Keyword auto-response best practices
Keyword auto-responses handle clear, single-word commands reliably—but they have a fundamental limitation that affects your compliance exposure.
Legal disclaimerThis documentation describes how to configure keyword auto-responses in Bloomreach. It isn't legal advice. Bloomreach provides tools to help you capture and process opt-out requests. It's your responsibility to configure them in a way that complies with applicable laws and regulations, including TCPA, CTIA guidelines, and any other requirements relevant to your business and jurisdiction. Work with your own Legal and Compliance team to determine what opt-out language, keywords, and processes you need to support.
Keyword auto-responses only match when the keyword appears at the start of the message. Phrases like "please stop sending me texts" or "take me off your list" won't trigger an opt-out—even though they clearly signal one. Under US TCPA/FCC guidance, brands are expected to honor reasonable opt-out language, not just the exact keyword "STOP."
This guide explains how to combine keyword auto-responses with scenarios to build more robust opt-out coverage and set the right expectations with your customers.
For technical setup, see Keyword auto-responses.
Auto responses vs. scenario-based
| Auto-keyword responses | Scenario based | |
|---|---|---|
| Trigger speed | Immediate | 1 minute or more |
| Works for unknown numbers | Yes | No—existing customers only |
| Can create new profiles | Yes (grant consent) | No |
| Handles natural language opt-outs | No | Yes |
| Case-sensitive matching | No | Yes—requires adding each variant separately |
Understand the gap
Keyword auto-responses work well for:
- Customers who follow your instructions exactly—for example, "Reply STOP to opt out."
- Clear, high-intent signals at the start of a message.
They won't catch:
- Natural language opt-outs that don't start with a keyword—for example, "can you please stop texting me."
- Messages where the keyword appears mid-sentence — for example, "I want to STOP getting these."
- Variations you haven't explicitly configured.
Relying on STOP alone—or even a broad keyword list—won't catch all real-world opt-out language. Combining keyword auto-responses with scenarios gives you better coverage.
Build a safer opt-out setup
Expand your keyword list
Start with the standard opt-out keywords and add natural-language variants your customers are likely to use. The list below is a starting point. It's not exhaustive, and it's not legal advice.
Suggested opt-out keywords (examples only): STOP, QUIT, UNSUBSCRIBE, CANCEL, END, ARRET, BASTA, STOP ALL, PLEASE STOP, REMOVE, OPT OUT, NOT INTERESTED
Work with your Legal and Compliance team to define the full set of words and phrases your business needs to treat as opt-outs, and configure them as keyword configurations in Bloomreach.
Use scenarios to catch what keywords miss
Scenarios let you react to inbound messages that contain opt-out language but don't start with a keyword. Set up a scenario that triggers on inbound SMS messages and checks whether the message body contains opt-out language—even mid-sentence. The scenario has four steps. Each channel (SMS, MMS, WhatsApp) must be configured separately.
Step 1: Add a trigger for each keyword variant
Add one On event node per keyword variant you want to catch. Set the delay to 0 seconds.
Configure each trigger with these conditions:
-
campaign→action_typeequalssms -
ANDstatus equalsincoming_message -
ANDmessage containsyour keyword variant
Repeat for every variant you want to catch. For example, to catch "stop" in any form, add three separate triggers: one for stop, one for Stop, one for STOP. Do the same for each additional phrase—unsubscribe, Unsubscribe, UNSUBSCRIBE, please stop, and so on.
Case sensitivity in scenariosUnlike keyword auto-responses, scenarios are case-sensitive. Every capitalisation variation must be added as a separate trigger—"stop", "Stop", and "STOP" are treated as three different values. This isn't ideal, but it's the only way to configure this in scenarios.
Step 2: Add a validity check
Add a condition node to check that the customer has SMS Marketing consent. This prevents the scenario from processing opt-outs for customers who are already unsubscribed.
Configure the customer filter: SMS Marketing is true.
Connect all your trigger nodes to this single condition node. Only customers who match (consent = true) will proceed to the next step.
Step 3: Revoke consent
Add an Add event node on the Yes branch of the condition.
Configure it to track a consent rejection:
-
Event type:
consent -
category= your SMS marketing consent label (for example, Marketing) -
action=reject
This is the step that actually unsubscribes the customer in Bloomreach. Without this node, the scenario won't process the opt-out.
Step 4: Send a confirmation message
Add an SMS node after the unsubscribe event to send a confirmation.
ImportantSet the sender consent to General consent for this message—not SMS Marketing. The customer no longer has SMS Marketing consent at this point, so using it as the consent category will prevent the confirmation from sending.
Example confirmation message:
"Sorry to see you go! We'll unsubscribe you from all future marketing. If we got it wrong, reply JOIN at any time to opt back in to marketing texts."
This is a best practice for coverage—not a legal guarantee. Work with your legal and compliance team to define which phrases to include and how to handle edge cases.
Send a catch-all response for unrecognized messages
Set up a scenario that triggers when an inbound message doesn't match any keyword. Use it to guide customers to clear unsubscribe instructions.
Example auto-response:
"We received your message but couldn't process it automatically. To unsubscribe, reply STOP. For help, visit Support portal or call [phone number]."
This helps customers who tried to opt out but used phrasing that your keywords didn't catch, and gives them a clear path forward.
Confirm every opt-out
Always send a confirmation when someone unsubscribes—whether via keyword or scenario.
Example auto-response:
"You've been unsubscribed and won't receive further messages from us. If this was a mistake, reply JOIN to re-subscribe."
Recommended configuration
| Layer | What it covers | How to configure |
|---|---|---|
| Keyword auto-responses | Clear commands at the start of a message | Configure STOP, QUIT, UNSUBSCRIBE, and variants as opt-out keywords |
| Scenarios—opt-out detection | Natural language opt-out phrases that don't start with a keyword | Trigger on inbound message contains opt-out phrases |
| Scenarios—catch-all | Unrecognized messages that may be opt-out attempts | Trigger when inbound message doesn't match any keyword; send STOP instructions |
Use keyword data in scenarios
Once keyword events are tracked, use them to trigger further automation.
Re-engage customers who opted out
-
Trigger: Customer sends STOP.
-
Wait: 30 days.
-
Condition: Customer has email consent.
-
Action: Send a one-time re-engagement email with an SMS opt-in link.
Support high-intent customers
-
Trigger: Customer sends HELP or INFO.
-
Condition: Customer has made a recent purchase.
-
Action: Add to VIP segment and send early access to sales.
Maintain engagement across channels
-
Trigger: Customer opts out of SMS.
-
Condition: Customer has email consent.
-
Action: Increase email frequency to maintain engagement.
Proceed with caution. Re-engaging opted-out customers—even via a different channel—can create legal risk if it appears designed to circumvent their opt-out. Make sure any re-engagement complies with CAN-SPAM (for email) and doesn't pressure the customer to re-subscribe. Your legal team should approve any re-engagement workflow before launch.
International considerations
Keyword lists need to reflect the language and carrier requirements of each market you operate in.
-
Germany: "STOPP" is more common than "STOP."
-
France: Customers may use "arrêt," "arret," or "ARRET." If you add ARRET as a keyword, the system will automatically match accented variants like "arrêt" too—accent matching is handled automatically for keyword auto-responses. In scenarios, add each variant explicitly since scenarios are case- and accent-sensitive.
-
Other markets: Research local opt-out conventions and carrier-specific requirements before launch.
Add regional variants as separate keyword configurations and apply the same scenario-based approach in each language where you operate.
Set the right expectations
-
Keyword auto-responses won't catch everything. However, you handle it—automated scenarios, manual review, or both—monitoring for opt-out language that does.
-
Your keyword and phrase list isn't static. Customer language evolves. Review and update your opt-out keywords and scenario conditions regularly, in consultation with your legal team.
-
Opt-out processing isn't always instant. There's a small delay between when a number is blocked and when consent events are tracked. Build buffer time into your campaign scheduling where possible.
-
Bloomreach provides the tools—configuration is your responsibility. The right setup depends on your business, your markets, and your legal obligations. Your legal and compliance team determines what counts as a reasonable opt-out for your use case.
Updated about 1 hour ago
