Vending Site Selection: Screening Markets With Public Data

What to take away

Why federal data beats a host's sales pitch

A host who says yes is not a host who pays. Gut feel measures willingness. Employment counts measure repeat demand.

Federal statistics give you one yardstick across cities. Phoenix and Atlanta are measured the same way by the same agency, so a comparison between them is honest before you book a flight.

That matters when capital is tight. A machine, a card reader and a first fill tie up real money, and a wrong site costs more than a missed one.

Public data also survives a property manager's claim of 400 employees. You can check the tenant roster against occupation counts for that industry and see whether the number holds.

Use the data to build a shortlist, then visit. The visit confirms power, access and break room rules. The data decides which doors are worth knocking on.

Tourist foot traffic is the trap here. Tourists buy once. Employees buy every day.

Reading metro employment density from the Quarterly Census of Employment and Wages

The Quarterly Census of Employment and Wages draws on unemployment insurance records, so its counts reflect nearly every covered job. It reports employment and wages by industry for states, metros and counties.

Start at the metro level to compare markets, then drop to county where the data supports it. Download the annual singlefile, named like 2023.annual.singlefile.csv, or pull the same figures from the QCEW Data Viewer under annual averages by county.

Then look at industry concentration inside the metro. A metro where 30 percent of jobs sit in offices, hospitals and schools will out-earn a logistics hub of the same size.

Average weekly wage by industry is a rough proxy for spend per visit. Higher-wage tenants buy lunch without checking the price.

Write one row per metro into a spreadsheet: total employment, employment density, top three industries, average weekly wage. That sheet is the raw input for everything below.

Before you pitch a host on a metro that clears the screen, work through how to analyze vending machine locations at the building level.

Using OEWS occupation tables to size a building's daytime population

QCEW tells you how big a metro is. The OEWS occupational tables tell you who works inside a specific building type.

The tables break employment down by occupation and industry at national, state and metro levels. Use them to estimate headcount for a prospect without asking the host.

Download two files from the OEWS data page: the metropolitan file, named like oesm23ma.zip, and the industry file, named like oesm23in4.zip. The industry file carries one occupation inside one industry.

Then scale one building type: a 12-story medical office tower with 240,000 rentable square feet.

That result is an estimate, not a count. Treat it as a range and confirm it against a lobby directory.

OEWS also exposes shift patterns. Industries with heavy health care and manufacturing employment run multiple shifts, which stretches vending hours past the standard lunch window.

Pair occupation counts with the tenant mix. A tower of law firms and insurance offices produces steady midday demand. A warehouse produces concentrated break-time demand.

For the pitch and contract side of a host relationship, the site selection and location guide covers what to put in writing.

Matching industry mix to vending demand in American metros

Not every job produces the same vending spend. Industry mix is the second filter after density.

Office and professional services jobs cluster from 9 a.m. to 5 p.m. with two clear peaks. Hospitals and clinics run around the clock. Manufacturing and logistics skew toward shift changes.

Schools and universities bring seasonal swings that can halve summer revenue. Government offices are steady but slow, and often bound by stricter procurement rules.

Compare metros on that mix. Dallas and Houston carry large energy and logistics employment alongside growing office cores. Phoenix and Atlanta lean toward back-office, health care and distribution. Seattle concentrates tech and professional services.

A balanced mix gives you a route that does not depend on one industry. That is the goal.

Check state and local rules before committing. Sales tax registration runs through the state Department of Revenue, food permits come from the local health department, and both vary by state. Confirm the current requirement with the authority that sets it.

Weighing a second metro or a second route? Think about vending machine expansion carefully, because density gains shrink once drive time is included.

A yearly public-data screen for US vending markets

Use public data and a spreadsheet to compare candidate markets. Refresh the comparison once a year, or after any major release.

The table works one comparison from the QCEW annual averages. New York County holds about 2.3 million covered jobs on 22.7 square miles. Maricopa County holds about 2.2 million on roughly 9,200 square miles.

One 30 minute ring covers far more host buildings in New York County than in Maricopa County.

The score is a screen, not a verdict. Visit the top three metros and confirm that host types, parking and building access match what the data implies.

If a metro underperforms its score across two quarters, the cause is usually route geometry or host quality rather than the data.

Turning BLS geography data into a drive-time route map

The overview of BLS statistics by geography defines the metro boundaries you are ranking. Use the same definitions every time so comparisons stay honest.

Once a metro clears the screen, drop to the county level. QCEW county data shows which parts of the metro actually carry the employment.

Plot candidate buildings on a map and draw 15, 30 and 45 minute drive-time rings from your depot or home base. Set the working cap at 30 minutes.

Ten stops per shift inside that ring is a workable one-person route: roughly five minutes of driving and seven minutes of service per stop, plus loading and a margin for traffic.

Route economics set the stop count. Six stops inside 15 minutes beats twelve stops spread across 40 minutes, every week.

If you are unsure which metros deserve the first pass, look at markets where a vending machine business already shows proven demand patterns.

Set a hard cap on stops per shift and hold it. A marginal stop at the edge of the ring usually costs more in time than it returns in sales.

Redraw the map after every route change. Buildings get re-tenanted, employers move, and a site that scored well two years ago may now be half empty.

Where Census demographics inform market screening

Employment data covers the workday. Census data covers the rest of the day, and it can move a metro up or down the list.

Census demographic data for site selection includes population, age, household income and commuting patterns at the tract level, which is finer geography than most employment datasets.

Strong daytime employment with weak residential income still works if your machines sit inside secure buildings. Modest employment with dense, high-income housing supports lobby and apartment sites.

Age shifts product mix. Areas with a high share of residents under 18 or over 65 show different preferences and lower impulse purchase rates.

Commuting data tells you where workers actually live. If most employees reverse-commute from outside your drive-time ring, your after-hours options shrink.

Apply the demographic filter last. It should refine the market screen, not drive it.

Keeping the ranking current with the BLS release schedule

Federal data goes stale. QCEW publishes quarterly with a lag, and OEWS estimates update annually.

The BLS release schedule lists publication dates for employment, wage and occupation data. Put those dates in your calendar and re-run the ranking after each relevant release.

A simple cadence holds up. Full re-rank once a year after the annual OEWS and QCEW updates, light refresh each quarter when new QCEW data lands.

Keep a copy of the spreadsheet for each run. Comparing last year's ranking to this year's shows which metros are gaining employment and which are losing it.

Tie the refresh to your operations review. The performance indicators worth a monthly review show whether the ranking is turning into revenue per stop.

When a definition is unclear, the BLS glossary settles terms such as metro area, employment and wage before you compare two markets.

Common questions

How often should I re-rank metros?

Once a year in full, after the annual OEWS and QCEW updates. Add a lighter quarterly refresh when new QCEW data is released.

Can I use QCEW alone to pick a site?

No. QCEW gives metro and county employment. You still need OEWS for building-level occupation counts and Census data for the residential picture.

What employment density is too low?

Set the floor from your own route cost. A visit that takes 12 minutes of driving and service at a loaded cost of $30 an hour costs $6. Add the fill, card fees and spoilage. If the stop needs 250 visits a year, it must gross roughly $3,000 to $5,000 to earn its keep.

Divide that target by the average spend per worker at your comparable sites. That gives the number of jobs your ring must hold. A county under about 100 jobs per square mile is farm and forest, and one stop can eat an hour.

Do I need to register for sales tax before placing machines?

In most states, yes, through the state Department of Revenue. Food permits come from the local health department where they apply. Confirm your own filing with the state authority or a licensed CPA.

How many stops should one route cover?

Cap it at what you can service in a single shift including driving, restocking and collections. Density sets the number, not ambition.