Sequences
Sequences are automated multi-step outreach cadences. You define the steps, channels, and timing — ScendCore’s AI handles the execution, drafting personalized messages for each contact at each step.
How Sequences Work
A sequence is a series of ordered steps, each with a configurable delay. When you enroll a contact, ScendCore works through the steps automatically:
- The first step fires after its configured delay from enrollment
- Each subsequent step fires after its delay from the previous step
- The AI drafts personalized content for each step based on the contact’s profile, your agent’s persona, and any messaging overrides you set
Sequences can be paused, resumed, or archived at any time.
Creating a Sequence
- Go to Sequences in the left sidebar
- Click + Create Sequence
- Enter a name and optional description
- Click Create
Using Templates
If you do not have any sequences yet, ScendCore offers built-in templates to get you started:
| Template | Steps | Description |
|---|---|---|
| Cold Outreach | 3 steps | Intro, value prop, CTA cadence over 10 days |
| Follow-Up Series | 2 steps | Check-in and final nudge over 7 days |
| Re-Engagement | 2 steps | Sequence for contacts gone cold |
| Meeting Request | 1 step | Direct ask for high-intent leads |
| Voicemail + Email Combo | 3 steps | Email intro, voicemail drop day 3, follow-up email day 8 |
Click a template to create a sequence pre-configured with the right number of steps and delays.
Configuring Steps
Each step in a sequence has:
- Step number — The order in the sequence (auto-assigned)
- Job type — What kind of outreach this step performs:
- Initial Outreach — First contact message
- Follow-up Email — Continuation of a thread
- Re-engagement — Reaching back out to cold contacts
- Voicemail Drop — AI-generated voicemail left on the contact’s phone
- Channel — The communication channel for this step
- Delay — How long to wait before this step fires
Setting Delays
Delays can be configured in minutes, hours, or days:
- Minutes — For fast sequences (e.g., 30 minutes between steps)
- Hours — For same-day follow-ups (e.g., 4 hours later)
- Days — For standard cadences (e.g., 3 days between touches)
The delay on the first step is measured from the moment of enrollment. Delays on subsequent steps are measured from the previous step.
Tip: For testing, set all steps to 0 minutes delay. ScendCore’s sequence processor runs every 2 minutes, so steps will fire within minutes of enrollment.
Channels
Each step can use any of the following channels:
| Channel | Description | Prerequisites |
|---|---|---|
| Personalized email via sender profile | Sender profile configured | |
| SMS | Text message via Twilio | Twilio number configured in Settings → Phone Numbers |
| Voice | Phone call or voicemail drop | Twilio number + contact must have a phone number |
| WhatsApp message via Twilio | WhatsApp sender configured in Settings → Channels |
Voicemail Drop
Voicemail Drop is a specialized step type that leaves an AI-generated voicemail on a contact’s phone. When a sequence step uses the Voicemail Drop job type, the channel is automatically set to Voice.
Here is how it works:
- The AI generates a short spoken script personalized to the contact
- Twilio places a real phone call to the contact’s number
- Answering Machine Detection determines whether a human or voicemail answered
- If voicemail is detected, the AI message plays after the beep and the call ends cleanly
- The transcript appears in the contact’s conversation timeline like any other message
Prerequisites:
- The contact must have a phone number on file
- A Twilio number must be configured under Settings → Phone Numbers
Tip: Voicemail drop is most effective as a mid-sequence touchpoint. The recommended pattern is: email on day 0, voicemail drop on day 3, follow-up email on day 8. The “Voicemail + Email Combo” template uses this exact cadence.
Messaging Overrides
Each step supports optional messaging overrides to guide the AI:
- Angle — The approach or hook for this step
- CTA — The call-to-action to include
- Value proposition — The key value message to emphasize
These are hints for the AI, not literal copy. The AI weaves them into a personalized, natural-sounding message.
Sequence Settings
Each sequence has configurable behaviors:
- Stop on reply — Automatically pause the sequence for a contact when they reply
- Stop on meeting — Pause when a meeting is booked
- Sender profile — Which sender identity to use for this sequence
- Agent — Which AI agent drafts the messages
- Default sequence — Mark one sequence as the default for quick enrollment
Sequence Status
Sequences have four statuses:
| Status | Meaning |
|---|---|
| Draft | Being configured, not yet running |
| Active | Running and processing enrolled contacts |
| Paused | Temporarily stopped, can be resumed |
| Archived | Permanently stopped, kept for reference |
Enrolling Contacts
There are several ways to enroll contacts in a sequence:
Individual Enrollment
- Open a contact’s profile
- Select a sequence to enroll them in
- The first step queues immediately (or after its configured delay)
Bulk Enrollment
- Go to Contacts and select multiple contacts using checkboxes
- Click Enroll in Sequence from the bulk actions bar
- Choose the target sequence
From a List
- Create a contact list in the Lists section
- Open the list and click Enroll All in Sequence
- Select the target sequence
ScendCore checks for duplicate enrollments — a contact will not be enrolled in the same sequence twice while they have an active enrollment.
Default Sequence (Catch-All)
You can mark one active sequence as the default by toggling the Default option in sequence settings. When a contact enters the system — via import, form submission, or API — without a specific campaign or sequence assignment, ScendCore automatically enrolls them in the default sequence.
Only one sequence can be the default at a time. Setting a new default removes the flag from the previous one.
Tip: Use the default sequence as a safety net. Even if a contact slips through without a campaign assignment, they will receive your baseline outreach automatically.
Viewing Sequence Progress
Each sequence card shows real-time stats:
- Active enrollments — Contacts currently progressing through steps
- Completed — Contacts who finished all steps
- Unenrolled — Contacts removed before completing
Click a sequence to see the full enrollment table with per-contact status, current step, and timestamps.
Enrollment Statuses
| Status | Meaning |
|---|---|
| Active | Currently progressing through the sequence |
| Completed | Finished all steps |
| Paused | Temporarily paused (e.g., due to a reply) |
| Unenrolled | Removed from the sequence before completion |
How Execution Works
Behind the scenes, ScendCore’s queue worker processes sequence steps:
- The sequence cron runs every 2 minutes
- It checks all active enrollments for steps that are due (based on delays)
- For each due step, it creates an agent job assigned to the sequence’s AI agent
- The agent drafts personalized content for the contact
- Depending on the agent’s autonomy level, the draft either auto-sends or goes to the approval queue for human review
- After the step completes, the enrollment advances to the next step
If Stop on reply or Stop on meeting is enabled and the corresponding event occurs, the enrollment is automatically paused.
Visual Sequence Builder
In addition to the step-by-step editor described above, ScendCore offers a visual drag-and-drop builder for creating sequences with branching, conditions, and AI-powered routing. The Visual Builder lets you design complex multi-path flows on an interactive canvas, including:
- Condition nodes that route contacts down different paths based on field values
- AI Decide and AI Qualify nodes that let the AI choose the best path for each contact
- A/B Split nodes for testing different approaches
- Notify Rep nodes to alert your team at key moments
You can also use Quick Mode as a bridge between the simple step editor and the full canvas — it provides a form-based interface for building linear sequences that can be promoted to the visual canvas when you need branching.
For the full guide, see the Visual Sequence Builder documentation.