A wholesale distributor in the upper Midwest, three warehouses and a counter business, got a renewal notice from its ecommerce vendor in the form of an invoice. Not a letter, not a call from an account manager. An invoice, emailed to accounts payable, for a twelve-month term that had already been locked in by the time anyone in operations saw it. The controller paid it. Then she went looking for the contract, because the number was higher than she remembered and she could not tell from the invoice why.
What she found is worth walking through, because the document she was reading did not resemble the one the company signed a decade earlier. Same vendor category, same business problem, almost nothing else in common.
The invoice that arrived before the conversation
The contract was six years old with two renewals stacked on top of it. The original signature page covered a hosted storefront, a catalog feed, and a flat annual fee. By the time the controller sat down with it, the live agreement was the original document plus two order forms, a linked usage schedule that lived on the vendor's website, and a data processing addendum nobody at the company had read.
Three things were driving the increase. First, an annual uplift clause: the fee rose by a fixed percentage each term, compounding, with no requirement that the vendor flag it. Second, a usage tier. The company had crossed a threshold on monthly transactions, which moved it into a higher band automatically. Third, two modules that had come in free during an onboarding promotion had converted to paid line items at the second renewal, exactly as the order form said they would.
None of that was hidden. All of it was findable. The problem was that the findable version was spread across four documents and one web page, and the only date that mattered had passed.
This is the part that scales differently. A household subscribing to a photo backup service and forgetting to cancel loses a few dollars a month and can call the card issuer. A distributor with a platform wired into its ERP, its pricing rules, its customer-specific catalogs and its freight logic cannot simply stop paying. The switching cost is the leverage, and the renewal date is when that leverage is at its highest.
Ten years ago you bought a license. Now you rent a meter
The shape of the decision has changed in a way that rewards reading earlier and punishes reading late.
A decade ago, a mid-size distributor buying ecommerce software was mostly buying a perpetual or long-term license, often installed on hardware the company owned or leased, with a separate maintenance percentage on top. The money questions were upfront: implementation hours, who pays for customization, what happens when the vendor releases a major version. If the relationship soured, the software kept running. Badly, unsupported, but running. That was the floor.
Today the same company is buying a subscription to something it does not possess. The floor is gone. If the term lapses or payment stops, the storefront goes dark, and with it the ability for customers to place orders at two in the morning. That single structural change is why the renewal clause has become the most expensive paragraph in the document.
The pricing basis moved too. Ten years ago, the variable was seats. Now it is almost always consumption: orders processed, gross merchandise value, API calls, catalog records, storage. Modern b2b ecommerce platform software tends to price against the thing that grows when the business succeeds, which means a good year produces a renewal quote the finance team did not budget for.
And the list of things the vendor now promises has lengthened. Uptime commitments with service credits. Security representations. Subprocessor lists. Breach notification windows. A decade ago many of those clauses did not exist in a distributor's software agreement at all. Now they are the clauses with real money behind them, and they are the ones most worth negotiating while there is still time on the clock.
The same clause does not mean the same thing in every state
Here is where the distributor's review got genuinely useful, because it operates across state lines and its contract did not.
Automatic renewal is regulated at the state level, and the rules are not uniform. A number of states require clear and conspicuous disclosure of auto-renewal terms and advance notice before a term rolls over, with specific windows and specific delivery methods. Others say far less. Several of those statutes were written with consumer subscriptions in mind, and whether they reach a negotiated business-to-business agreement depends on the statute's own definitions and on how courts in that state have read them. The Federal Trade Commission oversees negative option and automatic renewal practices at the federal level, and the subject has been an active area of its rulemaking attention for years.
The practical consequence for a company, as opposed to a household, is that the governing law clause stops being boilerplate. If the agreement says it is governed by the law of the vendor's home state, the notice protections a buyer might have expected under its own state's rules may not apply. The distributor's contract named a state neither party operated in. Nobody had noticed.
Two more regional threads turned up in the same read.
- Sales tax calculation. Since economic nexus became the standard for when a remote seller owes tax in a state, a distributor selling into twenty states may have filing obligations in most of them. The platform calculates the tax. The contract decides who is responsible when the calculation is wrong, and whose indemnity covers the assessment. Older agreements, written when tax turned on physical presence, often say nothing at all.
- Data handling. State privacy statutes now vary on consent, deletion requests, and what a service provider may do with data it processes. The data processing addendum the distributor had never read is the document that determines whether the vendor can meet a deletion request from a customer in a state with an active privacy law. A household never encounters this. A company with a customer list does, every time a request arrives.
What ninety days of lead time actually buys
The distributor rebuilt its approach around the calendar rather than the document. The renewal date went into the finance calendar with two upstream reminders: one at ninety days to start the review, one at the outer edge of the notice window with the actual required delivery method written next to it.
Then the team assembled a single packet and kept it in one place: the master agreement, every order form, every addendum, and a dated printout of any schedule that lives on the vendor's website. Terms that sit on a web page can be revised. A screenshot with a date is the only record of what the terms said when the company agreed to them.
From there the asks were specific. A cap on the annual uplift. Notice of any tier change before it takes effect rather than after. A stated price for the next two terms instead of one. A data export commitment with a format and a deadline attached, so that leaving is a project and not a hostage negotiation. And a governing law provision that at least sits in a state where one of the parties does business.
Most of that was granted. Not because the vendor was generous, but because the conversation happened in month nine rather than month twelve, while the account was still worth keeping and the alternative was still real.
The renewal came in lower than the invoice that started the whole exercise. The more durable result was that the company now knows which single date on its calendar carries the most negotiating power, and it stops being the vendor's date and becomes its own.