Sequencer

The Sequencer automates multi-step outreach across every channel. Define a series of steps — emails, LinkedIn actions, phone calls, SMS, WhatsApp — and graph8 executes them on a schedule for each contact in your list.

Channels

Sequences support multiple channels, each with its own resource allocation:

Channel	What It Does
Email	Send via Gmail or SMTP mailboxes
LinkedIn	Connection request, direct message, InMail, follow, view profile, like post (via HeyReach)
Phone	Voice AI calls or manual dialer
SMS	Text messages
WhatsApp	WhatsApp Business messages

Each channel requires an allocated resource — a mailbox for email, a LinkedIn account for LinkedIn actions, a phone number for calls/SMS, etc.

Building a Sequence

1. Go to Engage → Sequencer and click Create Sequence
2. Name your sequence and select a contact list
3. Add steps — each step has a channel, timing, and content
4. Assign resources (mailboxes, phone numbers, LinkedIn accounts)
5. Set a schedule for sending windows

Steps

Each step defines one action in the sequence:

Step Type	Channel	What It Does
Email	Email	Send a personalized email
Phone	Phone	Trigger a voice AI call
Manual Dialer	Phone	Queue contact for manual dialing
LinkedIn	LinkedIn	Execute a LinkedIn action (connect, message, InMail, follow, view, like)
SMS	SMS	Send a text message
WhatsApp	WhatsApp	Send a WhatsApp Business message

For each step, choose how the content is created:

• On-demand — write the content yourself
• AI-generated template — let AI generate the content
• Manual template — use a saved template

Set the time interval between steps (e.g., wait 2 days after the previous step before executing).

LinkedIn Actions

LinkedIn steps support multiple action types:

• Connection request - send a connection request with an optional message and fallback message
• Message - send a direct message to an existing connection
• InMail - send an InMail with subject and body
• Follow - follow the contact’s profile
• View profile - view the contact’s profile (creates a notification)
• Like post - like a recent post with a selectable reaction type

LinkedIn account requirement: Sequences with HeyReach LinkedIn steps require at least one LinkedIn account to be assigned as a channel. If no LinkedIn accounts are connected, the step creation form shows a prompt to connect an account first. You can add or remove LinkedIn accounts from the Channels tab in Sequence Details after the sequence is created.

AI Instructions (LinkedIn)

LinkedIn steps support an AI Instructions mode.

Instead of writing a static message, you can provide instructions and let graph8 generate a personalized message for each contact at send time.

The AI uses:

• Contact data (name, role, company)
• Global Context
• Campaign documents

This allows dynamic, per-contact personalization and keeps messages relevant and up to date.

Sequence Settings

Setting	What It Controls
Finish on reply	Stop the sequence for a contact when they reply
Send in same thread	Keep all emails in the same email thread
Wait for new contacts	Pause the sequence and resume when new contacts are added to the list
Shared	Make the sequence visible to your entire team

Running a Sequence

Sequences move through these statuses:

Status	Meaning
Drafted	Created but not yet started
Scheduling	Processing and scheduling contacts
Live	Actively executing steps
Paused	Temporarily stopped — can resume
Waiting	Paused, will resume when new contacts are added
Completed	All contacts have finished the sequence
Terminated	Manually stopped — cannot resume

Controls:

• Start — begin executing the sequence
• Pause / Resume — temporarily stop and restart
• Terminate — permanently stop the sequence
• Duplicate — create a copy of the sequence

Pre-Launch Validation

Before launching a sequence, graph8 checks that all required fields are available for the selected channels.

If any required data is missing, you’ll see a warning with details about what needs to be fixed before launch.

Contact States

Each contact in a sequence has their own state:

State	Meaning
Queued	Waiting to begin
In progress	Currently executing steps
Paused	Manually paused
Completed	Finished all steps
Failed	A step failed
Bounced	Email or message bounced
Removed	Manually removed from the sequence
Replied	Contact replied (sequence stopped if finish-on-reply is enabled)

You can pause or resume individual contacts without affecting the rest of the sequence.

Schedules

Schedules define when steps can execute:

• Sending windows — set the hours and days of the week when messages can be sent
• Timezone — respect each contact’s timezone for delivery timing

Create reusable schedules in Engage → Schedules and assign them to any sequence.

Templates

Save time with reusable templates:

• Sequence templates — save an entire sequence structure (steps, timing, channels) for reuse
• Step templates — save individual step content as templates

Reports & Analytics

Every sequence tracks detailed performance:

Metric	What It Shows
Overview	Total contacts, active, completed, success rate, overall progress
Step breakdown	Per-step counts: queued, in progress, completed, failed, bounced, replied
Conversion funnel	Contacts entered vs. completed at each step
Step performance	Sent, delivered, replied, bounced counts and rates per step
Response rates	Reply rates across time frames (24h, 48h, 7 days)
Engagement trends	Daily reply rate and bounce rate trends

Skipped Contacts

The step progress also tracks Skipped contacts.

Skipped contacts are those that were not processed due to:

• Missing required data (e.g., no email or phone)
• Deduplication rules
• Step conditions not met

The skipped count appears when there are affected contacts.

Step Conditions

Attach conditions to any step so it only runs for contacts who meet specific criteria. Contacts who fail the condition are marked as Skipped and continue to the next step.

Condition Types

Condition	Example
Field value	Job title contains "VP"
Previous step outcome	Step 1 email was opened
Engagement threshold	Opened ≥ 2 previous emails
Custom field	ICP score > 80
List membership	Contact is in list "Enterprise ICP"
CRM field	HubSpot lifecycle stage = MQL
Time since event	Website visit within last 7 days

Adding a Condition

1. Open a step in the sequence editor
2. Click Add Condition
3. Pick the condition type and value
4. Choose what happens if the condition fails:
   • Skip step — move to the next step
   • Exit sequence — remove the contact from the sequence
   • Wait and retry — re-evaluate after N days (useful for “wait until they visit the site”)

Tip

Use conditions to personalize sequences without branching. One example: “If contact opened email 1, send LinkedIn message; otherwise, send email 2.” Keeps the sequence simple while adapting to behavior.

Branching

Branches split a sequence into multiple paths based on contact attributes or behavior. Each path can have its own steps, content, and timing.

When to Use Branches

• Send different content to different personas (VP vs. IC)
• Fork sequences by channel preference (email-first vs. LinkedIn-first)
• Treat replied contacts differently from non-repliers
• Escalate to phone after 2 ignored emails

Adding a Branch

1. In the sequence editor, click Add Branch on any step
2. Define the split condition (same types as Step Conditions)
3. Add steps to each branch
4. Paths can re-converge into a shared follow-up step or end independently

Branch Analytics

Each branch reports independently:

Metric	Scope
Contacts per branch	How the split distributed
Per-branch conversion	Reply rate, meeting-booked rate per path
Winning branch	Side-by-side comparison for A/B test patterns

Provider Fallback

When a primary channel resource fails, graph8 can fall back to a secondary resource automatically.

Fallback Modes

Mode	Behavior
Strict	If the assigned resource fails, skip the step
Rotate	Try the next resource in the assigned pool
Cross-channel	Fail over to a different channel (e.g., email → LinkedIn)

Common Use Cases

• Mailbox at send cap → rotate to a different mailbox in the pool
• LinkedIn account rate-limited → wait 4 hours and retry on the same account
• Phone number rejected (do-not-call list) → skip and continue sequence
• WhatsApp template rejected → email fallback with the same content

Configure fallback per sequence in Sequence Settings → Fallback.

Rate Limits and Send Caps

graph8 enforces multiple caps to protect sender reputation:

Cap Type	Default	Purpose
Per-mailbox daily cap	40–100 emails	Warmup-aware; adjusts based on reputation
Per-LinkedIn daily cap	50 connect + 50 messages	Matches HeyReach platform limits
Per-phone daily cap	100 outbound calls	Configurable per number
Org-wide burst cap	5× daily volume over 1 hour	Prevents spikes that trigger spam filters
Per-contact cap	1 email per day per sequence	Prevents over-messaging

When a cap is hit, the step re-queues for the next available window. Caps are visible in the sequence’s Resource Allocation view.

Deliverability

Sequence performance depends on the underlying mailbox health. Before launching a high-volume sequence:

• Warm up mailboxes for 2–3 weeks so cold sends land in the inbox, not spam
• Verify domain DNS (SPF, DKIM, DMARC) is configured and active
• Run placement tests weekly to catch deliverability drops early

Deliverability is configured in Settings, not Engage. See Email Warmup → for the full warmup workflow, placement testing, and reputation monitoring.

Studio Integration

Campaigns created in Studio can be launched directly into the Sequencer. When launched, the sequence inherits the campaign’s messaging, personas, and channel strategy. The link between Studio campaign and Engage sequence is maintained for end-to-end performance tracking.

Tip

Start with 3–5 steps per sequence. Multi-channel sequences (email + LinkedIn + phone) typically outperform single-channel ones. Use the conversion funnel report to identify where contacts drop off.

Next Steps

• Inbox → — Manage replies and conversations
• Dialer → — Power dialer for outbound calls
• Connections → — Connect your CRM for contact sync
• Email Warmup → — Build sender reputation before launching