Suppression and blocking
Suppression is a fact about one address or number: it will not be written to, and the refusal happens at the boundary where messages leave. A block is an instruction about behaviour: stop pursuing this company, this topic, this line of work. One is enforced by the send path, the other by the reasoning. Using the wrong one is how a gap appears.
Two refusals, side by side#
| Suppression | Block | |
|---|---|---|
| What it applies to | One identity — an address, a number | A relationship, a subject or a kind of action |
| Where it is enforced | At the send boundary, every time | In how Connect reasons about what to do |
| Where it comes from | A bounce, an unsubscribe, a STOP, or a person recording it | A person writing it down as a durable instruction |
| What it stops | The message going out on that identity | The intention forming in the first place |
| Reversibility | Varies by origin — an unsubscribe is not somebody else's to undo | Editable like any other standing instruction |
| What it does not do | Stop Connect thinking about that person | Stop a message to an address you have not suppressed |
The last row is the practical heart of it. A block that says *leave this account alone* does not put a stop at the send boundary; a suppression on one address does not stop Connect drafting to a colleague at the same company. A situation that needs both should have both.
Where suppression comes from#
Suppression is usually not a decision somebody sat down to make. It arrives, and the origin is part of the record because the origin decides what may happen next.
- An unsubscribe
- The recipient asked. That is theirs to reverse, not yours, and treating it as an internal setting is the mistake worth avoiding.
- A STOP on messaging
- The same thing, arriving over a different channel. It is honoured on the identity that sent it.
- A hard delivery failure
- The address does not work. Continuing to send at it damages the sending reputation of every mailbox that shares it.
- A do-not-contact entry
- Recorded deliberately, usually because something went wrong or somebody asked in person. It is not cleared casually, and the Connect Assistant cannot clear one at all.
What a block is good at#
A block is a durable instruction, so it works on intentions rather than on envelopes. It is the right tool when the thing you want stopped is not *messages to an address* but *work of a kind*.
- A competitor you would rather not be researched. No address to suppress; a clear instruction to record.
- A subject Connect should not answer on. The boundary belongs in the rules, not at the outbound edge.
- An account in a delicate phase, where you want no initiative at all until somebody says otherwise.
- A pattern you keep correcting. If you find yourself rejecting the same draft repeatedly, the correction belongs as an instruction rather than as a series of decisions about single messages.
What a block is bad at is guaranteeing silence. It shapes behaviour, and behaviour has more paths through it than a send boundary does. When silence on a specific address is the requirement, suppression is what enforces it.
Choosing correctly the first time#
Ask whether you are protecting a person or a policy.
Result A person who asked not to be written to is a suppression. A decision about how the business behaves is an instruction.
If a recipient asked, record it as a suppression even when you also write a rule.
Result The request is honoured at the point it can be enforced, rather than depending on Connect's reasoning being right every time.
For an account you want left alone entirely, use both.
Result The instruction stops the initiative and the suppression stops the message, and neither has to be perfect on its own.
Review suppressions with their origins in view rather than as a list of addresses.
Result A bounce that has since been fixed is a different case from an unsubscribe, and only one of them is yours to reconsider.
If a send was refused and you are not sure which of the two caused it, a recipient is suppressed shows the origin along with the refusal.
Questions#
Does suppression remove the contact?
No. The person, their history and their records all remain — what changes is that the suppressed identity is not written to. That is the correct outcome: deleting somebody who asked not to be contacted would also delete the evidence that they asked, and the next import would recreate them.
Can I suppress a whole company?
Suppression works on identities, so a company with several addresses needs each one suppressed. When the intent is *this organisation, entirely*, pair the suppressions with an instruction so that Connect also stops forming plans involving them rather than only stopping at the last moment.
Does an import undo either of these?
No. Suppression and do-not-contact entries survive an import, which is what stops a refreshed spreadsheet quietly re-enabling contact with somebody who opted out. A standing instruction is unaffected by an import too — it is about behaviour, not about rows.