Cost per unit can make a weak project look healthy

Cost per unit is attractive because it reduces a complex development into a number that can be compared quickly. It is useful for orientation, but dangerous as a governing metric. The ratio says nothing about how much building, land, parking, infrastructure, or time is required to create each unit. It can therefore make a spatially inefficient or operationally weak project appear competitive.

The basic problem is denominator compression. A “unit” is treated as though every unit has the same size, value, construction burden, and share of common infrastructure. In reality, a development may contain studios, family units, accessible units, townhouses, penthouses, affordable units, and service or amenity obligations that carry very different economics. A project can improve its reported cost per unit simply by increasing unit count, even when the resulting product becomes smaller, harder to operate, or less valuable.

The ratio hides the amount of building behind each unit

Two projects with the same unit count can require very different gross floor areas. Corridor length, core size, lobby area, mechanical rooms, amenity, storage, loading, refuse, back-of-house functions, and parking all contribute to the amount of construction that supports the residential program. When gross area per unit rises, cost pressure often follows even if the headline cost per unit remains within a familiar range.

That is why cost per unit should always be read beside cost per gross square foot, cost per rentable or sellable square foot, and net-to-gross efficiency. Those metrics expose whether the unit count is being supported efficiently or buried inside excessive nonrevenue area.

Unit mix can manipulate the comparison

A project dominated by smaller units will generally report more units within the same building area. That can lower cost per unit while increasing cost per rentable square foot and creating a product with different turnover, management, amenity, and market characteristics. A family-oriented project may show a higher cost per unit because each unit is larger, yet it may produce stronger retention or serve a different market need.

The correct comparison must normalize for unit size and mix. Average unit area, bedroom distribution, target household, and revenue per square foot should be visible. Otherwise, cost per unit rewards density of count rather than quality of product or efficiency of capital.

Parking can dominate without appearing in the denominator

Structured and underground parking can add substantial area, excavation, structure, ventilation, fire protection, waterproofing, and circulation. Yet every stall is typically absorbed into the project cost and then divided by the residential unit count. A development with an expensive parking solution may appear to have high residential cost even though much of the burden is generated by transportation policy, site constraints, or market assumptions.

The model should separate residential building cost, parking cost, and site infrastructure. It should also report spaces per unit, area per stall, and cost per stall. This makes it possible to determine whether the residential product is inefficient or whether the parking strategy is the true cost driver.

Amenities and service standards change the product

Common amenities, staffed lobbies, package rooms, fitness areas, coworking space, roof terraces, pools, concierge functions, and extensive landscape can increase cost per unit. They may also support pricing, leasing velocity, retention, or brand position. Treating these features as pure inefficiency ignores their revenue function. Treating them as automatically value-creating ignores operating expense and market saturation.

A decision-ready model should identify which shared spaces are required, which are market differentiators, and which are oversized or weakly connected to value. Amenity area per unit and operating burden should be visible so that the project can distinguish strategic investment from decorative scope.

Site and infrastructure costs distort cross-project comparisons

A constrained urban infill site, a steep site, a brownfield parcel, and a large suburban tract may produce similar unit counts but require different foundations, retaining systems, utility extensions, stormwater work, access improvements, environmental remediation, and construction logistics. Cost per unit blends those conditions into the same ratio and can make the building design appear responsible for costs generated by the site.

For portfolio comparison, the numerator should be disaggregated into building, parking, site, off-site infrastructure, land, soft cost, and financing layers. The combined ratio remains useful, but the decomposition explains why it differs.

Schedule and phasing are also hidden

Cost per unit rarely reveals how long the project takes to entitle, design, finance, build, and stabilize. A project with a lower nominal construction cost can require more financing carry, longer general conditions, more escalation, or a slower lease-up. Phased projects may also allocate shared infrastructure inconsistently, making early phases appear expensive or later phases artificially cheap.

The ratio should therefore be tied to a date and delivery condition. Is it based on current dollars, midpoint-of-construction dollars, completed units, occupied units, or stabilized units? Without that context, comparisons across time and phase can be materially misleading.

Low cost per unit can signal a weak product

Reducing unit size, compressing storage, narrowing circulation, minimizing acoustic separation, simplifying facade, reducing durability, or cutting amenity can lower cost per unit. Those actions may be appropriate, but the metric alone cannot distinguish disciplined simplification from value destruction. The project may meet the cost target while creating a product that underperforms in rent, sales, operations, maintenance, or resident experience.

Cost per unit should therefore be paired with revenue per unit, revenue per rentable square foot, operating expense per unit, replacement reserve, and expected absorption. Capital efficiency is meaningful only in relation to the value and performance produced.

Use a metric set, not a favorite ratio

A robust early screen should report total development cost, cost per unit, cost per gross square foot, cost per rentable or sellable square foot, gross area per unit, net-to-gross efficiency, parking burden, site cost, soft cost, and schedule exposure. It should also show how those metrics change under alternative unit mixes and planning schemes.

No single ratio should be allowed to declare the project healthy. The purpose of metrics is not to simplify the decision until complexity disappears. It is to reveal where the project creates or consumes value. Cost per unit becomes useful when it is one lens in a coordinated diagnostic system.