Nothing is being sent, and nothing is broken

Here is a failure with no error in it.

Your workflow triggers on schedule. It builds the message, calls the email platform, and gets a normal response. The run is green. Every run that day is green. And not one customer receives anything, because the account behind that integration has been suspended for sending, and the part of the platform your automation talks to is still perfectly happy to take instructions.

Mailchimp describes suspension as something its system does automatically once a send crosses industry thresholds — thresholds it deliberately does not publish, because they vary between mailbox providers and change over time. The triggers are the familiar ones: too many spam complaints, too many hard bounces, a spike in unsubscribes, an unauthenticated sending domain, direct complaints to providers.

The important detail for anyone with automation attached is what stays alive. Sending stops and list imports stop. Signup forms, API access and the rest of the account carry on working. Which means the check your integration would naturally perform — did the call succeed? — answers yes.

Why this is worse than a crash

A crash is loud, dated and specific. This is none of those things.

It does not appear in your logs. The decision was taken about your account, inside the platform, not inside your scenario. Your automation history shows a normal week.

It is not one message. Everything routed through that account stops at once: the campaign, but also the receipts, the appointment reminders, the password resets, the invoice chasers. Businesses tend to discover the blast radius during the incident rather than before it.

It ends by conversation, not by retry. Reinstatement means answering questions in the account and then waiting for a compliance person, in business hours, in their time zone. There is no button, and no amount of retry logic helps.

The warning went to one mailbox. Usually there is a warning first. It goes to the account contact address — which, on a platform set up years ago by someone who has since left, may be a mailbox nobody opens. This is the same weakness as not knowing whose accounts your automation runs on, seen from the other side.

Notice it in an hour, not in a week

The general rule: never treat "the platform accepted my request" as "the customer got the message". Those are different facts, and only the first one is cheap to check.

Watch the outcome, not the call. Delivered counts, opens on a transactional stream, replies — any number that only moves when a real message lands. A platform that has stopped sending produces a flat line in every one of them while your workflow reports success.

Set a floor alarm. If yesterday's delivered count is zero and it is a working day, something is wrong, and it does not matter what. This single check catches suspension, expired billing, a broken domain and a mistyped list ID with one rule.

Split transactional from marketing. Different accounts, ideally different vendors. Marketing is what gets a domain suspended; receipts and reminders are what customers actually need. Sharing one account means a marketing problem becomes an operations problem within minutes.

Make the platform contact address a mailbox two people read. Not the founder alone, not a person who left. The warning email is the one piece of advance notice you get.

Keep bounce and complaint rates in front of you monthly. They move slowly and then all at once, usually after an imported list. If you cannot say roughly what yours are, you cannot see the cliff coming.

The bit that hurts commercially

Suspension arrives at the worst time by construction: a big send is what triggers it, so the platform goes quiet immediately after you did the marketing that was meant to produce a week of work.

Then there is the second wave. Reminders stop, so people miss appointments. Invoice chasers stop, so payments slip. The order confirmations that told customers everything was fine stop, so your inbox fills with people asking whether their order exists. None of that shows up as an email problem when it lands on you — it shows up as a bad week.

What to do this week

Two hours of work, no project required.

List every automation that sends a message and mark which account it sends through. If a single account carries both a newsletter and anything a customer is waiting for, that is the risk in one line — and the fix is a second account.

Then add the floor alarm on delivered counts, and check who receives the platform's warnings. If you want that done as a review of everything sending on your behalf, with the list of what stops when each account stops, that is part of a process audit: $299, three business days.

Related: why automations fail silently, and how to make them speak up is about failures inside your workflow. This is the version where your workflow is fine and the channel has been switched off underneath it.