Vending Commission Rates and Contract Terms 2026: What Public Contracts Show

What do vending contracts actually say? Private placement agreements are confidential, but public agencies post theirs. We read 28 public vending solicitations, agreements and awards from cities, counties, school districts, a housing authority and public colleges and recorded what each one states: the commission, the term, insurance minimums, healthy-vending rules, card readers and service frequency. Every row links to the original.

Download all 28 contracts (CSV) · Full table · Methodology · Cite this report

What public agencies get paid

Most public solicitations do not print a rate at all: in 13 of the 28 documents the bidder names the commission, and the agency picks the best offer. The rest fix a split, set a floor, or record what was actually signed. Across the 7 documents that state one rate for every machine, the median is 25% of sales (range 10% to 52%). Counting only signed or awarded rates (n=4), the values are 52%, 25%, 10%, 22%. Where rates differ by product line, beverages carry the higher rate: 15%, 40%, 25% for drinks against 10% and 8% for snacks in the documents that split them.

Two cautions. Miami-Dade's 52% came from a sealed bid where price decided the award, and the next two offers were 43% and 42.5%; that is a countywide contract awarded on price alone, not a benchmark for a single break room. And Redmond's 10% is a state rule: in Washington, machines in public-access areas pay at least 10% of gross to the Department of Services for the Blind. Federal and many state buildings have similar blind-vendor priority rules; see how to get vending machines in government buildings.

How long the contracts run

The usual shape is a short first term with renewal options at the agency's discretion. The median initial term is 3 years (n=21, range 1 to 10), and 5 years if every renewal is taken (n=17). One-year terms with annual renewals are common in small cities and school districts; the longest (10 years or more) are university beverage-and-vending contracts (Jackson State, Missouri State) and the Los Angeles parks concession once its two 5-year extensions are counted.

Insurance, card readers, healthy items and service

Insurance. 7 documents name a general-liability amount, and every one asks for $1,000,000 per occurrence; 4 add a $2,000,000 aggregate and 4 require $1,000,000 of auto liability. Budget for that certificate before you bid.

Card readers. Of the 17 documents that address payment, 13 require card acceptance on every machine, 3 require it only in busy spots or on request, and 1 (City of Oak Harbor, 2023) lists coins and bills only. Gila County and Takoma Park also name tap-to-pay or mobile payments.

Healthy items. 16 documents mention healthy products; 8 put a number on it, with a median of 50% of items (range 20% to 75%). The Wake County school district points to the USDA Smart Snacks standard instead of a percentage.

Service. 13 documents say how often to restock or how fast to respond. Restocking runs from seven days a week (Erie County) to at least monthly (Seattle Housing Authority); Gila County scales it to traffic (twice a week at busy sites, every two weeks at remote ones). Where a repair response time is set it is 24 hours (City of Takoma Park), 24 hours (City of Redmond), 8 hours (County of Erie).

All 28 documents

Every document we used, newest first, with a link to the original. "Not stated" means the document does not say; we did not fill gaps from other sources. The CSV has the same rows with the exact notes.

Methodology and limits

• Sources. 28 public documents (requests for proposals, invitations to bid, signed agreements, board awards, an internal audit and one county announcement) found through web search on October 8, 2026, starting from the Takoma Park 2026 vending RFP and Gila County's GIL-RFP-2026-008. By issuer: 8 cities and villages, 8 counties, 7 public colleges and universities, 3 school districts, 1 utility district, 1 housing authority. Dates run from 2015 to 2026; 24 are from 2023 or later.
• Extraction. Each field was read from the document text. A field the document does not state is recorded as "not stated" and is never estimated. Where a document showed a number whose role we could not confirm (current rate, minimum or example), the number was left out.
• Medians. One value per document. A median is only reported with at least 3 documents behind it. Rates that apply to one product line (snacks or drinks) are reported separately from rates that cover every machine.
• Public-sector skew. Every document here is from a government agency, school district, housing authority or public college. Public buyers publish their terms, private landlords and employers do not, and public contracts tend to be larger, multi-site and competitively bid. Expect private offices, apartments and small businesses to settle lower: the typical commission by location type guide covers that side.
• Small sample. This is a compiled set of documents, not a random sample of every public vending contract in the US. Treat the medians as a reference point for bidding, not an industry average.

Using this when you bid

Price your offer from your own per-machine numbers, then check it against these ranges. To find public buildings and other locations near you, VendBuddy's Lead Finder searches by ZIP and location type, and the 2026 Location Index shows which location types operators look at most. For the bid itself, see how to get vending machine contracts.

More research: all VendBuddy reports.