outbound-sequences

name	outbound-sequences
description	Build or audit a compliant multi-touch cadence across email/LinkedIn/phone. Trigger: 'build me a sequence', 'set up a cadence', 'enroll <contact>'. All steps draft-only ahead of the send gate.
origin	ESCC

Outbound Sequences

The multi-touch cadence engine. This skill assembles compliant, channel-mixed sequences — email, LinkedIn, call, and voicemail — pulling structure from playbook-library, approved proof from product-knowledge, and voice from brand-voice. Every sequence produced here is draft-only; no step is live until it passes outbound-reviewer and clears the fail-closed pre:outbound-send-gate hook.

Governing rules: rules/common/outbound-compliance.md (unsubscribe + sender identity on every commercial email, suppression check before enroll), rules/common/messaging-style.md (one CTA, < 120-word emails, no soft closes), rules/common/selling-principles.md (evidence-first, no fabricated claims, no false completion).

When to Activate

Activate this skill when:

• An SDR needs a complete multi-touch cadence for a segment, persona, or named account — standard, event-triggered, or ABM variant.
• A rep asks to enroll a contact in a sequence (suppression check first, then build or select the appropriate sequence).
• A sequence needs an audit or refresh — checking for stale proof points, compliance gaps, or low-response steps.
• outreach-analytics signals that a step is underperforming and a variant needs to be drafted and swapped in.

Do not activate for a single first-touch only (use cold-outreach for that). Do not activate to log or track sequence enrollment — use crm-operator for enrollment logging.

Default Cadence Structure

The baseline SDR cadence is 7 touches over ~12 days across three channels. Deviate with reason; document it in the sequence header.

Step	Day	Channel	Goal
1	0	Email	First touch — personalized opener, value hook, CTA
2	1	LinkedIn	Connection request or InMail — short, no pitch yet
3	4–5	Email	Follow-up — new angle or proof point, same CTA
4	5	Call	Live dial — 30-second value framing, ask for 15 min
5	5	Voicemail	20–25 seconds if no answer — reference email, one clear ask
6	10–12	Email	Break-up / pattern-interrupt — reframe or different persona
7	12	LinkedIn	Soft engagement (comment, share) or final InMail

Day 0 is enrollment day. Steps 4 and 5 are same-day (call first, leave VM if no answer). Email body limits: < 120 words per messaging-style.

Event-Triggered Variant

Compress the cadence when a relevant trigger fires (funding round, job change, product launch, conference attendance, intent signal). Reduce to 4–5 touches over 7 days; lead with the trigger in step 1.

Step	Day	Channel	Note
1	0	Email	Trigger-led opener ("saw your Series B announcement…")
2	1	LinkedIn	Connection + one-line relevance ref
3	3–4	Email	Proof point + CTA
4	4	Call + VM	Live attempt
5	7	Email	Break-up

ABM Variant

For a named account with multiple stakeholders, build per-persona threads (Champion, Economic Buyer, Blocker) that run in parallel and reference each other. Coordinate timing so the same account does not receive three emails from three reps on the same day. Each thread still follows the compliance and CTA rules below.

Workflow

Step 1 — Suppression check (MANDATORY before enroll)

Before adding any contact:

1. Call crm-operator (read) to check the contact against the suppression list (prior opt-out, DNC, hard bounce, complaint, or legal hold).
2. If suppressed: stop — do not enroll, do not contact. Log the block.
3. If clean: proceed to sequence selection or build.

This is not optional. post:outbound-style-check audits enrollments retroactively; a suppressed contact in a sequence is a compliance violation.

Step 2 — Sequence selection or build

1. Check playbook-library for an approved sequence template matching the segment + persona + product motion. If a match exists, start there.
2. If no template: build from the default cadence structure above, selecting the appropriate variant (standard / event-triggered / ABM).
3. Pull value props and proof points from product-knowledge via its specificity ladder — role + segment + competitor (the contact's role resolves from jobtitle), approved entries only, with provenance. If no approved proof exists for the slot, soften to a question; do not invent a metric.
4. Apply voice and phrasing from brand-voice. Do not invent a new tone.

Step 3 — Draft each step

For every step, define:

• Channel (email / LinkedIn / call script / voicemail script)
• Timing (day relative to enrollment)
• Goal (what outcome this step is driving toward)
• Draft content (full copy — see format requirements per channel below)
• Compliance block (email: unsubscribe + sender identity, every time)

Email step format requirements

Subject: [< 50 chars, honest, no clickbait]
Body: [< 120 words, one idea, CTA last]

---
[Unsubscribe link — mandatory on every commercial email]
[Sender full name, title, company]
[Physical/postal business address]

No email step ships without the unsubscribe + sender-identity block. This is enforced by post:outbound-style-check and required by outbound-compliance.md.

Call script format (30 seconds)

Opening: "[Name], this is [Rep] from [Company] — is this an ok time for 30 seconds?"
Value: "[One sentence on the specific pain or trigger you saw.]"
Ask: "I'd love 15 minutes to show you how [outcome]. Worth a quick chat this week?"
Objection pocket: [One prepared response to "send me an email" or "not interested"]

Voicemail script format (≤ 25 seconds)

"Hey [Name], [Rep] at [Company]. [One-line trigger reference.] Left you an email —
subject is [subject line]. If it's relevant, I'm at [number]. No worries if not."

Never leave a voicemail pitching the full product. Reference the email; keep it < 25 seconds.

LinkedIn step format

• Connection request: < 200 chars, no pitch, one-line relevance.
• InMail: < 150 words, same structural rules as email minus the compliance block (LinkedIn's own unsubscribe mechanism applies); still requires accurate sender identity and no fabricated claims.

Step 4 — Compliance review of every email step

Before the sequence is considered ready for outbound-reviewer:

• Unsubscribe link present on every commercial email step
• Sender full name, company, and physical address present on every email
• No fabricated metrics — every number traces to an approved product-knowledge entry with provenance
• No deceptive subject lines
• One CTA per email step, no stacked asks
• Each step adds a new angle or value — no re-pitch in different words

Step 5 — Hand to outbound-reviewer

Pass the full sequence draft. outbound-reviewer checks confidence across compliance, claim accuracy, and tone before it is eligible for the send gate. A sequence is draft-only until outbound-reviewer clears it and a human approves the live send (recorded in the state store).

Step 6 — Enroll via crm-operator

Once approved:

1. crm-operator logs the enrollment event in HubSpot against the contact record — sequence name, start date, step count.
2. Bulk enrollments (> 1 contact) require a review pack and are capped by ESCC_BULK_SEND_MAX.
3. Track variant performance in outreach-analytics; use that data to retire underperforming steps and promote winners.

Examples

Standard sequence — RevOps persona, mid-market:

SEQUENCE: RevOps Mid-Market / Q3 Forecast Pain / 7-step
Enrolled: 2026-06-16 | Suppression check: CLEAN

Step 1 — Day 0 | Email
Subject: Forecast accuracy for {company}

Hi {first_name},

RevOps teams at companies like yours spend the last week of every quarter
manually reconciling pipeline because reps update CRM inconsistently.

Our customers close that gap in under a day — first clean forecast typically
happens on day one after setup, going off our onboarding data.

Worth 15 minutes to see if the problem looks similar here?

{Rep first name}
{Title} | {Company}
{Physical address}
Unsubscribe: {link}

---
COMPLIANCE: ✓ unsubscribe ✓ sender identity
PROOF: product-knowledge PP-031 (approved, verified 2026-05-02)
ANGLE: forecast pain — day-one time-to-value

Step 2 — Day 1 | LinkedIn
Request note (< 200 chars):
"Hi {first_name} — work with RevOps teams on forecast accuracy.
Saw you're at {company}; thought it worth connecting."

---
Step 3 — Day 4–5 | Email
Subject: How {similar_company_type} handles late-quarter reconciliation

Hi {first_name},

One thing that surprises RevOps teams: the reconciliation problem is almost
always a CRM hygiene issue, not a forecasting tool issue.

Happy to share how other teams have approached it — 15 minutes this week?

{Rep first name}
{Title} | {Company}
{Physical address}
Unsubscribe: {link}

---
COMPLIANCE: ✓ unsubscribe ✓ sender identity
PROOF: no specific metric used — framed as pattern, not stat
ANGLE: reframe root cause

Step 4 — Day 5 | Call script (30 sec)
"Hi {first_name}, this is {Rep} from {Company} — 30 seconds ok?
Sent you a note about forecast reconciliation for RevOps teams.
We help teams like yours get a clean forecast on day one — I'd love 15 minutes
to show you how. Anything on your calendar this week?"

Step 5 — Day 5 | Voicemail (≤ 25 sec, if no answer)
"Hey {first_name}, {Rep} at {Company}. Left you an email — subject is
'How {similar_company_type} handles late-quarter reconciliation.' If that's a
live problem, I'm at {number}. No worries if not."

Step 6 — Day 10–12 | Email (break-up)
Subject: Closing the loop

Hi {first_name},

I'll stop reaching out — clearly the timing isn't right.
If late-quarter pipeline chaos becomes a priority, you know where to find us.

{Rep first name}
{Title} | {Company}
{Physical address}
Unsubscribe: {link}

---
COMPLIANCE: ✓ unsubscribe ✓ sender identity

Step 7 — Day 12 | LinkedIn
Engage with a recent post (like or comment) or send a final InMail:
"Closing the loop — happy to revisit if forecast accuracy lands on the radar."

Suppression block example:

Requested: enroll [e-mail] in RevOps Mid-Market sequence.

Suppression check → BLOCKED.
Reason: prior opt-out logged 2026-04-12 via email unsubscribe.

Action: contact not enrolled. Event logged in HubSpot (crm-operator).
Do not re-add to any sequence. Do not attempt manual outreach.

Outreach-analytics-driven step retirement:

outreach-analytics reports: Step 3 (Day 4 email) reply rate 0.8% over 90-day
window — below 2% threshold.

Retiring Step 3 body. Drafting variant:
  ANGLE: Peer comparison ("what your counterparts at similar companies told us")
  PROOF: product-knowledge — pull approved social-proof entry, if available.
  If none: frame as question, not a claim.

Variant drafted → hand to outbound-reviewer before promoting to live sequence.

Anti-patterns

• Skipping the suppression check. Adding a suppressed contact is a compliance violation. The check is non-optional, every time, before any enrollment.
• Missing the compliance block on any email step. One step without an unsubscribe link fails the whole sequence. post:outbound-style-check will catch it, but do not rely on the hook to find what you should have included.
• Inventing a metric to make step 1 land. A specific number with no approved product-knowledge entry is fabrication. Soften to a question or a framed pattern; never invent a stat.
• Re-pitching the same angle. Each step must add a new angle, reframe, or proof point. "Just following up" with the same body is a soft-close anti-pattern (messaging-style.md explicitly bans it) and will be flagged by outbound-reviewer.
• Claiming the sequence is live. A sequence is a draft until outbound-reviewer clears it and a human approves the live send. Do not report it as enrolled or sending until crm-operator logs the tool-result confirmation.
• Stacking CTAs. One ask per touch. "We could also jump on a call, or maybe you'd prefer a demo, or I could send a one-pager" is three asks and guarantees zero action.
• Ignoring outreach-analytics signals. If a step is persistently below the reply-rate threshold, retire it. Running a known-bad step because "it might still work" wastes send quota and degrades domain health.

Related

• cold-outreach — first-touch personalization workflow; feeds step 1 of this skill.
• playbook-library — approved sequence templates; check here before building from scratch.
• product-knowledge — source of all approved proof points; never fabricate.
• brand-voice — sets tone and phrasing for all copy.
• outbound-reviewer — must clear every sequence before the send gate.
• outreach-analytics — reply-rate and conversion data; drives step retirement and variant promotion.
• crm-operator — the sole write-capable agent; logs all enrollments and activity.
• opt-out-handling — processes unsubscribe requests and updates suppression.
• rules/common/outbound-compliance.md — the legal/regulatory floor.
• rules/common/messaging-style.md — structural bar for all outbound copy.
• Command: /sequence.