reputation. At 2–12 transactions per year, these effects accumulate agonisingly slowly — and require constant capital injection to survive the gap.
Platforms that tried to solve the frequency problem by expanding categories — more items = more bookings — walked straight into the next trap.
5. The “Everything, Everywhere, Now” Trap
Many P2P rental platforms launched with a promise to make “anything” rentable. The logic was seductive: more categories = more potential lenders = more supply = more demand = faster scaling.
In practice, this strategy produces a platform that is too thin in every category to achieve liquidity. Someone searching for a drill finds 8 items in their city. Someone searching for a camera finds 2. Someone searching for party equipment finds none. The result: poor fulfillment rate in every category, high churn, and no organic growth — because there’s nothing to generate word-of-mouth.
The contrast case: Fat Llama built its early reputation heavily around tech and camera gear. That allowed them to achieve real depth in one segment before expanding. Hygglo in Sweden has remained more focused on outdoor and household equipment than on everything simultaneously.
The correct sequencing: own one category or one city first. Then expand.
This is also true geographically. “Launching in the UK” means nothing if you have 40 items in London, 3 in Manchester, and 1 in Leeds. Your national coverage is a liability if it obscures how thin you are in every individual market.
Platform Autopsies: What Specifically Went Wrong
Five documented cases, with specific failure modes drawn from public sources:
SnapGoods (USA, 2010–2012) SnapGoods was one of the earliest P2P rental pioneers, cited as a poster child in early sharing economy journalism. In August 2012, the platform disappeared — no announcement, no post-mortem, no farewell. Steven Hill of the New America Foundation described it as vanishing “poof, without a trace.” The irony: multiple journalists continued citing SnapGoods as an active example of the sharing economy for years after it had shut down, unaware it no longer existed. Core failure: Too early for smartphone adoption; never reached local density threshold.
Lumoid (USA, 2012–2017) Y Combinator alumni, ~$6M in funding, and a signed partnership deal with Best Buy. Lumoid had every signal of a success story. In December 2017, it shut down. Founder Aarthi Ramamurthy: “Sometimes customer demand, a deal with Best Buy, and investors believing in your vision just aren’t enough, especially when it comes to scaling the business.” The next financing round needed to execute the Best Buy deal never closed. Core failure: Physical equipment rental unit economics; critical funding gap at the decisive moment.
Omni (USA, 2014–2019) $35M in venture capital, a storage-plus-rental model with real demand, and a pivot to whitelabelling for merchants. In November 2019, Omni shut down. TechCrunch called it “another victim of a venture capital-subsidized business offering a convenient service at an unsustainable price.” The merchant pivot required “too big of a shift in behaviour” from both merchants and users. Core failure: Margins too thin against Amazon delivery speed; pivot arrived too late and required too much behavioural change.
Zilok (France, ~2008–2023) Zilok was one of Europe’s oldest P2P rental marketplaces — over 15 years of operation, an extreme outlier for this category. In 2023, even Zilok shut down. Cited reasons: growing regulatory burden, rising operational costs, trust erosion through unresolved disputes. Core failure: Fee structures required for secure transactions eroded already thin margins; regulatory requirements (insurance, tax reporting) grew heavier over time rather than easier.
Circutus (Finland, 2023–2025) The most recent documented failure: Circutus launched in March 2023 and filed for bankruptcy in December 2025. Founder Lotta Lilja: “The user base was growing, but we hadn’t yet had time to accumulate enough customers to make the revenue stream clearly positive.” A committed third angel investor pulled out at the last moment. Core failure: Starting capital insufficient to reach critical mass; a single missing investor decided the outcome.
What the Next Generation Must Do Differently
The five diagnoses above produce five concrete requirements. Not optional best practices — survival conditions.
1. Organic CAC as a strategy, not a cost-cutting measure SEO-optimised listings that rank for local search queries are the only scalable route to lender acquisition. Every platform that relies primarily on paid channels is funding its own failure. The math simply doesn’t close.
2. Full trust stack from day one Identity verification, digital rental contract, payment escrow, damage insurance. Not one of these. All four. Platforms that plan to add these components sequentially lose the lenders who matter most during the interim.
3. City first, country later The first city has to work — measurably, with a real fulfillment rate above 70% in core categories. Only then expand. “Available nationwide” is not a competitive advantage if local density is absent.
4. One category deep before going broad Tool rental in one city is a solvable problem. “Rent anything across Europe” is not a problem — it’s a hundred problems running in parallel with insufficient resources for any of them.
5. Automation makes low frequency economically survivable At 2–12 bookings per year per lender, every transaction step — request, contract, payment, handover documentation, invoice — must run automatically. Every manual touchpoint per booking destroys unit economics at this frequency. Automation isn’t efficiency; it’s the structural fix that makes the model viable at all.
Related Resources
- Sharing Economy Marketplace Blueprint — The practical counterpart to this analysis: how to build for lenders first, solve cold start, and deliver on the three promises that actually make a P2P rental platform work
- P2P Marketplace Economics Calculator — Run your own numbers on the revenue waterfall, LTV/CAC analysis, and break-even lender count for your specific take rate and booking value
- P2P Rental Market Statistics Europe — Market size data, platform landscape, and documented platform shutdowns with sources
- Collaborative Consumption Glossary — Definitions for CAC, LTV, Take Rate, Trust Stack, and other terms used in this article
This analysis is based on publicly available data on failed platforms, academic research on the sharing economy, and operational observations from running Mietzekater. Additions and corrections welcome: hi@mietzekater.de. Last updated: June 2026.