How the estimate is made

The steps

Find the property's record. We match the address or parcel ID to the jurisdiction's parcel record: its property type, building and land size, assessed value and recorded sales.
Choose comparable sales. We look for recorded sales of properties of the same type (office with office, retail with retail), the most similar first, nearby sales preferred. A sale is left out when its price is nominal (a token amount far below the property's value) or its date can't be read, and a sale recorded on several parcels at once is counted once. A land sale is also left out, treated as the price of more than its lot, when the lot is under an acre, the price is more than five times the lot's assessed value and its price per square foot of land more than two and a half times the middle of the other land sales (for a lot under 1,000 square feet with no assessed value on record, more than five times the middle of the other land sales). Smaller independent cities also draw on the surrounding county: for example, the City of Fairfax draws on Fairfax County sales.
Keep to a size band. A sale is used only when the building is between one fifth and five times the size of the subject's, and sales between one third and three times its size are preferred. Land is compared on land area. See size band.
Adjust each sale for market timing. Each sale's price per square foot is moved to today's market using the trend in recorded sales of the same property type in the same jurisdiction over the last ten years. Where there are too few sales to measure a trend, a fixed regional rate is used. The yearly trend is limited to between −8% and +12%, and each sale's total adjustment is capped. See time-adjusted $/SF.
Weigh the sales. The sales that match the subject most closely (on type, size, location, sale date and zoning) count the most. Zoning counts only when both properties have zoning on record. When at least two sales are of the same specific kind, such as two office condos, those are used on their own. With four or more sales, a price per square foot below 40% or above 250% of the middle one is set aside. The weighted price per square foot is then multiplied by the subject's size. If fewer than three usable sales fall within the report's time window, older sales are used and the report notes it.
Reconcile with the public assessment. We compare the result with the property's assessed value, adjusted by the local ratio between recorded sale prices and assessed values for that property type (the market-adjusted figure). The weaker the comparable sales, the more weight it gets, and the report notes when this happens. For a property type where a test written in advance, on sales made after the latest assessments, showed that a newer blend comes closer to sale prices, the market-adjusted figure keeps a share of the value even when the sales are strong, and the report says so. When there are too few usable sales, or they disagree too much, the value is the public assessment itself (or, for a property type where such a test showed it comes closer to sale prices, the market-adjusted figure), and the report's method says so. A value that rests on the assessment shows both figures: the assessment, with its tax year (for the District of Columbia, the date it values the property at; for Maryland, its three-year cycle), and the market-adjusted figure, with the number of recent sales behind its ratio and whether those sales carry the jurisdiction's own validity code. The market-adjusted figure is labelled with its share of the value where it was blended in, and "not used in the value" where it was not.
Use the property's own recent sale. If the property itself sold in the last five years for a real price (not a token amount, and at least a quarter of its assessed value), and the comparable sales are not strong, that sale is the value, moved to today's market with the same trend. A sale recorded on several parcels at once is shared out by each parcel's share of the assessed value. A farm, a large-acreage tract, a house on commercially classed land or a rural business that sold on its own in the last 36 months is valued on that sale at its recorded price, with the assessment as a cross-check. The report shows the sale with its date, and says so when no County validity code verifies it (it then passed our own arm's-length checks).

Each report names the method that set its value, and lists every comparable sale considered, whether it was used and, if not, why.

The Excel download may also show an income cross-check, built from benchmark market rents and capitalization rates, for comparison only. It never changes the estimate.

How to read the confidence

Each estimate carries a confidence, shown as a word and a percent, such as High (74%).

High: 70% or more.
Moderate: 45% to 69%.
Low: below 45%.

Confidence rises with the number of usable sales and how closely they match the property, and falls when their prices per square foot disagree. Blending in the assessment raises it a little. A value taken from the assessment alone is Low. A value taken from the property's own recent sale starts high and falls as the sale gets older. The low-to-high range around the value is wider when confidence is lower; around a value taken from the assessment alone it is at least plus or minus 20%.

Confidence describes the evidence behind the estimate. It is not a probability that a sale would close at the estimated value.

Special cases

Land is compared on price per square foot of land, not of building.
Several parcels as one property. A range address or an assemblage is valued as one property: comparable sales applied to the combined size, checked against the parcels valued one by one. A sale recorded on more than one of its parcels is counted once.
Properties that aren't commercial (for example a church, a school, a farm or a house) aren't compared with commercial sales. Their value is the public assessment, except a recent sale of their own as described above. A farm assessed on its agricultural use, as Maryland and Prince William County assess farm land, gets no market value from that assessment: the report shows it as an agricultural use assessment, not a market value.
A house on commercially classed land where the County puts 80% or more of the value in the land, and the house is small for its lot or worth little beside the land, is compared with commercial land sales, and the report says that a house stands on it.
A building owned by a public body or a tax-exempt organization whose use sells on the market (an office, a store, a warehouse) is valued from sales of buildings of the same use, checked against its full assessment. Its own sales are never used as comparable sales for other properties. Rent-restricted public housing is valued on the assessment, and so is vacant land a public body holds, which may be dedicated or restricted.
Land size. Where the County records the land size more than once (its assessment record and its parcel map), the report says whether those figures agree, always with the unit, and lists them in Recorded facts.
Places with thin public data. Some jurisdictions publish less, such as no property types or no recorded sales. Those are marked "Live, limited data" on the live jurisdictions page.
A member's own comp selection. A member can leave out sales, swap in the report's alternates, add a named sale or enter an off-market sale, and see a figure of their own, worked out from those sales with the report's own weights and market trend. It uses the sales only, never the assessment or the property's own sale. It is the member's opinion, labelled with their name and firm, not Locumetric's estimate and not an appraisal. Locumetric's estimate doesn't change, and sales a member enters never enter it.

Where the data comes from

Each section of a report names its sources and, where the source gives one, the date it was current. By kind of record:

Parcels, assessments and recorded sales: each jurisdiction's assessment and land records and its mapping (GIS) data.
Owners: the owner of record on the parcel, and state business registries such as the Virginia SCC and the DC business registry for the companies behind it. Names are shown as filed.
Loans and leases: SEC CMBS filings, HUD insured multifamily loans, SBA 504 loans and the GSA lease inventory.
Zoning and land-use cases: county and city zoning maps and planning records.
Environmental records: US EPA, Virginia DEQ, the Maryland Department of the Environment and the DC Department of Energy and Environment.
Towers and billboards: the FCC Antenna Structure Registration, county telecom records and Maryland SHA sign permits.
Site and location: FEMA flood zones, Opportunity Zone designations (the 2018 designations, from HUD's layer, which run through December 31, 2028, and the U.S. Treasury's designations announced November 5, 2026, which run from January 1, 2027 through December 31, 2036, on 2020 census tracts; a report shows both, each with its dates), the National Register of Historic Places, transit stations and traffic counts from transportation agencies, and county map layers.

Records that are matched to a property by address or by map location, rather than by parcel ID, can belong to a different property. Reports label those "Possible match".

How current it is

Public sources publish on their own schedules, and we load them periodically, so a new filing can take some time to reach a report. Each section of a report names its sources and, where the source gives one, the date it was current ("as of").

A report is a snapshot of the records on file when it ran. Its values don't change afterwards. A newer report on the same property uses the records on file then, and shows what changed since the earlier one. When a measured change to the method switches on between the two, What changed says so, with its date, so the move isn't read as the market.

Live in 16 jurisdictions across Virginia, Maryland, and Washington, DC. Nationwide coverage is next. See each jurisdiction and its data notes.

Limits

Public records can be late, incomplete or wrong, and we don't check them in person.
A recorded price can include more than the real estate, such as a business or other parcels.
Private deals, rents and loans that aren't in public filings aren't in the estimate.
Few comparable sales, or sales far apart in price, make the estimate less certain. The confidence and the range say so.
The zoning, environmental and site sections report public records. They are not a zoning determination, an environmental site assessment or a survey.