Texting rules by state: why the strictest one is the one to follow
Several states regulate business texting more tightly than federal law, and the state that counts is the recipient's. Why one strict rule beats tracking fifty, what it costs you, and the five settings that implement it. Not legal advice.
Sohail Safdar · · 6 min read
Federal law is the floor. Several states have built on top of it, usually with a narrower window for when you may contact somebody and a limit on how often, and the state that counts is generally the recipient's rather than yours.
That last point is the one that catches small businesses. A single-location salon is not outside all of this just because it has one address: if it texts a customer who lives over a state line, the customer's state may be the one that applies.
This is a summary of the shape of the problem, not legal advice, and the laws in this area change and generate litigation.
What state laws tend to add
A shorter window. 8am to 8pm in several states, against the federal 8am to 9pm. An hour does not sound like much until you notice it is the hour most businesses do their evening catch-up in.
Limits on attempts. A cap on how many times the same person may be contacted about the same subject in a day.
Broader definitions. What counts as an automated system, or as a solicitation, is wider in some state laws than in the federal one. This is where a business that believed itself exempt discovers otherwise.
A private right of action. Customers can sue directly in several states, which is why this area produces litigation rather than only regulatory letters. It is also why the practical advice is conservative: the downside is not a warning.
Most of these laws are aimed at marketing. Replies to a customer's own message, and reminders about an appointment they booked, are generally treated differently. "Generally" is carrying weight in that sentence, and the definitions are the part to check.
Why one strict rule beats a map
The alternative to a single strict setting is maintaining a per-state matrix, keeping it current as laws change, and knowing which state each customer is in. The third of those is the one that defeats it.
You usually do not know where a customer lives. An area code is a hint, not an address: people keep their mobile numbers when they move, so a Chicago number in Denver is entirely ordinary. A known service address is better evidence, and plenty of customers do not have one on file.
So a matrix built on area codes is a matrix built on a guess, and a guess in this area is wrong in both directions: it applies a rule to somebody it does not cover and misses somebody it does.
One strict rule has none of that machinery and costs you almost nothing. Nobody has ever complained about a text that arrived at 8:05 the next morning instead of at 8:30 the previous evening.
The five settings that implement it
- Quiet hours set to the narrowest window you reach. If any state you text into is 8pm, use 8pm everywhere. 9pm to 8am is a common default and 8pm to 8am is the strict version.
- A frequency cap at or below the lowest limit that applies to you, counting automated messages rather than conversation.
- Marketing only with express written consent, recorded with the wording, the date and the source, and kept.
- STOP honoured instantly and across the whole business, not per campaign and not per conversation.
- The recipient's timezone for the window, not the practice's. A 7:30pm send in California reaches Florida at 10:30pm, and the clock the law cares about is theirs.
Point 5 is the one most systems get wrong, including some that get points 1 to 4 right, because the business's own timezone is the one that is easy to know.
The frequency cap deserves its own paragraph
A cap feels irrelevant to a business that would never knowingly send six messages to one person in a day. That is exactly why it matters, because the way it happens is never deliberate.
A customer books, so a confirmation goes. They reply to move it, so a new confirmation goes. The 24-hour reminder fires for the old appointment before the automation catches up. A review request from last week's visit lands the same afternoon. A follow-up sequence sends its second step. Each of those is individually reasonable and nothing in any single rule can see the total.
A cap is the only thing that can, because it is the only rule that looks at the person rather than at the message.
One important exception: a cap must not apply to a reply in a live conversation. Capping an answer to a customer who is mid-conversation is the same category of mistake as applying quiet hours to it, and it is worse, because the customer is sitting there waiting.
When to get advice rather than a blog post
If you send marketing by text, if you text into many states, or if you use purchased or lead-generated lists, have counsel review the setup. The combination of a private right of action and statutory damages per message is why this is the one compliance area where a small business should not be improvising.
In Frontiva
Every outbound message goes through one check before it is sent, and the check stores its outcome on the message rather than leaving the decision to be re-derived later. That matters more than it sounds: re-deriving the answer next year gives you next year's answer, and consent may have been withdrawn since. The honest record is what was true at the time.
Quiet hours are a per-business window, 9pm to 8am by default, that can only ever move a message later (a shift that moved one earlier would be the exact harm the rule exists to prevent, so the database refuses it as well as the code).
And they narrow to the recipient's state where that state is stricter, which is the strictest-wins rule above implemented rather than recommended: Florida and Oklahoma are configured at 8pm against the default 9pm, and a recipient whose state cannot be determined gets the practice's own window unchanged rather than a guess. The window is then evaluated on the recipient's own clock, derived from their number.
A per-contact frequency cap applies to automated sends, six in a rolling day by default and configurable per business, and conversational replies are deliberately not counted.
Marketing by SMS is refused outright today. Not rate-limited, refused: there is no basis recorded for promotional texting, so the send check declines it and names the reason. Review requests are the one narrow exception and they run on their own separate opt-in, which is its own rule in the engine precisely because the fix for "no review consent" is to ask that customer, while the fix for "no marketing consent" is a decision the business has to make.
TCPA consent explained covers the federal floor, consent revocation covers what happens when somebody withdraws it, and the automations page describes the single send check.