Methodology and data sources
Coefficients current as of 2026-Q4
How the model works
The calculation runs in seven steps, in this order. Each step feeds the next, so a change to any input propagates all the way through to the final total and the return figures.
- Unit mix to rentable area. Unit counts are allocated from the requested mix using largest-remainder rounding, so the counts always sum to the exact number of units requested. Rentable area is the sum of individual unit areas — never an average times a count.
- Rentable area to gross area. Net rentable area is divided by an efficiency factor (84%–86%% for single-storey, 70%–72%% for multi-storey) to give gross building area.
- Hard cost. Gross area multiplied by a published cost per gross square foot for the building type, then multiplied by a state construction index and a structural system multiplier.
- Soft costs. Hard cost multiplied by 25%, 32% or 40% for the low, mid and high cases.
- Land. Building footprint divided by buildable area per acre gives a default lot size; multiplied by the market tier's price per acre gives land cost. Footprint, not gross floor area, is used — which is why a two-storey building needs about half the land.
- Revenue. Rentable area multiplied by the achieved rent per rentable square foot — asking rate less the move-in discount — then by physical occupancy. Operating expenses are subtracted to give net operating income.
- Returns. Yield on cost is NOI divided by total project cost; payback is the inverse. Stabilised value capitalises the same NOI at a market cap rate, and the development spread is yield on cost minus that cap rate — the test a developer actually underwrites to.
Hard cost anchors
Three building types are anchored to published cost ranges. The intermediate climate-control settings are interpolated between those anchors and are labelled as derived wherever they appear.
| Building type | Low | Typical | High | Basis |
|---|---|---|---|---|
| Single-storey drive-up | $55.00 | $70.00 | $85.00 | Published anchor |
| Single-storey climate-controlled | $80.00 | $95.00 | $120.00 | Published anchor |
| Multi-storey climate-controlled | $105.00 | $137.00 | $170.00 | Published anchor |
Dollars per gross square foot. Hard cost only — excludes soft costs and land.
Soft costs and the trade split
Soft costs are modelled as a percentage of hard cost, at 25%, 32% and 40% for the low, mid and high cases. This covers site development overage, architecture and engineering, permitting, legal, security design, financing carry and contingency.
Hard cost splits across trades with a fixed shape. The HVAC share moves with the climate setting, and the residual is absorbed by "other" so the shares always sum to exactly 100%.
| Trade | Share of hard cost |
|---|---|
| Steel structure & building shell | 27.0% |
| Concrete, foundations & slabs | 14.0% |
| HVAC equipment, electrical switchgear & asphalt paving | 16.0% |
| Doors, partitions, fire protection, plumbing, security & general conditions | 43.0% |
Fixed facility area
Office, retail and staff area sits outside the rentable unit area but inside the construction cost. It scales sublinearly with facility size — 400–800 sq ft, scaled at 3 sq ft per unit — capped at the top of the range because beyond that the area is a design choice rather than a requirement. A 200-unit facility carries 600 sq ft. This is a stated assumption, not a surveyed figure.
It matters because it is the only cost in the model that does not scale linearly with unit count, and therefore the only source of economy of scale in the results: the office is a much larger share of a 50-unit project than of a 500-unit one.
Revenue assumptions
| Climate setting | Asking, low | Asking, typical | Asking, high | Achieved, typical |
|---|---|---|---|---|
| None (drive-up) | $1.15 | $1.29 | $1.40 | $1.10 |
| Partial | $1.28 | $1.41 | $1.50 | $1.20 |
| Full climate control | $1.40 | $1.53 | $1.65 | $1.30 |
Dollars per rentable square foot per month, before the state rent index is applied. Asking is the advertised street rate; achieved is asking less 15%, which is what a facility actually collects once move-in discounts are applied. Revenue is modelled on the achieved column at 86% physical occupancy. Using the asking rate against occupancy — the common shortcut — overstates income by about a fifth, which is why this site does not do it.
State construction index
Each state carries a construction cost index relative to the national average of 1.00, and a land market tier that selects the default land price band. Both are adjusted independently: a state can be expensive to build in and cheap for land, or the reverse.
The index is not a substitute for a local estimate. Construction costs within a state vary more between urban and rural submarkets than the state average suggests. Use it to compare markets, not to price a specific parcel.
| State | Construction index | Land tier | Rent index |
|---|---|---|---|
| Alabama | 0.84 | 1 | 0.88 |
| Alaska | 1.35 | 2 | 1.20 |
| Arizona | 0.94 | 3 | 1.02 |
| Arkansas | 0.83 | 1 | 0.85 |
| California | 1.26 | 4 | 1.42 |
| Colorado | 0.98 | 3 | 1.10 |
| Connecticut | 1.14 | 3 | 1.12 |
| Delaware | 1.00 | 3 | 1.00 |
| District of Columbia | 1.16 | 4 | 1.35 |
| Florida | 0.91 | 3 | 1.04 |
| Georgia | 0.88 | 2 | 0.94 |
| Hawaii | 1.45 | 4 | 1.38 |
| Idaho | 0.92 | 2 | 0.92 |
| Illinois | 1.02 | 3 | 1.02 |
| Indiana | 0.90 | 2 | 0.90 |
| Iowa | 0.90 | 1 | 0.86 |
| Kansas | 0.88 | 1 | 0.86 |
| Kentucky | 0.86 | 1 | 0.87 |
| Louisiana | 0.89 | 2 | 0.90 |
| Maine | 1.03 | 2 | 0.95 |
| Maryland | 1.02 | 3 | 1.08 |
| Massachusetts | 1.22 | 4 | 1.32 |
| Michigan | 0.93 | 2 | 0.90 |
| Minnesota | 0.96 | 2 | 0.96 |
| Mississippi | 0.82 | 1 | 0.82 |
| Missouri | 0.88 | 1 | 0.87 |
| Montana | 0.94 | 1 | 0.90 |
| Nebraska | 0.89 | 1 | 0.86 |
| Nevada | 0.97 | 3 | 1.00 |
| New Hampshire | 1.05 | 2 | 1.02 |
| New Jersey | 1.12 | 4 | 1.18 |
| New Mexico | 0.90 | 2 | 0.88 |
| New York | 1.28 | 4 | 1.40 |
| North Carolina | 0.88 | 2 | 0.93 |
| North Dakota | 0.92 | 1 | 0.90 |
| Ohio | 0.91 | 2 | 0.89 |
| Oklahoma | 0.87 | 1 | 0.85 |
| Oregon | 1.10 | 3 | 1.08 |
| Pennsylvania | 1.00 | 3 | 0.98 |
| Rhode Island | 1.12 | 3 | 1.10 |
| South Carolina | 0.87 | 2 | 0.90 |
| South Dakota | 0.90 | 1 | 0.85 |
| Tennessee | 0.88 | 2 | 0.92 |
| Texas | 0.92 | 2 | 0.93 |
| Utah | 0.93 | 2 | 0.97 |
| Vermont | 1.06 | 2 | 0.98 |
| Virginia | 0.93 | 3 | 1.00 |
| Washington | 1.14 | 3 | 1.16 |
| West Virginia | 0.85 | 1 | 0.82 |
| Wisconsin | 0.95 | 2 | 0.92 |
| Wyoming | 0.93 | 1 | 0.90 |
Unit catalogue
Rentable area per unit comes from the unit's nominal dimensions. Gross area adds the unit's share of circulation, derived from the efficiency factor for the building type.
| Unit size | Rentable sq ft |
|---|---|
| 5 × 5 | 25 |
| 5 × 10 | 50 |
| 10 × 10 | 100 |
| 10 × 20 | 200 |
| 10 × 30 | 300 |
Assumptions with no published source
Two kinds of number sit below. Some have a published range, and the model picks a point value inside it — the range is cited, the choice is ours. Others have no published source at the granularity a pro forma needs; those are stated outright rather than attached to a document that does not contain them. Either way they are inputs to challenge, not findings.
| Assumption | Value used | Basis and why it is an assumption |
|---|---|---|
| Operating expense ratio | 35% / 38% / 45% of EGI | Published range for independent operators is 35–45% of EGI (UnionMetric feasibility study; SEA CON reports NOI margins of 55–65%, the same thing inverted). REIT same-store portfolios reach the high 20s, but that is a scale effect a new single-asset project does not get, so the band starts at 35%. |
| Capitalisation rate | 5.5% / 6.5% / 7.5% | Marcus & Millichap reports core-market climate-controlled at 5.75–6.50% and tertiary drive-up at 6.50–7.50%; TractIQ's listed-offer median was 6.79% in October 2026. The midpoint is set between them. Transaction cap rates are reported after the fact and vary by asset quality, market and debt terms, so a single national figure is a choice rather than a measurement. |
| Structural system multipliers | 1-storey 1.00 / 1.02 / 1.04 · 2-storey 1.00 / 1.06 / 1.12 | Derived from the spread between published steel and masonry delivered cost. Note the multiplier is a national average: in some markets masonry is competitive with steel for single-storey drive-up, so a masonry project there would be overpriced by this model. |
| State construction index | 0.82 – 1.45 | Constructed to match published regional premiums (Northeast and coastal California at the top, the rural South at the bottom). It is not an official index such as RSMeans' city cost index, and within-state variation is larger than the state average suggests. Use it to compare markets, never to price a parcel. |
| State rent index | 0.82 – 1.42 | Constructed from the shape of published metro rent differences, applied to the national Street rate. No source publishes a complete state-level self storage rent index. |
| Land price per tier | $100K – $1.80M per acre | Only two anchor points are published: suburban $50k–500k/acre and urban above $1M/acre. Tiers 2 and 4 rest on those anchors; tiers 1 and 3 are derived between and beyond them. No national survey of self storage land prices by market tier exists. |
| Office and retail area | 600 sq ft at 200 units | Midpoint of the 400–1,200 sq ft range in design guidance. It is the only cost that does not scale with unit count, so it is the only reason cost per unit falls as a facility grows. |
| Site coverage and land per acre | 42.5% coverage · 18,500 sq ft footprint/acre | Footprint per acre is 43,560 × coverage. The published figures are rentable square feet per acre and disagree by about 30%: MakoRabco/Trachte report 17,500–19,602, Inside Self-Storage reports 13,000–15,000. At 84% efficiency this model implies roughly 15,500 rentable sq ft per acre — inside the spread and on the conservative side, because more land means more cost. |
| Interpolated cost rows | 0 of 3 building types | Partial climate control and non-climate multi-storey are linear midpoints between the three published anchors. They are labelled "derived" wherever they appear. |
Sources and their verification status
Every coefficient group carries a source reference. A reference is marked verified only when the figure has been traced to a specific published document. Unverified groups are presented as industry benchmark ranges rather than attributed to a named report, because attributing a number to a document that does not contain it is worse than citing nothing.
| Coefficient group | Publisher | Supports | Published |
|---|---|---|---|
| Average cost to build a self storage facility | Terrapin Construction Group | Hard cost $55–85 / $80–120 / $105–170 per gross sq ft; climate premium $20–40/gsf; soft costs +25–40%; single-storey efficiency 80–88%, multi-storey 65–75% | May 2026 |
| Self storage construction cost guide 2026 | SEA CON LLC | Independently reports the same $55–85 / $80–120 / $105–170 per gross sq ft bands and the 25–40% all-in premium | Aug 2026 |
| 2025 Contractor Survey (n = 12 general contractors), p.7 | DXD Capital | Hard cost composition: steel 27%, concrete 14%, all other 43%, and a combined 16% for asphalt, electrical switchgear and HVAC equipment | Jan 2025 |
| Self storage construction cost dataset (100+ projects, 18M sq ft) | TractIQ / ARCO-Murray | Median hard cost $84.23/gsf for class A projects ≤ 3 storeys; net rentable efficiency 75.0% area-weighted; site work 17.1% of hard cost in 2025 (10.0% in 2021) | Sep 2026 |
| Self storage market data | TractIQ | Street rate $1.41/sq ft/month; 10×10 non-climate $1.29, climate-controlled $1.53; web rate $1.17; physical occupancy 85.7% TTM | Sep 2026 |
| Self storage occupancy data and trends | TractIQ | Physical occupancy 84.4% (Q4 2025) for CMBS-reporting facilities — a narrower universe than full-market or REIT same-store figures | 2026 |
| Matrix Self Storage National Report | Yardi Matrix | Stabilised street rate $16.91/sq ft annualised ($1.41/sq ft/month) | Sep 2025 |
| Eight dangers of postponing your desired self storage build | Inside Self-Storage | Land anchor points: suburban $50k–500k/acre, urban parcels often above $1M/acre; no national survey by market tier exists | Dec 2025 |
| Navigating the self storage site selection process | Inside Self-Storage | Land price is only meaningful relative to what is built; 13,000–15,000 rentable sq ft per acre; $100k and $333k/acre case examples | Apr 2020 |
| Determining capacity for your self storage building project | MakoRabco / Trachte | 40–45% site coverage ratio, equivalent to 17,500–19,602 rentable sq ft per acre | 2022 |
| Form 10-K, fiscal year 2025 | Public Storage | Developments underwritten to roughly 8% yield on cost; physical stabilisation in 2–3 years, revenue stabilisation in 3–4 years | Feb 2026 |
| Climate-controlled self storage feasibility study (SBA 504) | UnionMetric | Lease-up curve: ~40% physical occupancy at the end of year one, ~72% by end of year two, ~90% at stabilisation (representative case, not a survey) | 2026 |
What the model excludes
- Construction financing, interest carry, loan fees and closing costs
- Tenant improvements, move-in concessions and marketing during lease-up
- Operating reserves and working capital
- Property tax reassessment after construction
- Impact fees, which vary by municipality and can exceed $1M on large projects
- Site-specific conditions: rock, high water table, wetlands, utility relocation
- Environmental remediation, if any
These are not rounding errors. Financing and carry alone can add 5–10% to the all-in cost of a project. Treat the model's output as a construction budget and revenue screen, then build the full sources-and-uses statement with your lender.
Update schedule
Construction coefficients are reviewed quarterly, revenue benchmarks monthly, and state indices twice a year. Coefficients are compiled into the application rather than stored in a database: every change is a reviewable commit, which gives the site an audit trail for its own numbers.
When a coefficient changes, every table, FAQ answer and calculator result on the site changes with it. Nothing is hard-coded.
Corrections
If a figure on this site contradicts a published source you can point to, we want to know. Send the source and the page and it will be corrected or the claim withdrawn.
Open the calculator