How to Set Up Sequences in Front (2026): Outbound Automation, Step by Step

What a Front sequence actually is

A sequence is an automated, multi-stage outbound email flow sent to a list of people from one of your channels. Per Front's guide to understanding sequences, a sequence can run 1 to 10 stages (labelled A through J), and each stage carries a sending delay that only fires if the recipient hasn't replied within a set number of days. That last clause is the whole point: a sequence keeps following up on your behalf, and the moment someone answers, it gets out of the way.

Two things are worth setting expectations on before you build. First, sequences are a Professional-plan-or-above feature — they are not on Front's entry tier. Second, a sequence sends from a real email channel, not from the shared workspace abstractly: a personal sequence sends from an individual inbox that contains an email channel, and a workspace sequence sends from a shared inbox that is an email channel. If no such channel is connected, the sending dialog has nothing to send from — which is exactly the state the screenshot below captures.

Step 1 — Create the sequence and pick a From channel

Start from the Sequences area, not from a conversation.

1. Open Sequences in Front's main navigation menu.
2. In the left sidebar, choose the scope you want: Personal (only you send and see it) or a Workspace name (shared with the team).
3. Click Create sequence to open the New sequence dialog.
4. Fill in the required fields: a Name (internal only — recipients never see it), the From channel you'll send through, and optionally a Sender name, plus a Cc and Bcc (each capped at a single address).

The From picker is the gate. It only lists email channels you're actually allowed to send from, so on a workspace that has no personal sending channel connected, opening it returns No results and the dialog's confirm button stays disabled — you cannot advance to the stage builder until a channel is chosen.

That "No results" state isn't a bug — it's Front telling you the prerequisite hasn't been met. Connect a personal email channel (or run the sequence from a shared inbox that is an email channel) and the picker fills in, unlocking everything that follows.

Step 2 — Add your recipients (manual or CSV)

Once the channel is set, you load the people the sequence will reach. Front gives you two ways:

• Type them in. Enter email addresses or contact lists directly; pressing Enter/Return starts a new line.
• Upload a CSV. This is the real workhorse for outbound. Front requires the first column to be email, accepts up to 200 addresses per CSV file (you can upload several files if your list is larger), and supports up to 20 custom variables/columns — the extra columns become the personalization tokens you'll use in Step 4.

The CSV limit is the design decision to plan around. A 200-address cap per file keeps a single sequence deliberately small and human-scaled — this is Front's collaborative-inbox lineage showing, not a bulk cold-email cannon. For a sales team running a focused prospect list or a CS team onboarding a cohort, that's plenty; for a marketing blast to tens of thousands, it isn't the right tool.

Step 3 — Build the stages and set per-stage delays

Now the multi-step part. Each stage is one email in the follow-up chain, and the delay between stages is what makes the whole thing feel patient rather than pushy.

1. Click Next: Stage A to open the first stage, then set its subject line and draft the body.
2. Set the stage's delivery timing — Stage A typically goes out immediately (or at a scheduled time), and every later stage carries a delay measured in days.
3. Click Add stage to append Stage B, C, and so on, up to the 10-stage maximum. Repeat the subject/body/delay pattern for each.
4. When the whole chain is built, click Start sequence; Front shows a confirmation prompt before anything actually sends.

The delay is conditional, which is the elegant bit: a later stage only fires if the recipient hasn't replied yet. So a three-stage sequence with 3-day delays quietly becomes a one-email sequence for anyone who answers Stage A. Behind the scenes Front also throttles sends to protect deliverability — roughly 7.5 seconds between Gmail or Office 365 messages and about 1 second for other channel types — so a 200-person stage trickles out rather than hammering the mail server in one burst.

Step 4 — Personalization variables

A follow-up that opens with "Hi there" is fine; one that opens with the person's real first name, company, and the thing they actually signed up for lands far better. Front handles this with variables, documented in its guide to variables in templates.

• Type {{ anywhere in the subject or body to open the variable dropdown.
• For fields you uploaded in your CSV, use the {{sequence...}} variables — these pull from your spreadsheet columns.
• For data stored on a saved Front contact, use {{recipient...}} variables (for example {{recipient.first_name}} or a custom field via {{recipient.custom.field_name}}).
• Guard against blanks with a fallback: {{recipient.first_name | 'there'}} renders "there" whenever the first name is missing, so no message ever ships with an empty gap where a name should be.

The {{sequence...}} versus {{recipient...}} distinction is the single most common thing people get wrong. If your data came from the CSV, the {{recipient...}} tokens will read as blank because there's no saved contact behind them. Match the token family to the data source and personalization "just works."

Step 5 — Stop-on-reply (and the analytics-inflation trap)

Stop-on-reply is on by default, and it's the feature that keeps a sequence from embarrassing you. Per Front, once a recipient sends a real reply, the sequence automatically stops for that recipient — no later stage will send. Crucially, auto-replies such as "Out of office" do not count as a reply and will not stop the sequence, so a vacation responder won't accidentally kill your follow-up chain.

Here's where honest reporting matters. Because stop-on-reply hinges on Front detecting an inbound reply, your outbound metrics only stay clean if the replies actually route back to the same conversation. Two things quietly inflate the numbers:

• Replies that land somewhere else. If a prospect answers from a different address, or their reply gets routed to another inbox by a rule, Front may not register it as a stop — so the next stage sends to someone who already responded, and your "delivered" and "sent" counts overstate real outreach.
• Bounces and auto-replies read as activity. OOO messages, mailer-daemon bounces, and auto-acknowledgements aren't engagement, but if you eyeball raw send/open counts without filtering them, a sequence looks busier and healthier than it is.

The fix is to judge sequences on reply rate and true positive replies, not on stages sent or opens. Front's own reporting can help you separate the two — see Front analytics explained for what its metrics do and don't capture — but the discipline is yours: a sequence that "sent 600 emails" told you nothing until you know how many were real conversations.

The honest limits — and where an AI layer picks up

Front sequences are a genuinely good, native way to automate patient, personalized follow-up — and for a support, sales, or ops team already living in Front, not having to bolt on a separate outbound tool is a real win. But it's fair to name the ceiling.

The gating is real: sequences need the Professional plan or above, and every sequence needs a connected personal or shared email channel to send from — without one, you get the "No results" From picker in the screenshot above and can't even reach the builder. The scale is deliberately modest — 200 addresses per CSV, 10 stages max — which is right for focused outreach and wrong for mass campaigns. And the deeper limit is what a sequence fundamentally is: a pre-written script. It sends the words you wrote, in the order you wrote them, to whoever's on the list. It can personalize a {{first_name}}, but it can't read the reply that comes back, understand what the person is actually asking, check their account or order, and answer them. The moment a recipient replies, the automation's job is done and a human's begins.

That seam is where an AI agent layer fits — not to replace sequences, but to handle the reasoning-heavy part they structurally can't. The category of AI agents for customer service exists for exactly the work a merge field can't do. Macha is one such layer: it runs on top of the Front you already use through the live Macha–Front connector — it does not replace Front, your inboxes, or your sequences. Your sequence keeps doing what it's good at: sending timely, personalized outbound and stopping the instant someone replies. Then Macha's agent reads that reply, understands intent rather than keywords, and drafts or sends a grounded response — pulling a real order, account, or renewal status through a custom tool that turns your REST API into something the agent can call. A sequence starts the conversation; the agent finishes it. If you want the mechanics, connecting Front to Macha to route conversations to AI walks through it, and Macha's credits are consumed per AI action, never per reply — sending and reasoning have different costs, and it's honest to price them that way.

For the concepts underneath this how-to, see Front sequences explained and Front message templates explained; for what unlocks on which tier, Front pricing explained.

FAQ

How many stages can a Front sequence have? Between 1 and 10 stages, labelled A through J. Each stage after the first carries a sending delay measured in days, and that delay only fires if the recipient hasn't already replied.

What plan do I need for sequences in Front? Sequences require the Professional plan or above. You also need a connected email channel to send from — a personal sequence sends from an individual inbox with an email channel, and a workspace sequence from a shared inbox that is an email channel. Without one, the From picker returns "No results" and you can't open the stage builder.

Why are my personalization variables showing up blank? Almost always a mismatch between the variable family and the data source. Use {{sequence...}} variables for data you uploaded in a CSV and {{recipient...}} variables for fields on a saved Front contact. Add a fallback like {{recipient.first_name | 'there'}} so a missing value never ships an empty gap.

Does a sequence stop if someone replies? Yes. When a recipient sends a genuine reply, Front automatically stops the remaining stages for that person. Auto-replies such as "Out of office" messages do not count as a reply and will not stop the sequence.

How many recipients can I add to a sequence? Up to 200 addresses per CSV file, with the first column required to be email, and up to 20 custom variables/columns. You can upload multiple CSV files if your list is larger, but the 200-per-file cap keeps sequences scaled for focused outreach rather than mass blasts.

Can I add AI to Front sequences without replacing Front? Yes. An AI agent layer like Macha connects to Front as a live connector and runs on top of your existing inboxes and sequences — it doesn't replace them. Your sequence keeps sending personalized outbound and stopping on reply; the agent then reads the replies that come back, understands intent, and drafts or sends a grounded response.

Ready to turn "they replied" into "it's handled"? Start a free trial of Macha and connect it to your Front in minutes.