$2.2 Billion Went Through Amazon. None of It Was Consumers

Institutional demand doesn't arrive as an RFP. It arrives as a purchase order from a buyer you never merchandised for — and most consumer brands are structurally unable to accept it.

Ask a consumer brand why it doesn't sell to schools, cities, or state agencies, and the answer is always the same: RFPs, bonding, eighteen-month cycles. That reasoning is sound, and it answers a question nobody asked.

Institutional demand arrives through three doors, not one. The door most brands are refusing is the third. The first requires no government sales expertise at all — only catalog work most brands have already half-completed without knowing what it was for.

The buyer is already there

Public institutions have quietly made a marketplace their default source for routine goods.

In 2025 the Institute for Local Self-Reliance analyzed purchasing records from 128 cities, counties and school districts serving 51 million residents, plus 122 state agencies. Its estimate: local governments spent roughly $2.2 billion through Amazon in 2023 — close to four times their 2016 spend. The report is a critique of that concentration, not a celebration of it. Its central factual claim is not in dispute.

The purchase does not run through a procurement officer negotiating a contract. It runs through a school safety coordinator with a purchasing card and an approval workflow, buying the same ASIN a household would buy, at the price the algorithm is showing that afternoon.

Where the conversion breaks

A meaningful share of institutional demand is already crossing listings built entirely for consumers — and losing at checkout.

It breaks for reasons that have nothing to do with procurement law. The buyer needs a quantity tier a consumer pack doesn't offer. Tax exemption applied at checkout. A quote for 400 units on a listing built for one. Section 889 attestation on file before federally-funded dollars can clear. The offer visible inside a punchout catalog, reaching them through their own e-procurement system rather than requiring them to leave it.

Amazon Business supplies all of this infrastructure — punchout and punch-in integration, single sign-on, guided buying rules, bulk quantity quotes, tax exemption enrollment. The platform has done the work. Most brands have simply not configured their side of it.

The contract that replaces ten thousand bids

Win one solicitation. Thousands of agencies can then issue a purchase order immediately.

Public agencies must competitively bid above a threshold — that requirement is what makes government selling slow. But most states have adopted legislation permitting interlocal agreements and cooperative contracts: an agency may buy from a contract that a different public agency already solicited and awarded, and satisfy its own competitive bidding obligation without running a bid of its own.

Sourcewell holds hundreds of these competitively solicited contracts across more than 50,000 participating government, education and nonprofit organizations, with roughly $5.85 billion in contract sales in FY24. OMNIA Partners runs the same lead-agency model at $30B+ reported annual volume.

For a district with one procurement staffer, a cooperative contract turns a solicitation into an afternoon. The brand on that contract is not winning on price. It removed the buyer's hardest constraint.

Why readiness categories compound fastest

Demand is compliance-driven, not discretionary — and the product expires.

In California, Education Code sections 32280 through 32289 require every school to maintain a Comprehensive School Safety Plan with disaster procedures, and state law requires an earthquake emergency procedure system in every public school building with an occupant capacity of fifty or more students or more than one classroom. No statute says buy a kit. Taken together, the framework makes the kit the practical expression of compliance — and some version of that structure exists across most states.

Then there is shelf life. Emergency water, food rations and batteries expire on a schedule the buyer cannot defer. A brand selling a durable good into an institution makes one sale and waits a decade. A brand selling consumables into a compliance mandate is not accumulating a series of wins. It is building an installed base.

What it costs

Three things get worse. Plan for them rather than discover them.

Cash conversion lengthens — institutions pay on purchase order terms, frequently net 30 to net 60, and a marketplace business accustomed to rapid settlement feels the working capital drag immediately.

Revenue becomes calendar-bound. Most state fiscal years run July through June, so institutional buying concentrates against budget availability rather than consumer seasonality. It does not smooth a Q4-weighted retail business; it adds a second, differently-shaped peak.

And public contract pricing is public record. A brand that has spent years enforcing MAP across its marketplace channels now has a published institutional price sitting in a searchable database. That is manageable with deliberate pack and SKU architecture, and unmanageable without it.

The readiness test

The honest test for door one is not strategic. It is clerical, and it can be answered this week.

A brand that cannot clear this list is not being blocked by the difficulty of government selling. It is being blocked by seven pieces of paperwork.