Outreach Sequence

TL;DR. An outreach sequence is a structured series of touches, messages, delays, and channel switches, that runs against a list of prospects on autopilot. The unit of work in modern B2B outbound. A well-designed multichannel sequence produces 2–3× the reply rate of a single-touch send and 4–6× the pipeline at the same effort cost. This guide covers what sequences are, the canonical shapes, what each step is for, where teams build broken sequences, and how to operate them at scale without burning accounts or list quality.

What is an outreach sequence?

An outreach sequence is a multi-step automated cadence of touches sent to a list of prospects. Each step in the sequence is defined by:

• Channel, email, LinkedIn DM, LinkedIn connection request, LinkedIn InMail, voice memo, sometimes phone or video.
• Delay, time since the previous step or the start of the sequence (e.g., "+3 business days").
• Content, the message body, or in modern tools, the AI generation rule that produces the body.
• Exit conditions, what makes a prospect leave the sequence (reply, accept, meeting booked, unsubscribe, manual pause).
• Branch logic, what happens conditional on the prospect's behavior (e.g., "if they opened but didn't reply, send X; if they clicked, send Y").

A sequence operates as a small state machine. A prospect enters at step 1, advances through the steps based on delays and behavioral triggers, and exits when an end condition is met.

Why sequences exist

Three reasons modern outbound runs on sequences instead of one-off sends:

1. Follow-up math

Across thousands of B2B outbound programs, the distribution of replies by step in a 4-step cadence is roughly:

• Step 1: 45–55% of total replies
• Step 2: 20–30%
• Step 3: 10–15%
• Step 4 (breakup): 8–15%

Skipping steps 2–4 leaves roughly half of all available pipeline on the table.

2. Repeatability

A good sequence captures the best parts of how your best rep would manually run outbound, and scales it. Without sequences, every rep invents their own cadence, the team-wide pattern is invisible, and you can't measure or improve.

3. Channel switching

Modern B2B buyers don't live on one channel. Some open every email; some never; some reply on LinkedIn but ignore email entirely; some need a voice memo to engage. A sequence that touches multiple channels reaches more prospects without requiring you to know the right channel for each one upfront.

The canonical sequence shapes

Three sequence archetypes cover most B2B use cases. Pick one based on your ICP, channel preference, and target cycle length.

The LinkedIn-only sequence (4 steps over 12 days)

1. Day 0: Connection request with a personalized note (1–3 sentences referencing a specific signal).
2. Day 2 (if accepted): Free DM that builds on the connection note, usually a specific question or a 30–60 second voice memo.
3. Day 6 (if no reply): A second DM that adds new information, a relevant case study, a useful resource, or a single specific question.
4. Day 12 (if still no reply): Soft breakup, "Closing the loop on this. Happy to revisit in Q3 if timing's better."

Best for: founders, executives, people who live on LinkedIn but rarely check work email. Reply rate range: 18–35%.

The email-only sequence (5 steps over 18 days)

1. Day 0: Initial cold email, signal-grounded opener, lookalike proof, tiny ask.
2. Day 3: Value-add follow-up, a 1-paragraph insight or resource relevant to their role. No CTA.
3. Day 7: Social-proof follow-up, one customer quote or stat. Soft CTA.
4. Day 12: Reframe, same problem, different angle.
5. Day 18: Breakup, permission-based close.

Best for: CFOs, operators, IT/security buyers, technical roles, anyone whose primary work surface is email. Reply rate range: 8–28%.

The multichannel sequence (6–7 steps over 18 days)

1. Day 0: LinkedIn connection request with personalized note.
2. Day 1: Initial cold email, references the LinkedIn touch.
3. Day 3 (if connection accepted): LinkedIn DM, voice memo or specific question.
4. Day 6: Email follow-up #1, value-add.
5. Day 10: Soft LinkedIn engagement, comment on a recent post, share their content.
6. Day 14: Email follow-up #2, social proof + soft CTA.
7. Day 18: Multichannel breakup, short email + short LinkedIn message acknowledging the close.

Best for: high-value ICP, complex sales, accounts where you'd happily pay $2–5 in tooling per touch to reach the right buyer. Reply rate range: 22–45%.

The five jobs of a sequence

A well-designed sequence does five distinct things, each step usually doing only one of them.

1. The trigger / opener

Step 1's job is to earn the right to step 2. Not to book a meeting, not to sell, just to land a relevant message that doesn't feel like spam. A great opener references a specific signal from the prospect's recent activity.

2. The proof

One step (usually step 2 or 3) exists purely to establish lookalike credibility. "We've helped 8 RevOps teams at NYC SaaS shops ramp new SDRs to 40%+ reply rates in their first month." Specific. Plausible. Adjacent to the prospect's context.

3. The reframe

A reframe step takes the same underlying problem and presents it from a different angle. If step 1 framed the problem as "long ramp time," the reframe step might present it as "missed Q4 quota risk." Same root cause, different surface.

4. The value-add

A step that gives the prospect something useful without asking for anything in return. A short essay, a benchmark, a resource link, a tip. Pure giving builds reciprocity.

5. The breakup

The final step explicitly closes the loop. Counterintuitively, breakup messages have higher reply rates per send than middle-sequence steps, partly because they're often the most clearly-written, partly because they create a sense of "last chance" without being pushy.

Where sequences go wrong

1. The five-paragraph step 1

Too much information in the first touch. Step 1 should be the shortest message in the sequence, not the longest.

2. The "I'm just bumping this up" follow-up

Step 2 that adds no new information. This is the most common reason reply rate drops off after step 1. Every step should give the recipient a new reason to respond.

3. Too many steps

Sequences over 7 steps see diminishing returns and rising spam-mark rates. A 9-step sequence isn't a sign of patient outbound, it's a sign of a target prospect who has decided not to respond.

4. Wrong delays

Back-to-back daily follow-ups feel desperate. 14-day gaps feel cold. The 3–6 day window between mid-sequence steps is where reply rate is highest.

5. No exit conditions

Sequences that keep sending after a prospect has replied are the fastest way to burn a relationship. Every sequence needs an unambiguous "stop on reply" rule, plus manual-pause and unsubscribe paths.

Branch logic and behavioral triggers

Modern outbound platforms (Linkziy, Outreach, Salesloft, Apollo) support branching: a sequence can split based on the prospect's behavior on a previous step.

Common branches

• If opened but didn't reply: Send a slightly more direct follow-up. The prospect saw your message and chose not to act, additional context might unblock them.
• If clicked but didn't reply: Send a follow-up that builds on what they clicked. They showed interest in a specific topic.
• If didn't open: Send from a different subject line, or switch channels. Don't keep sending more from the same envelope.
• If LinkedIn connection request was accepted: Branch into the LinkedIn DM track. If declined, stay on email.
• If prospect engaged with your content: Send a warmer, less salesy follow-up.

Branches are powerful but expensive in time to build and maintain. A reasonable rule: every sequence should have at most 2 branches. Beyond that, the sequence becomes a maze that's hard to debug when reply rates drop.

Operating sequences at scale

A handful of sequences run by one rep is easy. A library of 20 sequences across 5 reps, all targeting different ICPs and updating with new copy, is genuinely hard. The teams that operate well share three habits.

1. One sequence per ICP segment

Don't build one mega-sequence that tries to speak to everyone. Build a small sequence per ICP segment, tied to a single saved list filter. When the ICP changes, the sequence changes. The mapping is 1:1.

2. Auto-pause on broken steps

Any sequence step whose reply rate drops below 12% for three consecutive days should auto-pause. This stops broken steps from continuing to burn list quality while a human investigates.

3. Quarterly sequence reviews

Every quarter, review every active sequence's reply rate by step. Retire sequences with bottom-quartile performance. Promote winning steps into new sequences. Treat the sequence library as a product, not a one-time setup.

The sequence vs the playbook

A sequence is a specific cadence with specific copy. A playbook is the higher-level strategy, when to use which sequence, on which list, with which sender. Mature teams write both.

Example playbook entry: "Sequence X is used by AE Tier-1 for accounts that score 90+ on our ICP fit model, in the 30 days after a trigger event. Tier-2 accounts get Sequence Y, which has a longer cadence and a lower-stakes ask."

How AI changes sequence design

Five years ago, the entire copy of every step was static. Today, every step's copy can be generated per-prospect by an LLM grounded in that prospect's profile. The implications:

• Sequence templates become structural, not literal. "Step 1 is a signal-grounded opener with a tiny ask", not "Step 1 is the exact sentence below."
• Personalization can happen at every step, not just step 1.
• Reply rate variance between reps shrinks, the AI handles the personalization work that used to be a "good rep / bad rep" differentiator.
• The bottleneck moves from "writing the message" to "picking the right prospect." Which makes the ICP even more central.

How Linkziy builds sequences

The Outreach Automation module ships with a visual sequence builder. Each step is a card on a flow canvas. Channel, delay, AI rule, and exit conditions are configured per step. The AI fills in the copy per prospect at send time, grounded in the prospect's LinkedIn profile and any signals you've enriched.

Sequence-level analytics show reply rate by step, by sender account, and by ICP segment. Sequence steps auto-pause at the floor. Each prospect's exit reason is logged so it's clear why a sequence finished, replied, accepted-and-replied, unsubscribed, sequence-ended.

Bottom line

A great outbound program is a library of well-targeted sequences, not a pile of one-off messages. Sequence design is a craft, every step needs a job, every delay needs a reason, every branch needs to earn its complexity. The teams that take sequence design seriously consistently produce 3–5× the pipeline of teams that treat sequences as a glorified mail-merge.