Minnesota's Auto-Renewal Law Doesn't Ban Retention Offers. It Requires Consent Before You Can Show One.

Every state auto-renewal fight in the last two years has been about the same target: the cancel button. How many clicks. Whether the exit channel matches the signup channel. How many days of notice before a renewal charge hits. Minnesota looked at the same problem and regulated something none of the other states touched — not how a subscriber gets out, but what you're legally allowed to say to them on the way out. Specifically: you can't offer them anything to stay until they've told you they want to hear it.

If you build or run a cancellation flow, this is worth understanding at the statute level, not the summary-blog-post level, because the mechanism it creates is unusual: it doesn't ban anything you're currently doing, it just tells you the order you're allowed to do it in.

What the statute actually restricts

Minn. Stat. § 325G.58, subdivision 4(a) says that once a seller receives a cancellation or termination notice, it cannot "make or provide additional benefits, contract modifications, gifts, or similar offers to the consumer until the seller has obtained permission from the consumer, granted by the consumer after notice of cancellation or termination was given." Subdivision 4(b) adds a second constraint most compliance write-ups skip past: "A seller can only seek a consumer's permission under this paragraph once per cancellation or termination attempt." One ask. If the consumer says no, or doesn't respond, that's the end of it for that attempt — you don't get a second framing of the same question three screens later.

Read literally, that's a two-step gate on the exact moment a cancellation flow is built around. Step one: the consumer says they want to cancel. Step two, before any specific incentive language can appear, the seller has to ask permission to show one — and can only ask that once. Only after a yes can a discount, a bonus credit, an extended trial, or any other "gift or similar offer" actually be displayed. Show the offer before the ask, or ask a second time after a no, and you're outside the statute regardless of how generous the offer itself is.

What's exempt — and it's a useful line to design around

Subdivision 5 draws the boundary precisely, and it's narrower than most teams assume when they first hear about this law. Four things stay available with no consent gate at all:

That split matters more than it looks like at first read. A pause or downgrade offer — the two retention responses that don't involve a new incentive, just a cheaper existing path — can sit on the base cancellation screen with no gate in front of it. A discount, by contrast, is exactly the kind of "gift or similar offer" the statute targets, and it can't appear until the consumer has separately agreed to see options to stay. Two of the three levers in a typical pause-discount-downgrade sequence are unrestricted; the third is the one that needs a consent step inserted in front of it.

How this compares to every other state's approach

We've covered California's, Colorado's and New York's, and Virginia's auto-renewal laws individually. Lined up next to Minnesota's, the difference in what each one actually touches is stark:

Every other state law in this space assumes the retention offer itself is fine and regulates the path around it — how many clicks, which channel, how much notice. Minnesota is the only one that reaches into the offer's timing and sequencing directly. If you've been tracking auto-renewal compliance as a checklist of "put the cancel button somewhere obvious and match the signup channel," Minnesota's law sits in a completely different category of requirement, and a compliant California flow can still violate it.

Why regulators are looking at this specific moment

Minnesota didn't invent this problem out of nowhere. ICPEN's global sweep of 642 subscription services in early 2024 — the same sweep the FTC cited in its own dark-patterns review — found that cancellation flows routinely stack incentive offers as an obstacle rather than a courtesy, and that the friction starts well before the cancel screen:

A consent gate on incentive offers is a fairly direct response to that pattern. If the documented problem is sellers using the retention moment to wear a subscriber down with offer after offer rather than a single relevant one, requiring an explicit yes before any offer appears — and capping the ask at once — removes the wearing-down mechanic by construction. It doesn't matter how the offer is worded if it can only be shown after the consumer has already agreed to look at it.

Enforcement is real even though the statute has no built-in damages number

Sections 325G.56 through 325G.63 don't create their own private lawsuit or set their own dollar penalty — a detail that's led some compliance write-ups to undersell how much exposure this actually carries. Enforcement instead routes through Minnesota's general consumer-protection framework: the Attorney General can seek a civil penalty of up to $25,000 per violation under Minn. Stat. § 8.31, subd. 3, and because a violation of this chapter also counts as a deceptive trade practice under § 325F.69, an individual consumer can bring a private claim under § 325F.70 — with attorney fees available — if the action "benefits the public," the same standard used across Minnesota's broader consumer-fraud statute.

The statute's good-faith safe harbor (§ 325G.63) only shields you from civil penalties, and only if you made a genuine compliance effort. It does nothing about the self-executing consequences that apply regardless of intent: a gift or benefit provided without the required permission becomes the consumer's unconditional property, a contract modification pushed through without proper notice is void and unenforceable, and a seller that skipped the required confirmation or notice paperwork loses the ability to dictate how the customer is allowed to cancel at all. None of those depend on whether you meant to get it wrong.

What to actually change in your flow

The fix isn't a legal disclaimer — it's a sequencing change, and it's small enough to ship without redesigning the whole flow:

• Move the ask ahead of the offer. After a subscriber submits a cancellation reason, insert one screen: "Would you like to see options before you go?" Only render discount, credit, or contract-modification copy after a yes. Pause and downgrade options can stay visible on the base screen either way — they're exempt.
• Cap it at one ask, and mean it. If the consumer declines or ignores the prompt, that's final for the attempt. Don't re-surface the same question reworded on the next screen, and don't auto-advance past a "no" into an offer anyway — both are exactly what subdivision 4(b) is written to stop.
• Log the consent event separately from the cancellation-reason event. "Reason submitted" and "permission granted to see an offer" are two different, timestamped actions. If you're ever asked to demonstrate a good-faith effort under § 325G.63, that log is the evidence.
• Segment Minnesota residents if your product has any consumer tier. The chapter only reaches "consumer" agreements for personal, family, or household purposes, so a pure B2B contract is out of scope. But billing address or account-level location data is enough to flag which subscribers need the gated sequence and which don't.

There's also a case for building the consent gate everywhere, not just for Minnesota residents. Our own research on cancellation flow dark patterns found that stacked, un-asked-for retention offers measurably cut trust and usability scores even in flows that are entirely legal where they run — Minnesota just turned a UX best practice into a statutory requirement for its own residents. Building "ask once, offer once" as your default sequence means you're never retrofitting a state-specific version of your flow if another legislature copies this approach, which is exactly the pattern that turned California's button-placement rule and Colorado's B2B extension into a multi-state trend within a single legislative cycle.

None of this changes the actual economics of a retention offer — our break-even math on when a save is worth making still applies once the offer is on screen. What Minnesota changes is who gets to decide whether that screen appears at all: the consumer, once, before you get to make your case. If you're running CancelFlow already, that's a sequencing change inside the flow you have, not a rebuild — and it's a cheap way to be ahead of a requirement that, on current form, other states are likely to look at next.

Frequently asked questions

Stop losing subscribers today

One script tag. One function call. A live cancellation flow in under 10 minutes.