Back to all modules
Marketing

Campaigns

Multi-step outbound programs across email and SMS, with scheduling, audience rules and goals that stop the sequence.

On this page
What it isWhat you can configureHow to set it upWhere it fits in a buildLimits and gotchas
Want this configured for you?

Bring your stack to a call and we will set this module up against how your business actually runs.

Book a call
01

What it is

A campaign is a planned sequence of outbound messages sent to a defined audience over time. You choose who receives it, what each step says, how long the gaps are, which channel each step uses, and what makes someone stop receiving it. Then it runs, either immediately to a list or on a schedule, and reports on what happened.

The distinction worth holding is between a campaign and a workflow. A workflow is triggered by something a contact did and is built around that individual's journey. A campaign is initiated by you, on an audience you chose, around something the business wants to say. In practice they overlap heavily and campaigns are often implemented as workflows under the hood, but the intent differs: workflows react, campaigns broadcast on a cadence.

Treat a campaign as a program with a purpose and an end, not a tap that stays open. Every campaign should have a stated goal, an audience definition you can defend, an exit condition that fires the moment the goal is met, and a date it stops. The failure mode at every level of sophistication is the same: campaigns that keep running, to audiences nobody re-checked, saying things that are no longer true, with no exit for people who already converted.

02

What you can configure

Audience

  • The source list, smart list or segment the campaign sends to
  • Inclusion and exclusion rules, including excluding anyone in another active campaign
  • Consent and channel eligibility: who has agreed to email, who has agreed to texts
  • Whether the audience is a snapshot taken at launch or re-evaluated as the campaign runs

Steps and channels

  • An ordered series of steps, each an email or an SMS
  • The gap before each step, in hours or days
  • Per-step conditions so a step only sends to part of the audience
  • Branching by engagement, for example a different follow-up for those who opened

Scheduling

  • Immediate send, scheduled start, or recurring
  • Send windows and allowed days, so nothing arrives overnight or on a Sunday
  • Timezone basis: the account's timezone or each contact's own
  • Throttling, so a large audience is sent in batches rather than all at once

Goals and exits

  • The conversion event that removes a contact from the campaign: booked, purchased, replied, clicked, stage reached
  • Unsubscribe and opt-out handling per channel
  • A hard end date after which no further steps send

Content and testing

  • Templates per step, with personalization and fallbacks
  • Subject line and content variants for testing, and how the winner is chosen
  • Tracked links and campaign-level tracking parameters for attribution

Reporting

  • Per-step delivery, open, click, reply and unsubscribe figures
  • Goal completions attributed to the campaign
  • Bounce and failure detail by channel
03

How to set it up

  1. Write the goal as one sentence with a measurable outcome, and agree it with the client before building anything.
  2. Define the audience as a saved segment rather than an exported list, so you can see and re-check who is in it.
  3. Check consent and eligibility for the channels you intend to use. Email consent is not SMS consent, and an old list is the fastest route to complaints and deliverability damage.
  4. Confirm the sending infrastructure is healthy before the first send: an authenticated sending domain, a warmed sending reputation, and a messaging number with any required registration complete.
  5. Plan the sequence on paper: how many steps, which channel each uses, the gap between them, and what each one is for. Three well-spaced steps beat seven.
  6. Build the content for each step, keeping SMS short and making sure each message stands alone for someone who missed the previous one.
  7. Set personalization with fallbacks, and verify every merge field and every link against a real test contact.
  8. Set the exit conditions. The goal event must remove the contact immediately, and this is the step most often skipped.
  9. Set the exclusion rules, including anyone already in another active campaign, anyone who converted recently, and anyone the client has flagged as sensitive.
  10. Configure the schedule: start, send windows, allowed days, and throttling for the audience size.
  11. Set the campaign's hard end date.
  12. Send the whole sequence to a small internal test audience and read every message on a phone and in a desktop inbox, checking rendering, length and unsubscribe handling.
  13. Launch to a small slice of the real audience first, check delivery and bounce figures and read the first replies, then release the rest.
  14. Watch the first forty-eight hours for bounces, unsubscribes, complaints and replies, and be prepared to pause. The ability to stop a campaign quickly is part of the build.
  15. At the end, record what the campaign produced against the stated goal, and archive it rather than leaving it live.
04

Where it fits in a build

Campaigns consume audiences from Contacts and Smart Lists, send through Email and SMS, hand replies to Conversations and the AI agent, fire exits off Pipeline and Appointment events, and report through Reporting and Attribution.

Promotion to an existing customer base

A segment of past customers who have not purchased recently receives a three-step campaign: email, then SMS two days later, then a final email. Anyone who books or buys exits immediately. Replies are triaged by an AI step in a workflow so the team only handles real conversations. The campaign has an end date, so it cannot keep running into next quarter.

Event or webinar sequence

Registrants from a form enter a campaign that confirms, reminds the day before, reminds an hour before by SMS, and follows up afterwards with a split between those who attended and those who did not. Attendance status comes from the appointment or a tag set after the event.

Long-horizon nurture with a hand-off

Leads who were not ready enter a low-frequency campaign of useful content. The exit condition is any reply or any booking, at which point a workflow assigns an owner and creates a task. The campaign's job is to stay present; the workflow's job is to act the moment the lead moves.

05

Limits and gotchas

  • No exit condition is the cardinal sin. Continuing to chase someone who already bought or booked is the most damaging thing a campaign can do, and the client always hears about it.
  • Consent is per channel and it expires in practice. A list collected for email does not permit texting, and a list nobody has contacted in two years will generate complaints regardless of what was agreed originally.
  • Deliverability is earned before the campaign, not during it. Sending a large volume from a cold domain or an unwarmed number damages reputation for every later send, including transactional ones.
  • Snapshot versus live audiences behave very differently. A re-evaluated audience can pull in contacts mid-sequence who then receive step four as their first message. Know which mode you are in.
  • Overlapping campaigns stack. Two campaigns and a workflow can all message the same contact in one morning. Exclusion rules across active programs are not optional at scale.
  • SMS regulation is real. Registration requirements, opt-out keyword handling and quiet-hours rules apply, and the penalties fall on the sender. Treat SMS as the channel with the lowest tolerance for error.
  • Timezone confusion sends messages overnight. If the send window uses the account timezone and the audience is spread, some people get texts at four in the morning.
  • Throttling matters more than it seems. An unthrottled large send can hit provider rate limits, produce partial delivery and leave you unsure who received what.
  • Open rates are unreliable as a measure. Privacy features inflate and suppress opens unevenly. Judge campaigns on replies, clicks and goal completions.
  • Bounces need handling, not just reporting. Hard bounces should be suppressed automatically; left alone they degrade sending reputation every time you send.
  • Campaigns outlive their offers. A recurring campaign referencing a promotion that ended is a quiet, ongoing embarrassment. End dates and review dates prevent it.

Configure it with us.

Every module on this site is something we set up, test and hand over with the SOPs your team needs to run it.

Book a call