Co-working
A spread business dressed as real estate: take space on a long lease, fit it out, and sell it by the seat at a markup. The spread is thin, the capex is real, and the whole model turns on occupancy holding above the rent line.
Realisation per seat minus rent and amortised fit-out is the entire economics. Above roughly 48% rent-to-revenue on a fitted-out property, the maths stops working — so occupancy, lock-in and centre size decide everything.
Four parties, five formats, one fragmented market.
Who is in the deal
- LandlordOwns the asset. Under profit-share structures, also funds the fit-out.
- OperatorTakes the space, fits it out, sells seats. The equity in the spread.
- ClientFrom single desks to whole-floor anchors.
- Brokers / IPCsCBRE, JLL, Cushman. A channel with a cost attached.
Formats, cheapest to richest
- Virtual officeAddress only, often for GST registration. Negligible capex.
- Co-workingRetrofit, furniture, fittings, amenities. The classic shared format.
- Flexible workspaceShort-tenure, variable seat count.
- Built to suitA section dedicated to one company.
- Managed officeEntire space taken by one client — from the operator's standpoint, the cream: fully occupied, better tenure.
~400 operators
Top 10 control roughly 60% of seats. Fragmented below that.
1,250 vs 400+ centres
Tier 1 versus Tier 2. Bangalore and Pune are the two largest markets.
~40% IT & ITES
Followed by services and finance. GCCs, MNCs, large corporates and SMEs.
The spread, and the thresholds that protect it.
Around 90% of the industry runs on a straight lease. The operator signs a long lease, funds the fit-out, and carries the occupancy risk. The alternatives are asset-light: a managed-aggregation or profit-share structure where the landlord funds capex against a minimum guarantee — typically three to six months of rent or deposit — or an owned/fractional model.
Space is taken in three conditions. Bare shell is cheapest and needs the most capex; warm shell sits in between; fitted-out costs more in rent but almost nothing upfront. The trade is capex today against rent forever, and the rent-free period negotiated at acquisition is a real part of the return.
The two hard thresholds. A centre below roughly 6,000 sq ft is not operationally sustainable. On a fitted-out property, once rent exceeds about 48% of revenue it is difficult to make money. These are the fastest screens to apply to any new centre in the pipeline.
Industry standards
- Area per seat30–35 sq ft carpet
- Capex, bare shell~₹2,250 per sq ft carpet — about ₹2.5–3 crore for a 10,000 sq ft centre
- Average occupancy~72%, with only 10–15% floating seats. 100% is never reached
- Client tenure (ALOS)12–18 months for ~55% of clients; ~25% stay three years
- Stabilised occupancyRoughly 12 months post-launch
Capex composition
- AC — 25%
- Furniture & chairs — 20%
- Electrical — 15%
- Partitions & glass — 15%
- Flooring & carpet — 10%
- IT — 5%
- Painting — 5%
- Miscellaneous — 5%
Occupancy, tenure, and capital structure.
Occupancy above the rent line
At ~72% average occupancy and a 48% rent-to-revenue ceiling, there is little margin for error. Sustained occupancy in mature centres is the whole business.
Managed & anchor tenure
Managed offices are fully occupied with better tenures. Large clients — those above 100 or 300 seats — bring longer lock-ins and lower churn than the 12–18 month co-working norm.
Asset-light structure
Profit-share and managed-aggregation models push capex to the landlord and diversify risk. The payback period must sit inside the lease lock-in, or the operator is exposed.
What to answer before underwriting.
- →Model split. Straight lease versus managed aggregation versus owned. What share of centres sits in each, and where is the capex burden?
- →Does payback fit the lock-in? The single most important structural question. If payback runs past the lease lock-in, the operator carries uncompensated risk.
- →Occupancy by cohort. Mature versus recently launched centres, and across Tier 1 and Tier 2. A blended number hides the ramp.
- →Realisation per seat by micro-market. How much of seat growth is coming from lower-rent micro-markets, and what does that do to blended realisation?
- →Client concentration and lock-in. Top-five and largest-client share, weighted average lock-in, and fall-off rates at 12, 18 and 36 months.
- →Multi-centre clients. What share of occupied area comes from clients taking seats in more than one centre? A stickiness indicator.
- →Signed but not operational. LOIs and under-construction seats against the current base, and the construction-to-launch timeline.
- →Channel mix. Seats sold direct versus through IPCs and brokers, and what alternate channels are being built.
- →Landlord terms. Rent-free period post-acquisition, minimum guarantee, and where landlords push back on funding capex.
- →Minimum centre size. What threshold will the company not operate below, given the ~6,000 sq ft sustainability floor?
What to monitor, quarter by quarter.
| KPI | Calculation / source | Benchmark or read-through |
|---|---|---|
| Occupancy % | Occupied seats ÷ operational seats | ~72% industry average; split mature vs ramping centres |
| Realisation per seat | Revenue ÷ occupied seats ÷ month | Falls as the mix shifts to Tier 2 — check against seat growth |
| Rent ÷ revenue | Lease cost ÷ centre revenue | Above ~48% on fitted-out property, profitability is very difficult |
| Capex per sq ft | Fit-out cost ÷ carpet area | ~₹2,250 bare shell; overruns extend payback directly |
| Payback vs lock-in | Months to recover capex vs lease lock-in | Payback must sit inside the lock-in — the core structural test |
| Operational seats & pipeline | Live seats, signed, under construction | Signed-but-not-live is future revenue and future capex |
| Time to stabilised occupancy | Months from launch | ~12 months; slippage signals weak demand or poor location |
| Weighted average lock-in | Contract schedule | Longer lock-ins de-risk the spread; managed offices score best |
| Churn / fall-off rate | At 12M, 18M, 36M | Against the ALOS norm — ~55% at 12–18M, ~25% to three years |
| Client concentration | Top 5 and largest client share | Roughly 9% top-five and 3–4% largest is a healthy spread |
| Large-client cohort | Share of base from 100+ and 300+ seat clients | Higher share = longer tenure and lower churn |
| Direct vs broker mix | Seats sold by channel | Broker dependence is a permanent cost on the spread |
| Centre-level EBITDA margin | By centre, and blended | Blended margin hides loss-making centres — always ask for the split |
| Ancillary revenue share | Parking, F&B, VAS ÷ total | Higher-margin and less rent-sensitive than seat revenue |
How the thesis breaks.
- !Duration mismatch. Long leases against 12–18 month client tenures is the structural fragility of the model. A demand shock leaves the rent obligation intact.
- !Occupancy slipping below the rent line. With a 48% rent-to-revenue ceiling and ~72% typical occupancy, a few points of occupancy decides profit or loss.
- !Growth that dilutes realisation. Adding seats in low-rent Tier 2 micro-markets grows the headline seat count while lowering revenue per seat and possibly margin.
- !Sub-scale centres. Anything below ~6,000 sq ft is structurally unviable; a pipeline full of small centres is a warning.
- !Capex ahead of demand. Fit-out is sunk on signing. Speculative expansion into unproven micro-markets is where the capital goes.
- !Anchor client loss. A managed-office or large anchor departure leaves a large block of empty, fully-rented space at once.
- !Sector concentration. With ~40% of demand from IT and ITES, a tech hiring downturn hits occupancy across the portfolio simultaneously.
The figures, and where they stand.
| Metric | Value | Note | Basis |
|---|---|---|---|
| Operators in India | 400+ | Top 10 control ~60% of seats | Research note |
| Lease-model share | ~90% | Straight lease dominates the industry | Research note |
| Centres — Tier 1 vs Tier 2 | 1,250 vs 400+ | Bangalore and Pune the two largest markets | Research note |
| Client mix — IT & ITES | ~40% | Then services and finance | Research note |
| Area per seat | 30–35 sq ft | Carpet area | Industry standard |
| Capex, bare shell | ~₹2,250 / sq ft | ≈ ₹2.5–3 crore for 10,000 sq ft | Industry standard |
| Minimum viable centre | ~6,000 sq ft | Below this, not operationally sustainable | Industry standard |
| Rent-to-revenue ceiling | ~48% | Fitted-out property; above this profitability is very hard | Industry standard |
| Average occupancy | ~72% | Only 10–15% floating seats | Research note |
| Client tenure (ALOS) | 12–18M (~55%) | ~25% stay three years | Research note |
| Time to stabilised occupancy | ~12 months | Post-launch | Est. |
| Minimum guarantee to landlord | 3–6 months | Space or deposit, under profit-share structures | Research note |
| Healthy client concentration | Top 5 ~9% | Largest client 3–4% | Research note |