> ## Documentation Index
> Fetch the complete documentation index at: https://docs.getswan.com/llms.txt
> Use this file to discover all available pages before exploring further.

# How sequences work

> What a sequence is made of, how each contact moves through it, and what the statuses mean.

A sequence is a set of steps that Swan runs against a list of contacts. Each step is one of four kinds:

* **Email** — sends an email message.
* **LinkedIn connection request** — sends a connection invite.
* **LinkedIn message** — sends a direct message (requires an existing connection).
* **Wait** — pauses for a fixed amount of time before continuing.

Each contact enrollment is a separate **sequence contact** (one row in Outreach → Sequences). The same person can have more than one sequence contact across different sequences or senders; analytics can count those enrollments separately from distinct people.

Steps run in the order you laid them out, one after another. Swan skips a step it can't run for a given contact — for example, a LinkedIn step for a contact with no LinkedIn URL, or a channel that's already failed for that contact (see below) — and moves straight to the next one it can run. If a step gets skipped, a wait step that immediately follows it is skipped too, so you don't end up with a contact stuck waiting on a step that never happened.

## Contacts move through independently

When you add contacts to a sequence, each one gets its own progress through the sequence — one contact being three steps in doesn't affect another contact who just started. Each contact has:

* A **status**: pending, running, waiting on limits, finished, stopped, or error.
* A **reason**, when relevant, explaining why it's in that status (for example, a reply was received, the sender was disconnected, or a connection request was rejected).
* Its own current step, sender, and message history.

Common reasons you'll see include the contact replying (finishes the sequence), a sequence being paused because the recipient was reached manually, a connection request expiring or being rejected, or the contact already being in progress elsewhere. See [Replies and stop conditions](replies-and-stop-conditions.md) for the full list of what moves a contact out of "running."

A contact can only be in one active sequence per sender at a time — Swan won't enroll the same contact twice under the same sender while it's still in progress. The same contact can be in different sequences run by different senders simultaneously.

## Email and LinkedIn fail independently

If a sequence combines email and LinkedIn steps, the two channels are tracked separately for each contact. If LinkedIn stops working for a contact — the connection request was rejected, expired, or LinkedIn wouldn't allow it — Swan skips the rest of that contact's LinkedIn steps but keeps running its email steps normally, and vice versa. A contact only ends up in an error state once every channel its sequence actually uses has failed; a single failed channel just narrows what's left to run, it doesn't stop the sequence.

## Sequence-level status

The sequence itself also has a status, separate from any individual contact's status. See these, and each contact's own status and reason, at [Outreach → Sequences](https://agent.getswan.com/outreach/sequences) — open a sequence to see its contacts and their individual progress, or use the Progress filter to narrow the list:

* **Awaiting approval** — the sequence hasn't been turned on yet.
* **Enabled** — running normally; contacts move through it on schedule.
* **Stopped** — turned off; no further steps run for its contacts.

## Email threading

The first email Swan sends to a contact in a sequence needs a subject line, since it starts a new thread. Any follow-up email step configured to send "as a reply" reuses that same thread and subject automatically — it doesn't need (or use) its own subject.

## Common questions

**Why do some contacts skip a step that others go through?**
If a contact is missing an email address or LinkedIn URL, or the assigned sender doesn't have that channel enabled, or that channel already failed earlier for this contact, Swan skips steps it can't run and moves on to the next one it can. See [Replies and stop conditions](replies-and-stop-conditions.md).

**Why are all my sequence contacts showing as Finished?**
A contact finishes when it completes every step it can run, replies, or hits one of the stop conditions described in [Replies and stop conditions](replies-and-stop-conditions.md). Check the status reason on individual contacts for the specific cause.

**Do I approve each step, or the whole sequence?**
The whole sequence. Approval happens once at the sequence level — approving unlocks every step for that contact, and there's no per-step approval. By default a sequence waits for your approval before the first message sends; you can tell Swan to send a specific sequence automatically. See [Approvals](approvals.md).

**Can the same contact be in two sequences at once?**
Not from the same sender while one is still active, but yes across different senders or sequences.

**What sender does a contact use if a sequence has more than one connected account?**
Every sequence sends through a single sender — one connected account, which can have both email and LinkedIn attached, covering both kinds of steps. See [Sending limits](sending-limits.md).