How do you coach reps to build a multichannel outreach sequence in 2026?

A concrete scenario: the rep with fourteen touches and zero replies

A mid-market SaaS rep — call her Dana — is three months into quota and sitting at 0.8% reply rate on outbound. Her manager pulls her sequence in Outreach and finds fourteen touches across 21 days: nine emails, four LinkedIn connection requests with no note, one call attempt on day 19. Every email opens with "Hope you're doing well" and pivots to a feature list. The LinkedIn requests have no message. The single call has no voicemail logged. Dana is not lazy — she is running 60 new prospects a week and logging 400 dials a month. She is busy in exactly the way that produces nothing.

This is the most common coaching scenario in outbound, and it is almost never a motivation problem. Dana's manager makes the classic mistake first: he rewrites two of her emails himself, sends them, gets a reply, and declares the coaching done. Dana learns nothing, and her next sequence looks identical to the last one. The correct move is to slow down, diagnose why the sequence is failing, and rebuild her mental model of what a multichannel outreach sequence actually is.

The diagnosis takes one 25-minute conversation. The manager asks Dana to walk him through why touch two follows touch one. She cannot answer — she never thought about it. He asks her to name three pains her target persona has on a Tuesday morning. She names one, vaguely. He asks which steps she actually ran last week. She skipped the phone steps because calling is uncomfortable. That is three separate gaps in one rep: a skill gap (she has never seen a well-built cadence), a knowledge gap (she does not understand the buyer well enough to write relevance), and a will gap (she avoids the phone). Each one needs a different coaching response, and treating them as one problem — "work harder" — fixes none of them.

The rebuild takes three weeks. Week one, they co-build a single six-touch sequence together, touch by touch, and Dana ships it. Week two, she runs it at volume and they review per-step reply rates. Week three, she builds a second sequence alone and defends her reasoning out loud. By day 90 her reply rate is 6.4% and her meeting rate has tripled — not because she got a better tool, but because she learned to make touch four sound like it remembers touch one.

That arc — diagnose, co-build, run, measure, transfer — is the entire coaching model. Everything below is the detail.

How the mechanism actually works: diagnosing before you coach

The fastest way to waste a 1:1 is to coach the wrong gap. Reps build weak multichannel sequences for four reasons, and each requires a different intervention. Get the diagnosis wrong and you will spend six weeks coaching motivation at someone who simply has never been shown what good looks like.

Skill gap. The rep blasts the same paragraph across every channel — the LinkedIn note is the email with the greeting stripped off. They have never seen a real cadence built well, so they have no template in their head. Coaching response: model it. Build one sequence together, live, and narrate every decision you make. Then have them rebuild a different one from scratch while you watch.

Knowledge gap. The rep does not know the persona's pains, so every touch is generic. No amount of channel-stacking fixes a message with nothing to say. Coaching response: persona research and win-loss review. Pull three closed-won deals in the segment and have the rep interview the AE about what the buyer actually said in the first call. Relevance is a research output, not a writing skill.

Will gap. The rep knows how but skips the phone because calling is uncomfortable, and hides behind email where rejection is silent. Coaching response: role-play calls until the discomfort drops, then set an activity floor you review weekly. Will gaps close with reps and accountability, not with more training.

System gap. The sequence is fine, the rep is fine, but the data is bad, deliverability is tanked, the territory is saturated, or the tool is not configured to log steps. That is a manager and RevOps problem, not a coaching problem. Do not coach a rep for a system failure — fix the system.

Once you have the diagnosis, run the coaching conversation itself on the GROW model — Goal, Reality, Options, Will. Keep the rep building; you are the guide, not the author.

Goal — set the target out loud. Open with: "Before we touch the sequence, what are we actually trying to get — a meeting, a reply, a referral? Tell me the one outcome for this cadence." Then: "Who exactly is this for — title, company size, the moment they would care?" If the rep cannot answer in two sentences, stop and fix the targeting before you touch a single message. A sequence aimed at "VP Sales, any size company" is not a sequence, it is a spray.

Reality — look at the work, not the intention. Pull up the last sequence in Salesloft or Outreach and ask: "Walk me through why touch two follows touch one. What does touch two know that touch one did not say?" Most reps have never been asked this and discover in real time that their touches are seven unrelated cold starts. Then ask: "Which of these did you actually run last week, and which got skipped?" This surfaces the will gap honestly and without accusation — the tool logs the truth.

Options — co-build, do not dictate. Say: "Let's write the first touch together. One problem, one sentence of relevance, one soft ask, no pitch. What is the problem this persona has on a Tuesday?" Build a sample cadence with them live. A workable six-touch structure across 11 working days:

• Day 1 — Email, short. "Saw [trigger event]. Teams at [similar company] hit [specific problem] right after that. Worth a 15-minute compare?"
• Day 2 — LinkedIn, connect plus light note. "Connecting after your post on [topic] — curious how you are handling [problem] this quarter."
• Day 4 — Phone plus voicemail. "Hi [name], [rep] from [company] — left you a note on LinkedIn. Calling because [one reason]. I will send a two-line email so you have it in writing."
• Day 5 — Email, value, no ask. Share one specific insight or benchmark relevant to their world. Do not ask for the meeting.
• Day 8 — Video, 45 seconds. Walk one screen tied to their situation. Loom or Vidyard both work; the tool matters less than the specificity.
• Day 11 — Phone plus breakup email. "Last note from me — should I close the loop, or is the timing just off?"

Will — lock the commitment. End with: "What will you change in your next sequence, and when will it be live in Apollo?" Then: "What might stop you from making the calls — and how do we handle that now?" Write down the commitment and the date in the same doc. Coaching without a logged next step evaporates by Thursday.

The weekly engine that keeps this alive is a closed loop: observe a real touch, diagnose the weakest one, coach one change, role-play it, let the rep run the revised cadence, measure the per-step reply rate, repeat.

Real numbers, ranges, and benchmarks

Coach to leading indicators, because quota is a lagging number that arrives too late to do anything about. The numbers below are the ranges most teams see in outbound; use them as a sanity check, not a target to force, and always compare a rep against their own trailing baseline before you compare them to the team.

Sequence length. Six to nine touches across 11 to 18 working days is the productive band for most B2B outbound. Below five touches, you are quitting before the prospect has seen you twice. Above twelve, you are usually burning territory and domain reputation for marginal returns — and the extra touches are almost always more emails, not more channels.

Channel mix. Three to four channels is the sweet spot: email, phone, LinkedIn, and optionally video. The distribution that tends to work is roughly 40-50% email, 25-30% phone attempts, 20-25% LinkedIn, and 5-10% video. If a rep's sequence is 80% email, they do not have a multichannel sequence — they have an email sequence with decoration.

Cadence spacing. Two to five working days between touches. Tighter than two days reads as harassment; looser than five and the prospect forgets you between touches. Front-load: the first three touches inside the first week, then widen. Let engagement guide it — a prospect who opened three times but did not reply earns a faster next touch; one who never opened earns a channel switch, not another email.

Reply rates. Healthy outbound reply rates sit in the 5-15% range depending on segment, list quality, and offer. Per-step reply rate is the number that matters for coaching: if touch one gets 4% and touch five gets 0.3%, touch five is dead weight and you should replace it, not defend it. Meeting-booked rate from a cold sequence typically lands between 1% and 4% of prospects contacted.

Connect-to-conversation rate on calls. Of live connects, 10-20% converting to a real conversation is a reasonable band. Below that, the opener is the problem — coach the first fifteen seconds, not the whole script.

Step-completion rate. This is the will-gap detector. If a rep's sequence has four phone steps and the tool shows 40% completion, you have a will gap, not a messaging gap. Set a floor — every phone step gets attempted, every voicemail gets left — and review it weekly. Completion below 80% on any step means the sequence is a fiction.

Speed to first touch. Responding to an inbound or triggered lead within five minutes dramatically outperforms an hour later, and the gap widens from there. For outbound cold sequences the clock matters less, but time-to-first-touch after a signal event is still one of the highest-leverage numbers a rep controls.

Behavior change over time. The only metric that proves the coaching stuck: does the rep build the second sequence better than the first, without you in the room? Track the reply rate of sequence one versus sequence two versus sequence three. If the curve is flat, you coached a sequence, not a rep.

A 30/60/90 arc to make those numbers move. Days 1-30: the rep ships one multichannel sequence you reviewed touch-by-touch. Goal is message quality and channel logic, not quantity — one great sequence beats five mediocre ones. Days 31-60: they run it at volume while you review reply and meeting rates weekly and A/B exactly one variable at a time — subject line, day-one channel, or video versus no video. Changing three variables at once teaches you nothing. Days 61-90: the rep builds a second sequence solo and walks you through their own reasoning before they send it. You are transferring the diagnostic skill, not the script.

Trade-offs and alternatives

Multichannel sequencing is not free, and pretending otherwise produces bad coaching. Every channel you add costs the rep time, and every automation you add costs you authenticity. Coach the trade-off explicitly so reps make the choice on purpose.

Breadth versus depth. Four channels reach more of the buyer's day, but each channel demands its own craft. A rep who is excellent at email and mediocre at the phone will often do better running a tight two-channel sequence than a sloppy four-channel one. The alternative to breadth is depth: fewer channels, better messages, higher completion. For a rep in their first quarter, depth usually wins.

Automation versus personalization. Sequencing tools let you schedule, templatize, and scale. That is the point — and it is also the trap. An AI sequence-builder can produce fourteen fluent, forgettable touches in ninety seconds. The human edit for relevance is the entire value-add. Coach reps to automate the *timing* and personalize the *content*, never the reverse.

Volume versus territory health. Pushing "more touches" when the touches are generic scales the failure and burns the territory — more unsubscribes, worse deliverability, and a list that gets poisoned for the next rep. There is a real ceiling on how many cold touches a domain and a prospect list can absorb. When a rep is below target, the first instinct should be "which touch is weak," not "add three more."

Tool choice. Outreach, Salesloft, and Apollo are the common platforms; the differences that matter for coaching are reporting granularity and how easily a manager can inspect per-step performance. A tool the rep will actually keep updated beats a more powerful tool they abandon. Do not let a tool migration become the coaching project.

AI assistance versus AI authorship. AI is genuinely useful for drafting variants, summarizing research, and suggesting subject lines. It is not a coach. A rep who lets an AI write the cadence end-to-end produces touches that are grammatically perfect and emotionally empty. The rule to teach: AI drafts, human edits for relevance, rep owns the send.

Specialization versus generalist reps. Some teams split prospecting from closing, which lets a specialist run deep multichannel sequences all day. Others keep one rep on the whole cycle, which caps how much prospecting craft they can build. Neither is wrong, but the coaching plan has to match the model — a full-cycle rep needs a sequence they can run in 30 minutes a day, not a 90-minute daily ritual.

Common pitfalls and how to avoid them

Coaching the deal instead of the skill. Rewriting the rep's email yourself fixes today's sequence and teaches nothing. The rep watches you succeed and concludes they cannot. Co-build instead, then make them ship the next one alone. If you must write, write it out loud and narrate every choice.

Rescuing the rep. Jumping on the call "to show them how it's done" feels helpful and steals the rep's reps. Let them struggle inside a role-play where the stakes are zero, then let them run the real call. The discomfort is the training.

Coaching volume over message quality. "Make more dials" is the most common and least useful coaching sentence in outbound. If the messages are generic, more volume just burns the list faster. Fix the message, then scale the volume.

No follow-through. A 1:1 with no logged commitment and no review date is a conversation, not coaching. Every session ends with a written change, an owner, and a date. Put it in the same place you track pipeline so it cannot disappear.

Coaching everyone the same. Your will-gap rep needs accountability and an activity floor. Your knowledge-gap rep needs persona research and win-loss interviews. Running the same script for both wastes the session and frustrates the rep who needed the other thing.

Treating AI as the coach. Letting an AI sequence-builder write the cadence end-to-end produces fluent, forgettable touches. AI can draft and suggest; the human edits for relevance. If a rep cannot explain why a touch exists, the AI wrote it and the rep does not own it.

Skipping the diagnosis. The single most expensive mistake. Twenty minutes of diagnosis saves six weeks of coaching the wrong gap. Ask the three questions — can you describe a good touch, can you name three persona pains, are you running the phone steps — before you coach anything.

Measuring opens. Open rates are unreliable and easily inflated. Coach to reply rate, meeting rate, and step completion. A rep optimizing for opens will write clickbait subject lines and hollow bodies.

Ignoring deliverability. A rep with a perfect sequence and a burned domain gets zero replies and concludes the sequence failed. Check domain health, bounce rate, and spam complaints before you coach messaging. Sometimes the problem is infrastructure, and that is a RevOps fix.

Letting the sequence outlive its usefulness. A cadence that worked in Q1 can decay by Q3 as the market shifts and the list saturates. Teach reps to retire steps that stop performing instead of defending them. The sequence is a hypothesis, not a monument.

Related questions

How long should a multichannel outreach sequence be?

Six to nine touches across 11 to 18 working days suits most B2B outbound. Fewer than five and the prospect barely registers you; more than twelve usually means padding with extra emails that damage deliverability. Let reply data, not habit, decide when to cut a step.

Which channels belong in a modern sequence?

Email, phone, LinkedIn, and optionally short video. Email carries the volume, phone carries the urgency, LinkedIn carries the familiarity, and video carries the specificity. Three or four channels is the practical ceiling — beyond that, reps cannot maintain craft on any of them.

How do you know a rep needs coaching versus a better list?

Check step-completion rate and per-step reply rate first. If the rep is completing steps and messages are specific but replies are near zero, suspect list quality, targeting, or deliverability. If steps are being skipped or messages are generic, it is a coaching problem.

What is the manager's role in a rep's sequence?

Diagnose the gap, co-build one sequence touch-by-touch, review per-step data weekly, and transfer the diagnostic skill so the rep builds the next one alone. The manager owns the standard and the follow-through, not the writing.

Can AI build the sequence for the rep?

AI can draft variants, suggest subject lines, and summarize research. It cannot supply relevance or judgment. Let AI draft, have the human edit for relevance, and require the rep to explain why each touch exists before it ships.

FAQ

What is the first step in coaching a rep to build a multichannel outreach sequence? Diagnose before you coach. Ask whether the rep can describe a good touch, name three persona pains in thirty seconds, and whether they actually ran the phone steps last week. The answers tell you whether you are facing a skill, knowledge, will, or system gap — and each one needs a different intervention.

How many channels should a multichannel sequence include?

Three to four: email, phone, LinkedIn, and optionally video. The point is not the count but the logic — each touch should reference the previous one, and each channel should play to its own strength. Blasting the same paragraph everywhere is one channel wearing four costumes.

What cadence between touches actually works? Two to five working days, front-loaded so the first three touches land inside week one, then widening. Let engagement guide adjustments: a prospect who opened repeatedly earns a faster next touch, while one who never opened earns a channel switch rather than another email.

How do you make touches feel connected instead of robotic? Write one core message, then adapt it into channel-specific versions. The email can recap a LinkedIn comment; the voicemail can reference the email; the video can walk the exact screen the email described. The test is whether touch four sounds like it remembers touch one.

What tools help reps run multichannel sequences? Outreach, Salesloft, and Apollo are the common platforms for scheduling touches, logging steps, and reporting per-step performance. The tool matters less than whether the rep keeps it updated — a well-used simple tool beats an abandoned powerful one, and reporting granularity is what lets a manager coach.

How do you measure whether the sequence is working? Track per-step reply rate, sequence-to-meeting conversion, step-completion rate, connect-to-conversation rate on calls, and time-to-first-touch. Healthy reply rates land in the 5-15% range depending on segment. Ignore open rates — they are unreliable and coach the wrong behavior.

Sources

• Gong Labs — sales research and benchmark blog
• Outreach — sales engagement resources
• Salesloft — cadence and prospecting best practices
• RAIN Group — sales prospecting research
• Harvard Business Review — the new science of sales force productivity
• Sales Hacker — outbound and multichannel cadence guides
• Apollo.io — outbound sequence playbooks
• Winning by Design — outbound prospecting framework

Related on PULSE

• What is lemlist and why is it a hot RevOps multichannel outreach platform for 2027?
• What outreach sequence and messaging works best for win-back campaigns?
• How are AI-driven sales assistants reshaping the post-demo follow-up sequence for enterprise buying committees?
• How does the presence of an AI-powered 'procurement bot' change the sequence of sales touchpoints in 2027?
• What is the optimal number of AI-generated touchpoints per week in a 2027 enterprise nurture sequence?
• How can I ask a question that helps a rep structure a better follow-up sequence?