How to build a sales comparison grid for a commercial appraisal

The sales comparison grid is the part of a commercial appraisal that looks simple and takes the longest. The arithmetic is easy. What takes the time is finding the right comps, getting each fact about them right, and keeping the grid, the narrative and the exhibits saying the same thing. This guide walks through the order that keeps it manageable.

1. Decide what the grid has to show

Before picking a single comp, write down the property type, the highest and best use you are testing, and the unit of comparison the market uses for it. Improved commercial property is usually compared on price per square foot of building. Land is compared on price per acre or per square foot of land. Apartments are compared per unit. Choosing the unit first stops you collecting comps you cannot compare.

2. Choose the comps

Look for sales that are:

Recent enough to reflect the same market conditions, or close enough that a market-conditions adjustment is supportable.
The same kind of property in use, size and quality, in the same or a competing market.
Arm's-length. A sale between related parties, a foreclosure, or a transfer for nominal consideration says little about market value.
Confirmed. A number from a county list is a lead. Confirm it against the recorded deed, and where you can, with a party to the sale. See Florida sale data for appraisers.

Three to six well-confirmed comps are worth more than ten that you have not checked.

3. Set up the rows

A commercial sales grid usually carries, for each comp:

address and parcel number, sale date and sale price
land area and building area (say which building area: gross, net rentable)
price per unit of comparison
year built and condition
zoning and land use
the facts behind the transactional adjustments: property rights conveyed, financing, conditions of sale

Name the rows in your own words and keep the same order from one report to the next. A grid that always reads the same is quicker to check, for you and for a reviewer.

4. Adjust in the right order

The standard sequence has two stages.

Transactional adjustments come first, one after another, each applied to the price as already adjusted:

property rights conveyed
financing terms (a seller-financed sale at a below-market rate is worth less in cash terms than its price)
conditions of sale
expenditures made immediately after purchase
market conditions (time)

Property adjustments come second: location, physical characteristics, economic characteristics and use. Whether you quantify them in dollars or percentages, apply the same method to every comp and say so in the report.

5. Reconcile

The adjusted prices should land in a range, and the work is explaining which comps deserve the most weight and why. A comp that needs the largest adjustments is usually the weakest evidence, whatever its price per square foot says.

Mistakes that cost an afternoon

One price, several parcels. When a deed conveys two parcels for one price, county lists often show the full price against each parcel. Counted twice, the price per acre doubles. Check the deed.
The wrong building area. County figures often include covered porches, carports and canopies. Decide which area you are comparing and use it for every comp.
The wrong date. The deed date, the recording date and the date on a county list can all differ. Use the date you can document, and use the same kind of date for every comp.
Ignoring seller financing. If the seller holds a mortgage, the price may not be a cash price.
Retyping. Every time a comp is typed into a new spreadsheet there is a chance of a wrong digit. Keep one record per comp and build every grid from it.

Making it faster

If you build a grid a few times a month, the biggest saving is not in the arithmetic. It is in not retyping. ValuBids keeps each comp once, fills in the address, size, year built and last sale from a parcel number, and lets you check the comps you want and the rows you use. The grid exports to Word or Excel with your own headings, colors and column order, so there is no hour spent reformatting at the end.