The Direct Answer

The "flex versus core" debate is usually sold as a convenience choice. For the CFO it is a capital-versus-operating decision with a definable break-even point.

A traditional lease offers a lower monthly rent but demands a substantial upfront commitment to fit-out and, at the far end of the term, reinstatement. Enterprise co-working offers all-in pricing and near-zero capital outlay, but at a per-seat premium. Neither is inherently cheaper. The trade is capital-and-commitment against flexibility-at-a-premium, and the P&L feels those two things very differently across five years.

The real question is therefore not which model feels more flexible. It is at what headcount and what term the flexibility of a serviced office stops being a benefit and starts being a liability — a threshold that is specific to your growth plan and can be calculated rather than assumed.

Two Cost Structures, Side by Side

A traditional lease front-loads and back-loads cost around a low middle. The rent is the visible number, but the model also carries fit-out capital expenditure, a security deposit, ongoing management and service charges, and a reinstatement obligation that lands at expiry — all across a three-to-five-year commitment.

Enterprise co-working collapses nearly all of that into a single all-in operating cost per desk. There is minimal capital outlay, term flexibility, and speed to occupy. The premium per seat is the price of that optionality.

The distinction matters because the two structures fail in opposite directions. A traditional lease punishes you for taking space you do not grow into. Enterprise co-working punishes you for staying longer than you expected.

The Break-Even Framework: Where the Lines Cross

Seven variables move the answer:

1. Headcount — and how confidently you can forecast it
2. Term — the number of years over which capital is amortised
3. Fit-out cost per square foot — driven by specification and building grade
4. The all-in serviced rate per desk — the flex comparator
5. Reinstatement liability — the cost of handing space back
6. Space efficiency — square feet per head under your workplace strategy
7. Certainty of the growth plan — the variable that determines how much optionality is actually worth

The shape of the result tends to hold across a wide range of inputs. Below a certain combination of headcount and term, flex wins on total cost, because a traditional lease's capital and reinstatement burden cannot be amortised across enough desks or enough years to bring the effective cost per desk down. Above that threshold, the lease's lower run-rate overtakes the flex premium despite the upfront hit, and the gap continues to widen for the remainder of the term.

That last qualification matters. The lease advantage is monotonic within a term, not indefinitely: at expiry the fixed-cost base resets, through renewal at market rent, a further fit-out, or reinstatement. Any model spanning more years than the lease itself needs to price that second transaction.

Two levers move the crossover point more than any others, and they work differently.

Term sets the amortisation period. The same fit-out spread over five years rather than three is materially cheaper per desk per month, and reinstatement — a fixed sum falling due once — is diluted the same way.

Headcount works through scale rather than through area. Floor area grows roughly in line with headcount, so area-driven capital cost per head does not fall simply because the requirement is larger. The economies come from elsewhere: fixed professional fees and transaction costs spread across more desks, greater procurement leverage when the fit-out goes to tender, stronger incentive packages on larger lettings, and better core-to-usable efficiency on whole floors than on part floors. Separately, space efficiency — how many square feet each desk actually consumes — is the lever that converts a psf rent into a per-desk cost, and small changes to it move the answer more than most occupiers expect.

How to Build the Comparison

Model both scenarios over the same period, on a total cost of occupancy basis, and reduce each to a single effective cost per desk per month.

Traditional lease: (rent + service charge over the term, escalated for step-ups and reviews) + fit-out capital + professional fees + IT and furniture + reinstatement provision + occupier-retained operating costs, divided by (desks × months). Offset any rent-free period or landlord fit-out contribution.

That last input is the one most often left out. A lease rate does not include utilities, cleaning, reception, office management headcount, broadband and IT support, insurance, furniture replacement or consumables. All of them sit inside a serviced desk rate, so omitting them from the lease side is not a conservative simplification — it is a direct thumb on the scale.

Enterprise co-working: (all-in desk rate × desks × months) + set-up, customisation or exit charges, divided by (desks × months). In most cases this reduces to the rate itself, which is precisely why the model's real work sits on the lease side.

Three refinements separate a defensible model from an indicative one:

• Discount the cash flows. A lease front-loads capital and back-loads reinstatement; flex pays level. Summing undiscounted cash and dividing by desk-months treats a day-one fit-out dollar as identical to a month-sixty rent dollar, which flatters the lease. Run both at your cost of capital, and carry the opportunity cost of both deposits, not just the larger one.
• Escalate the rent. Savills has forecast CBD Grade A rental growth of 5% in 2026 and 5%–7% in 2027.² A five-year model with a flat rent line is inconsistent with the market it is modelling. Flex rates reset at renewal too.
• State your desks-per-head and square-feet-per-desk assumptions explicitly. Under hybrid working a lease is typically sized for peak occupancy while flex is bought per desk, so comparing the two without a stated sharing ratio compares different quantities.

Then run the traditional lease case again at your realistic downside headcount. This step is frequently skipped, and it changes the answer more often than any other, because a lease sized to an optimistic growth plan and occupied by a smaller team is among the most expensive outcomes available in this comparison.

A note on cost inputs. Indicative regional benchmarks by specification tier are set out in the Savills Projects Office Fit Out Guide 2025, which covers Asia Pacific and quotes in US dollars.¹ Because Singapore fit-out and reinstatement rates vary widely by building grade, scope and specification — and because construction costs have continued to escalate since that edition — the figures used in a live comparison should be priced by the Savills Projects team against your actual specification rather than taken from any published range. Our guide to decoding the "black box" of construction costs explains what drives the variance.

The rental side of the model can be anchored to current market evidence. Savills' Q2 2026 research recorded CBD Grade A rents at a record S$10.42 per sq ft per month, with CBD Grade A vacancy down to 5.6% and Grade AAA vacancy at 3.1%, the lowest in thirteen years.² When quality space is this tightly held, the practical constraint is availability rather than negotiating position — which is a reason to begin the analysis early, while options still exist.

When Flexibility Is an Asset — and When It Becomes a Liability

Flexibility is genuinely valuable where the future is uncertain. For early-stage businesses, teams with volatile or unproven growth trajectories, short-horizon projects, or any situation demanding occupation in weeks rather than months, the per-seat premium buys a real option — the ability to expand, contract or exit without a capital write-off. Market entrants testing Singapore before committing are paying for exactly the right thing.

Flexibility becomes a liability under the opposite conditions. For a stable, larger team with a multi-year horizon, the premium stops purchasing optionality that will ever be exercised and simply accumulates. Depending on the size of the premium and the floorplate, the five-year cumulative difference can approach the cost of fitting out a leased space outright — and at the end of it the occupier holds no fitted asset and no workplace of its own design.

There is also a strategic dimension that rarely reaches the spreadsheet. A serviced floor cannot easily express a brand, and workplace quality now sits inside the employee value proposition, which is changing how occupiers approach leasing decisions.

The Core-Plus-Flex Hybrid

For many mid-sized occupiers the right answer is neither pure model. A "core-plus-flex" structure commits a traditional lease to the stable, predictable portion of headcount — the base the business will carry through any plausible scenario — and uses serviced space to absorb growth, project teams and volatility above that line.

This converts the decision from a binary into a sizing question: how much of your five-year headcount is genuinely certain? That portion belongs on a lease, where the economics reward commitment. The uncertain remainder belongs in flex, where you are paying for the right to be wrong.

The Hidden Line Items CFOs Miss

A desk-rate-versus-psf comparison omits most of what determines the outcome:

• Reinstatement and dilapidations. A contractual obligation crystallising at expiry, and one that should be provisioned monthly across the term rather than met as a surprise. The scope is defined at the point of signing, so that is when to negotiate it — though in a market at 5.6% vacancy, with landlords holding firm on terms, occupiers should be realistic about how much can be bargained away.²
• Rent-free versus fit-out contribution. Landlord incentives can be taken as free months or as capital towards the fit-out, and the two have different cash-flow and accounting profiles. Our guide to decoding landlord incentives sets out how to value them.
• Capital expenditure depreciation. Leasehold improvements are depreciated over the shorter of their useful life and the lease term, affecting reported earnings differently from an equivalent operating cost.
• Tax treatment. Singapore's Section 14N deduction for renovation and refurbishment, and capital allowances on qualifying items, change the after-tax cost of a fit-out materially. A pre-tax comparison can point one way and an after-tax comparison the other.
• The predictability premium. An all-in operating cost is easier to forecast and defend to a board. Many finance teams regard that as worth paying for, even when it is not the lowest-cost option.
• Lease accounting — and a common misconception. Under IFRS 16 — applied in Singapore as SFRS(I) 16, issued by the Accounting Standards Committee under ACRA — leases are recognised on balance sheet as a right-of-use asset and a corresponding lease liability.³ The standard applies to leases of more than twelve months unless the underlying asset is low-value, and it is the short-term exemption for terms of twelve months or less that keeps most flexible arrangements off balance sheet. It does not follow that co-working is automatically off balance sheet: an enterprise agreement granting exclusive use of an identified suite for more than twelve months may well meet the definition of a lease and be recognised accordingly. If the balance-sheet outcome is part of the rationale, confirm the treatment with your auditor against the specific contract before relying on it.
• The option value of exit. The ability to leave early has a quantifiable worth. It is only worth paying for if there is a realistic scenario in which you would use it.
• Fitted space as a middle path. Taking a lease on already-fitted premises removes much of the capital outlay while retaining a traditional structure — a route we cover in beating rising renovation costs.

Strategic Recommendations (Risk-Adjusted)

How to Model Your Decision

The defensible basis for a five-year occupancy decision is not a view about flexibility. It is a total cost of occupancy model that runs both structures against your actual headcount plan, prices the capital and exit lines properly, and identifies the point at which the answer flips.

That break-even is specific to your business, and it moves with term, headcount, specification and the market. It is also the number that makes the recommendation defensible to a board — whichever way it points.

Speak to Our Office Leasing Team

The Savills Office Leasing team can build that model against your five-year headcount plan, working with our Project Management colleagues to price the fit-out and reinstatement lines from live cost data rather than published ranges. We advise across both structures, including flexible and co-working workspace, so the recommendation follows the modelling rather than the mandate.

For occupiers approaching a lease event, our commercial leasing timeline sets out when this analysis needs to begin.

Footnotes

¹ Source: Savills Projects, "Office Fit Out Guide 2025" (Asia Pacific edition; costs quoted in US dollars, excluding local taxes and levies). Direct PDF download, approximately 10 MB. https://pdf.savills.asia/fit-out-guide/savills-projects-office-fitout-v5-2.pdf

² Source: Savills Research, "Demand-Led Rental Inflation Overrides Global Economic Concerns: CBD Office Vacancy Falls to 5.6%" (Q2 2026 Office Report), 30 July 2026. https://www.savills.com.sg/insight-and-opinion/savills-news/237993/q2-2026-office-report

³ Source: IFRS Foundation, "IFRS 16 Leases". The Singapore equivalent, SFRS(I) 16, is issued by the Accounting Standards Committee (ASC). https://www.ifrs.org/issued-standards/list-of-standards/ifrs-16-leases/