The wrong number of vending machines is the most expensive configuration mistake an operator can make. Two machines in a high-traffic lobby and you have a queue problem; ten machines in a venue that only needs four and you have a utilization problem. Both look like bad machine design from a customer's perspective. Both are fleet-sizing mistakes a year before they get fixed.
Sizing comes down to four inputs: daily foot traffic, average dwell time, category mix, and the restock cadence you can actually sustain. None of those inputs is a number from a vendor spec sheet. They come from your venue — the people who walk in, what they reach for, and how often someone is going to be in front of the machine with a clipboard.
Step one — count the traffic, not the square footage
Square footage tells you almost nothing. A 2,000-square-foot coworking space with 60 daily visitors is a different machine count than a 2,000-square-foot waiting area that sees 600. The rule of thumb is one Core-1 per 200 to 400 daily visitors, scaled by dwell time: longer dwell means more visits to the machine per visitor, which means fewer machines per headcount.
Step two — pick a category mix before you pick a count
Decide up front whether the venue is a snack-and-beverage stop, a beverage-only stop, or a food-and-beverage destination. Snack-and-beverage stops need roughly one machine capacity per 300 daily visitors. Beverage-only can compress to one machine per 500 to 600 once you tune the slot ratio toward cans. Anything heavier — meals, fresh items — needs more machines per headcount, not fewer, because the dwell-time on the meal SKU is longer and you cannot safely double-stack capacity the way you can with a bottled drink.
Step three — restock cadence is the real ceiling
A machine is only useful if it is stocked. Five well-stocked machines perform better than ten machines where the operator has stretched theming across two staff restock shifts per week. If your staffing model is currently a twice-weekly restock, that is your fleet ceiling unless you add staff or switch to a route model.
Worked example
A regional airport terminal sees roughly 4,000 daily visitors across the post-security concourse, with a 45-minute average dwell. Divided cleanly into two zones (A and B), each zone sees about 2,000 visitors with a 45-minute dwell. The 45-minute dwell pulls more total machine interactions per visitor than a 10-minute dwell, so you can sustain a higher per-machine utilization rate — but each interaction is also lower-velocity. The right answer there is four machines total, two per zone, allocated across snack-and-beverage with a leaning toward beverage on the high-dwell side and snack on the connecting corridor. Four machines on a twice-weekly restock schedule is sustainable; eight is not.
The swap-friendly piece
Once the fleet is right, the per-shelf swap design of Core-1 keeps restock velocity high without compounding the failure risk. A single shelf going down does not take the machine offline, so a venue does not need to over-buy spare capacity to absorb a single-shelf failure the way it would with a service-locked machine. That is the second-order reason fleet sizing decisions feel easier when the failure mode is recoverable — you are sizing on traffic, not sizing on cushion.