Self-storage site selection and location guide: a closer look

What to take away

The order of the screen

Site work gets expensive in a predictable order, so run the cheap tests first and let them kill candidates before the costly tests begin.

Reversing this order is the common mistake. A parcel that has already absorbed survey and environmental spending acquires an advocate, and advocacy is a poor substitute for a supply count.

Defining the trade area

A storage customer chooses on convenience far more than on price, which means the catchment is a drive time rather than a radius. Draw the drive time in each direction at the hour customers actually come, which is not the hour you happen to be driving.

Typical working catchments run five to ten minutes of driving in dense urban markets and fifteen to twenty minutes in small towns and rural areas, where the next facility is further away. Measure at the hour the customer travels, not the average.

Then look for barriers. A limited-access highway with two interchanges, a river, a rail corridor, or a large industrial parcel can split one catchment into two.

A site on the wrong side serves half the population a radius would credit it with. Barriers are the single most common reason a market study overstates demand.

Households are only part of the picture. Small businesses, trades, and online sellers rent differently, stay longer, and want different sizes, and they cluster around commercial corridors rather than residential streets.

Counting supply properly

Count doors, not signs. Two facilities of the same footprint can differ by a factor of two in door count depending on unit mix, and it is doors that compete with yours.

A supply count in this form supports the lease-up arithmetic set out in the model for filling a new facility.

Start with the facility locators published by the large operators and the listing sites, then verify each site against the county assessor's parcel record and the building permits on file with the local planning office.

Self-storage specific market totals come from Yardi Matrix, Radius+, and the Self Storage Association. None of them publish door counts or unit mix, so Google Maps imagery and a site visit settle those.

The physical screen

Most candidate parcels fail here, and the tests are cheap.

The yield test

Two parcels of the same acreage can differ substantially in rentable square feet once setbacks, aisles, and building geometry are applied. Get a rough site plan before you get attached.

The number to compare is rentable square feet per acre, and the number to compare after that is land cost per rentable square foot. A cheaper parcel that yields less can easily be the more expensive one, and this comparison is where that shows up.

Typical yields run about 20,000 to 30,000 rentable square feet per acre for single-story drive-up buildings and 60,000 or more for a three-story climate-controlled building on the same land. Aisle width, setbacks, and how much of the site is buildable after drainage set where a parcel lands in that range.

Take two one acre parcels. One is $300,000 and yields 25,000 rentable square feet, which is $12 per rentable square foot. The other is $250,000 and yields 12,000, which is about $21. The cheaper parcel is the more expensive project.

Rentable area is also not the same as gross area. Corridors, stairs, offices, and mechanical space are all built and paid for and none of it rents. Ask the designer for the ratio on each candidate plan and treat a poor ratio as a permanent tax on the project.

Accessibility, records, and the other things settled at design stage

Accessibility is drawn, not retrofitted. Parking, the parking-to-office route, counter height, and door approach are easier to get right on paper.

The Department of Justice guide to the Americans with Disabilities Act for small businesses is the plain-language reference for what a small operation should consider. Give it to the architect during the first site plan, not after permitting.

Keep the site screen itself as a record: each candidate, the date it was screened, the supply count behind it, and the reason it was dropped. That file answers the lender's questions later and stops the same parcel being reconsidered on enthusiasm.

Basic expectations for business records are set out by the Internal Revenue Service in its page on the records a business should keep. Site diligence spending is deductible or capitalized depending on how the project proceeds, which is a question for your accountant while the file is still current.

The Small Business Administration guidance on market research and competitive analysis is a reasonable framework for the demand side of the study, though it is written generally and the door-count discipline above is what makes it useful for storage specifically.

Staffing follows the site, not the other way around

A site with poor visibility needs more marketing. A site with a complex layout or two buildings needs more attended hours. A site on a single frontage with a straightforward gate can often be run with fewer.

Decide the staffing model while the site is a candidate; it changes the operating budget you compare candidates on. Price the hours against published wage estimates for your own metro area, not a remembered figure.

The Bureau of Labor Statistics publishes employment and wage estimates by occupation and area, giving the payroll line in the comparison a defensible source.

Scoring candidates against each other

Build a table with one row per candidate and one column per factor that would actually change your decision. Weight the columns before you fill them in, because weighting afterward is just a way of confirming a preference you already had.

Useful columns include drive-time population and business count, competing doors by size band inside that drive time, permitted supply, rentable square feet per acre, land cost per rentable square foot, visibility, zoning risk, and utility capacity. Leave out anything you cannot fill for every candidate, since a column with gaps in it quietly stops being a comparison.

Common questions

How far will customers really drive?

Farther for price, less far than you would like for convenience, and the answer differs between a dense urban catchment and a rural one. Rather than guessing a distance, ask existing operators in the area where their customers come from, and check the barriers on a map.

Is visibility worth paying for?

Often, and the way to test it is to price the alternative. Work out what a year of paid demand capture costs at the volume the site would need, and compare that with the land premium. Sometimes the premium is cheap and sometimes it is not.

Can a poor site be fixed with a better building?

Layout and mix can compensate a little. Zoning, access, and a saturated size band cannot be fixed with construction, which is why they sit in the cheap screen at the front.

Should I buy land before the zoning question is answered?

Only with a contract that lets you walk away if the answer is no. Storage is not a permitted use everywhere it looks like it should be, and finding out afterward is the most expensive way to learn it.