What Is Coworking: Models & Economics 101

The entry page for anyone taking a coworking space from idea to open day. It answers two questions before you touch a single vendor: what business am I actually in, and how does the money work? Coworking looks like a hospitality or community business from the front desk, but underneath it is a real-estate arbitrage with a services layer bolted on. Understanding which model you're running — and the unit economics beneath it — is the difference between a space that compounds value and one that quietly bleeds. Terms in bold on first use are defined in the Glossary.

Coworking is the business of taking space, dividing it into smaller sellable pieces, wrapping those pieces in services, community, and design, and selling flexible access at a margin over what the space itself costs. The core trick is arbitrage: you pay for space in one large, long, cheap-per-unit block and re-sell it in many small, short, expensive-per-unit pieces. Everything else — the coffee, the events, the fast WiFi, the brand — exists to justify the markup and keep members from leaving.

That framing matters because it exposes the central risk. In the classic model your costs are long-term and fixed (a lease you signed for years) while your revenue is short-term and cancellable (members who can leave with a month's notice). Every business model below is, at heart, a different answer to who carries that mismatch.

The four business models

Operators run one of four deal structures with the property. They are not mutually exclusive — a growing operator often runs several across a portfolio — but each has a distinct risk profile, capital requirement, and control trade-off.

1. Lease arbitrage (the conventional model). The operator signs a conventional commercial lease — historically five to ten years — pays the landlord a fixed rent regardless of how the space performs, fits it out at their own cost, and keeps all the upside. Maximum control and maximum reward if the site fills; maximum risk if it doesn't, because the rent is due whether occupancy is 90% or 40%. This is the model WeWork scaled aggressively, and its collapse is the cautionary tale for the whole model type: WeWork took on long-term lease liabilities against short-term, cancellable revenue, and when office demand fell and rates rose the mismatch became fatal. Its bankruptcy filings showed roughly US$18.7 billion in liabilities against US$15.1 billion in assets, and the company never posted a quarterly operating profit in its history. The model isn't broken — most profitable independents run it — but it punishes over-expansion and thin balance sheets brutally. Sources: The Vertical, CreditRiskMonitor, Dave Friedman.

2. Management agreement (asset-light / revenue-share). Instead of leasing, the operator and landlord partner. The landlord owns the space, funds most or all of the capital fit-out, and carries the property risk; the operator brings the brand, the team, the systems, and the members, runs the space, and the two split revenue or profit on an agreed formula (sometimes with a minimum guarantee to the landlord). This transfers the cost-revenue mismatch onto the party best able to bear it — the building owner — and lets the operator grow without signing balance-sheet-crushing leases. It is the defining structural shift in flex over the last few years. Industrious built its business on it: an asset-light, profit-sharing partnership model that CBRE took to full ownership in early 2025 at an implied enterprise value of roughly US$800 million, with revenue compounding over 50% a year since 2021. The trade-off is control and margin ceiling — you share the upside, and the landlord has a seat at the table. Sources: OfficeRnD, technologywithin, CBRE, Facilities Dive.

3. Franchise. The operator buys the right to run a location under an established brand's name, systems, and sales network, paying an upfront franchise fee plus ongoing royalties. IWG (Regus, Spaces, HQ) runs the largest programme: initial investment for a US location runs roughly US$650,000 to US$1.7 million-plus per centre with a US$50,000 franchise fee, ongoing royalties around 6% of gross revenue plus a ~2% marketing levy, and franchisees typically needing net worth around US$1 million with US$350,000+ liquid. IWG quotes a global average of roughly US$550 per square metre to build out and a 3–4 year payback. You trade margin and independence for a proven brand, a booking network, and operational playbooks — attractive to real-estate investors entering flex without wanting to build a brand from scratch. Sources: IWG / TopFranchise, Franchise Gator, Sharpsheets.

4. Landlord-operated (owner-operator). The building owner skips the third-party operator entirely and runs flex space in-house, often under their own brand, to add flexibility to their portfolio and capture the full margin. Increasingly common as landlords watch operators earn the spread on their buildings and decide to keep it. The owner carries all the risk and needs to build operational capability they may not have — which is exactly why the management-agreement model exists as the middle path. Source: Cushman & Wakefield via Facilities Dive.

A note on members: license, not lease

Whichever model the operator runs upstream with the landlord, the agreement it signs downstream with a member is almost always a license to occupy, not a lease. A license grants permission to use space without conveying a legal interest in the property, which is what lets memberships be short, flexible, and cancellable on short notice. This distinction holds internationally and is the legal foundation of the whole flexible-workspace proposition — it's why a member can leave in a month while the operator is tied to the building for years. (Whether a given "coworking agreement" is legally a license or a disguised lease has been litigated; operators should get local advice rather than assume.) See the Glossary entries for License and Lease.

Asia-Pacific reads the models differently

APAC is where the model innovation is currently sharpest, and the region deserves specific attention rather than being treated as a footnote to US/EU practice.

• Awfis (India) has arguably industrialised the management-agreement idea into its Managed Aggregation (MA) model, which by FY25 accounted for 67% of its seats and 64% of its centres. The structure: the landlord contributes 80–90% of the fit-out capital, gives Awfis a minimum guarantee of 50% of market rent from roughly months five to thirteen, and after operating costs the profit splits 70% to the landlord and 30% to Awfis. Awfis reached over 134,000 seats across 208 centres, grew FY25 revenue 42% to ₹1,208 crore, and posted a 33.3% operational EBITDA margin — a working demonstration that asset-light can scale profitably. Sources: BW Businessworld, YourStory, The Flex Insights.
• JustCo (Singapore-headquartered) runs a premium serviced-office-plus-coworking model across roughly 51 locations in ten APAC hubs, backed by GIC and Frasers Property, and has pushed upmarket with its luxury "The Collective" brand while expanding into Vietnam and Japan. It illustrates the more capital-intensive, brand-and-location-led end of the spectrum. Sources: JustCo, PR Newswire.
• The Executive Centre and Servcorp anchor the corporate, private-suite-heavy business-centre tradition that predates "coworking" in the region — high service, high price, corporate clientele.
• Common Ground (Malaysia), now part of The Flexi Group (45+ managed locations across 11 cities), and WOTSO (Australia, ASX-listed) show two more variants: regional-champion consolidation, and a suburban/regional-focused operator that deliberately avoids the CBD-arbitrage game. Sources: TechNode, Coworking Insights.

The through-line: outside the US, the management-agreement / managed-aggregation structure isn't a niche — in India it is becoming the dominant way new flex capacity gets built.

The economics: where the money comes from and goes

Strip away the model and every coworking site is the same arithmetic: revenue per sold unit, times units sold (occupancy), minus a cost base dominated by rent and payroll. Here is how each piece behaves.

Revenue streams, ranked by what they actually contribute

The single most important economic fact for a new operator to internalise is that hot desks are not the business — private offices are. Commonly cited revenue-mix data puts private offices at roughly 70%+ of average coworking revenue and open-plan/hot-desk memberships at under 10%. On a per-unit basis one industry model puts a private office at ~US$1,500/month against ~US$250/month for a hot desk — a private office earning roughly six times a hot desk. The implication for space planning is direct: the ratio of private/dedicated space to open-plan space is one of the biggest levers on whether a site makes money. Sources: Optix, Sharpsheets, FinancialModelsLab.

Beyond desks, the ancillary streams are where margin hides:

• Meeting rooms — roughly 10% of revenue on typical mixes; high-margin because the room exists anyway.
• Virtual offices — business address, mail handling, occasional room access with almost no space cost. One of the highest-margin products in the whole model; some spaces draw upwards of 40% of revenue from it.
• Day passes / on-demand — drop-in desks, day offices, event-space hire; a growing slice (~7% of revenue in some data) and a top-of-funnel for full memberships.
• Services & events — printing, F&B, sponsorship, partner commissions.

Well-run spaces now generate roughly 15–25% of revenue from ancillary services (aggressive operators claim 25–45%). Diversifying beyond the desk is increasingly treated as the difference between a fragile and a resilient P&L. Sources: Optix, Allwork, Platform.hk.

What things cost members (and why it varies so much)

List prices are heavily market-dependent — the same product can differ 3–5× between a suburban secondary market and a prime CBD tower, so treat any single number as a starting point, not a rule. Broad 2025 ranges, with the sharpest divergence at the private-office end:

Product	Typical monthly range	Notes
Hot desk	~US$150–600	Singapore ~S$128–750; NYC (the most expensive global market) US$400–600 at premium sites
Dedicated desk	into the several-hundreds, up to ~US$1,200	~US$700/month is typical in Singapore; highly market-dependent
Private office (per desk)	~US$500–2,000+	Premium Singapore CBD suites reach S$1,500–1,950+ per desk
Virtual office	tens of dollars	Lowest cost to serve, highest margin

Dedicated desks in particular should not be thought of as cheap — in a mature Asian market like Singapore they run into the hundreds of dollars per month, around US$700 being typical, and it is a mistake to under-price them by benchmarking against a suburban US number. Sources: Launch Workplaces, DollarsAndSense, Workcentral, CoworkingCafe.

The cost base

For a lease-model operator, two line items dominate and are largely fixed:

• Rent — roughly 30–40% of revenue (prime locations 40%+, secondary markets 30% or less). This is the number the management-agreement and franchise models are designed to restructure.
• Payroll — roughly 25–30% of revenue.

Utilities, cleaning, maintenance, and operations add another ~15–25% combined, and variable costs (payment fees, consumables) run in the high-teens percent. Fixed costs — the lease and core salaried staff — commonly represent 70%+ of total operating cost, which is precisely why occupancy swings hit the bottom line so hard: above breakeven, most incremental revenue drops through to profit; below it, the fixed base still has to be paid. Mature, well-run spaces report EBITDA margins around 10–20% once filled. Sources: Sharpsheets, Business Plan Templates, FinancialModelsLab.

Upfront capital (fit-out)

Before any of that, there's the build-out. Industry estimates cluster around US$60/sq ft as a target blended fit-out cost for a moderate coworking build, versus ~US$280/sq ft for a medium-quality conventional office fit-out in early 2025 — coworking deliberately runs leaner. A representative 5,000 sq ft example: ~US$690,000 total initial CAPEX, of which ~US$300,000 is interior build-out, ~US$150,000 furniture and fixtures, and ~US$80,000 IT backbone. These figures are US-centric; APAC and European fit-out economics differ, so treat them as ballparks and price your own build locally. Sources: FinancialModelsLab, JLL 2025 Fit-Out Guide, Beaufurn.

The two metrics that decide whether you make money

Everything above collapses into two numbers an operator lives and dies by.

Occupancy — the share of sellable space (or seats) currently sold. A useful reading of the benchmarks: 75–85% is a healthy, stable band, and sustained ~90%+ signals a genuinely strong site (though very high occupancy can also mean you've under-provisioned meeting rooms or left pricing power on the table). The global average sat around 68% at the start of 2025 — well below "healthy," which is the honest state of a still-recovering, oversupplied-in-places market. European markets and strong Indian metros were running 78–85%.

A deliberate caution on breakeven occupancy: there is no universal breakeven or profitability occupancy threshold, and you should distrust anyone who quotes one as gospel. The occupancy at which a specific site covers its costs is driven mostly by its rent-to-revenue ratio, its price points, and its cost base — all of which vary enormously by market and by deal structure. A management-agreement site with no fixed rent breaks even at a completely different occupancy than a prime-CBD lease-arbitrage site. Commonly cited figures land in the 75–80% region, but that is a market-specific artefact, not a law — model your site's breakeven from your numbers. Sources: OfficeRnD, Optix, Speakwise.

Churn — the rate at which members leave. Because acquiring a member costs far more than keeping one, churn quietly determines whether occupancy compounds or leaks. Deskmag's long-standing benchmark is that spaces tend to be profitable around a ~5% monthly churn rate; retention above ~90% is considered good. The interaction with occupancy is the subtle part: a site at 90% occupancy with 40% annual churn is re-selling nearly half its base every year and living on a treadmill, while a site at 75% occupancy with 15% churn is compounding value from a stable membership. Occupancy without retention is a vanity metric. Sources: Deskmag / Coworking Statistics, Optix.

The market you're entering

For context on the opportunity: the global coworking market is valued at roughly US$21 billion, with about 42,000 spaces worldwide expected to reach ~44,000 in 2026, and an estimated 5.5–6 million members — a mainstream, globally adopted model rather than a niche. Around 54–58% of operators reported profitability in early 2025 with ~18% at a loss, and roughly 75% of spaces reach profitability after about two years. New-space openings are densest in Asia-Pacific. Sources: Allwork, Deskmag, Optix.

What to take from this page

If you remember five things: coworking is a real-estate arbitrage with a services layer, not a hospitality business that happens to rent desks. The deal structure you sign with the landlord — lease, management agreement, franchise, or self-operate — decides who carries the cost-revenue mismatch, and outside the US the management-agreement model is increasingly the default. Private offices and virtual offices, not hot desks, are where the revenue is. Rent and payroll will eat two-thirds or more of your revenue, so the whole game is filling the space and keeping it full. And there is no magic occupancy number that guarantees profit — model your own breakeven, then obsess over churn, because retention is what turns occupancy into a business.

---

Related: Glossary for every term above · Coworking Operating Systems (CAT-11) for the platforms that run the economics · Billing & Payments (CAT-13) · Marketing & Listing Platforms (CAT-16) and Broker & Space Aggregation (CAT-24) for demand. Forthcoming Start Here guides go deeper on feasibility, property, fit-out, and pre-launch.