How to Build a Sales Cadence: B2B Rep Step-By-Step

What a sales cadence actually is

A sales cadence is a sequenced set of outreach touches — across email, phone, LinkedIn, and video — aimed at one prospect over a defined window. That is the whole definition. Not a drip campaign. Not a mass send. Not a "nurture." One person, one path, one stop.

The confusion comes from the tools. HubSpot calls it a "sequence." Outreach calls it a "sequence." SalesLoft calls it a "Rhythm." Apollo calls it a "sequence." The object is the same. The strategy the object executes is the cadence — and the strategy is what most reps get wrong.

A rep sends a single email and calls it outreach. Three days later they send a "just checking in." A week later they forget the account exists. That is not a cadence — that is three touches with no plan. A cadence is what happens when you decide, before you send touch 1, what touches 2 through 8 will be, when they fire, what channels they ride on, and what ends them.

The math is unambiguous. HubSpot's 2024 state-of-sales data shows 80% of closed-won B2B deals took five or more touches to open. 44% of reps stop after one touch. The gap between those two numbers is the entire reason cadences exist — to take the decision out of the rep's hands on the morning of Day 5 when they are tempted to mark the account as "not interested" and move on.

Why most sales cadences die before touch 3

Most cadences die quietly between touch 2 and touch 4. The rep sends the first email, hears nothing, loses nerve, and goes looking for easier pipeline. The cadence becomes a two-touch stub. By Friday the account is back on the "not engaged" list and the rep is starting over with new logos next week.

Five patterns cause this. Each is fixable in one decision — usually made before the first send.

1. 1. The first email had no trigger. "Saw your LinkedIn" is not a trigger. Without a specific, dated signal the opener reads generic, the reply rate craters, and the rep loses faith in the cadence after touch 2.
2. 2. The cadence was email-only. Eight emails in a row is a mailing list, not a cadence. Single-channel cadences convert at roughly half the rate of multi-channel — and every email after touch 3 has less lift than a phone call or a LinkedIn DM would.
3. 3. The touches had no new angle. "Just following up on my email below" is a ping, not a touch. A touch that adds no new information trains the prospect to filter the next one.
4. 4. The spacing was wrong. Sending touch 2 six hours after touch 1 feels hostile. Sending touch 2 six days after touch 1 feels forgotten. 1–3 day spacing on the front half, 2–4 day spacing on the back half is the rhythm that works.
5. 5. There was no break-up. A cadence with no stop turns into a low-grade haunt. Break-up emails, done well, pull the highest reply rate in the sequence — 14–22% in our own outreach data — precisely because they signal release.

Notice the common thread: every failure is upstream of the send button. A rep running a well-designed cadence rarely "loses nerve" — the design carries them through Day 10 because Day 10 has a scheduled touch on a specific channel with specific copy. The design is the discipline.

The 7 decisions you make before touch 1

Seven decisions happen before the first send. Make them in order. Skip any and the cadence will fail at exactly the point the skipped decision should have covered.

1. 1

   Who you target

   One ICP, one persona, one trigger. "VP Sales at 50–300 FTE B2B SaaS" beats "anyone in sales." A broad ICP produces a broad cadence, and broad cadences read generic to every reader.

2. 2

   What trigger opens it

   A specific, recent event: new hire in the seat, funding round, pricing-page visit, competitor churn, job post naming a pain. No trigger means the first line is a guess and the first line dies.

3. 3

   Which channels

   Email is the backbone. Phone is the close-rate multiplier. LinkedIn is the warm-up layer. Video is the break-through touch. Pick at least three — single-channel cadences top out at half the reply rate.

4. 4

   How many touches

   Eight to twelve. Fewer and you miss the deals that need five touches to respond (80% of closed-won, per HubSpot). More and you are spamming — and your reply rate drops after touch 10.

5. 5

   How many days

   14 to 21 days. Compressed cadences feel stalkery. Cadences longer than three weeks get lapped by competitors. The 17-day cadence is the 2026 sweet spot across most reliable benchmarks.

6. 6

   What you say

   A reason (the trigger), a hook (the pain it implies), proof (one specific outcome, one customer you look like), and one ask. Four parts, 60–90 words per email. No paragraphs of context.

7. 7

   What ends it

   A break-up email on the final touch, and a 90-day cooldown. Every cadence needs a clear stop so the prospect does not feel hunted and so you do not burn the domain sending touch 17.

Write these seven answers on one sheet of paper before you open the sequence builder. If any answer is "I'll figure it out as I go," the cadence is already broken. The point of the decision sheet is that on the morning of Day 5, when you are tempted to skip the phone call because "they probably do not want to be called," the sheet says "Day 5 is a call, at 4:10pm, with this voicemail" — and you make the call.

How many touches. How many days.

The two numbers everyone asks about — touches and days — are the easiest to answer because the research has converged.

Touches. Eight to twelve. RAIN Group's research on first-meetings-with-cold-prospects lands at an average of eight touches. HubSpot's 2024 closed-won data says five or more. In our own outreach data the curve flattens sharply after touch 10 and turns negative by touch 14. Anything under five leaves money on the floor. Anything over twelve burns domain reputation for diminishing lift.

Days. 14 to 21. The common 3-7-7 cadence pattern (Day 0, Day 3, Day 10, Day 17) captures the bulk of the replies the cadence produces over 30 days — the last 10 days of a 24-day cadence usually yield under 10% of the replies. Shorter than 14 days and you stack too many touches in the same week; longer than 21 and you get lapped.

The one exception. Inbound reply motions compress this whole framework. When a prospect fills a form, a five-touch, ten-day cadence with the first email inside five minutes wins. The Lead Response Management Study's 21× qualification-rate lift on sub-five-minute responses is the most replicated number in sales research and the hardest one to live up to — because it requires a cadence that fires on the form fill, not on the rep remembering the form fill.

The channel mix that works in 2026

Four channels carry a modern cadence. Each one does one job. Use all four and you pull ahead of every rep still running email-only. Use one and you cap out at about half the reply rate you could otherwise hit.

Email carries the story. It is the one channel where you can run the full "reason → hook → proof → ask" arc in a single send. Keep emails at 70–90 words. Past 120 words the reply rate falls off a cliff in every benchmark cohort I have seen since 2022. One ask per email — "20 minutes Thursday at 3?" not "let me know what works."

Phone is the multiplier. A phone call between 4:00 and 5:00pm local time lifts connect rates above any other window in SalesHive's 2025 cold-calling benchmark set. Pair a voicemail with an email minutes later that says "tried your line just now" — the combo cites the voicemail and gives the prospect one move to make.

LinkedIn is the warm-up. A profile view plus a connection note, before email 2 goes out, warms the inbox without adding a send. In our rep data a LinkedIn message + profile visit combo consistently posts the highest single-touch reply rate, roughly in the low double digits. Use LinkedIn for the second hook, not the first — the prospect reads the email first and sees the LinkedIn visit second, which reads as pattern-matching, not random.

Video is the break-through. One 60–90 second Loom or Vidyard, dropped in at touch 5–7, pulls reply rates 2–3× the email average in most cohorts. Show your face, their logo, name the trigger, ask for the meeting in the last 10 seconds. Not every touch should be video — the effect relies on being the one video in a sea of text.

A 14-day cold outbound cadence you can copy

This is the version you can run tomorrow. Eight touches across four channels over 14 days. Built for a cold B2B outbound motion — ICP fit plus a real trigger. The scripts below are written for a rep selling a $10–40K ACV product to a director-or-above buyer, but the spacing and channel rhythm hold across most SaaS segments.

The first email, written for a VP Sales at a 200-person SaaS who just hired a new CRO:

That is 58 words. One trigger (the CRO hire). One hook (forecast roll). One proof (the "similar co"). One ask (Thursday at 3). Every touch after this one reuses the same shape — a new trigger or a new angle, the same hook, a different proof point, the same ask. The break-up email on Day 14:

Break-up emails signal release. They pull the highest reply rate in most sequences because the prospect finally has a reason to respond — either "yes, actually" or "not us, here is who." Both are wins.

4 cadences for 4 sales motions

One cadence does not fit every motion. The 14-day cold outbound version above is the default, but four motions need different shapes. Running the wrong cadence for the wrong motion is the second-biggest reason reps get low reply rates — the first being no trigger.

The inbound reply cadence is the one most teams underrun. A form fill is worth roughly 20× a cold email at the moment the fill lands — and roughly 1× a cold email 30 minutes later. Research from the Lead Response Management Study across 15,000 companies shows contacting inbound leads within five minutes lifts qualification rates 21× versus the 30-minute window. The cadence does not need to be complicated; it needs to fire inside five minutes, and most teams' sequence tooling does not allow for that without a signal-triggered layer.

The stalled-deal revive is the one most reps run backwards. The instinct is to send a polite "just wanted to check in" after two weeks of silence. The data says pattern-interrupt works better — a new hook, a competitor mention, a specific question that requires one word to answer. "Did you move forward with [competitor], or is this still live?" pulls replies in stalled-deal cohorts at 4–6× the rate of a "checking in" email.

Signal-triggered beats date-triggered

The 2026 shift in cadence design is moving from calendar triggers to signal triggers. In the old model, touch 2 fires on Day 3 because the sequence says so. In the signal model, touch 2 fires the moment the prospect views the pricing page, replies to a LinkedIn post, hires into a new role, or opens the first email three times in 20 minutes.

Seven triggers do most of the work in B2B:

• · A new hire in the buyer seat — VP Sales, CRO, Head of RevOps — within the last 30 days.
• · A funding round, especially Series A and B, where tooling budget opens up.
• · A pricing-page visit from someone at the target account within the last seven days.
• · A LinkedIn post that names a pain your product solves, engaged on within 48 hours.
• · A job post naming a pain point your product solves (e.g., "must be comfortable with manual CRM hygiene").
• · A competitor mention or review left publicly (G2, LinkedIn, Reddit).
• · A direct behavioral signal — opens, link clicks, time-on-site — from the rep's own domain.

The trade-off: signal-triggered cadences need tooling that watches for the signal and fires the sequence when it lands. Doing this manually scales to maybe 20 accounts per rep per week. With signal detection built into the workflow, a rep can comfortably monitor 200 accounts and run a cadence on the five that signal in any given week — the ratio flips from "spray and pray" to "watch and pounce."

Six common cadence mistakes that lower reply rates

Six mistakes show up in almost every underperforming cadence I audit. Each one has a one-line fix. Fix the cadence, not the copy — copy is downstream of design.

1. 1

   1. No trigger, no opener

   "I came across your profile" is not an opener — it is a flag that you did no research. Fix: every first email names a specific, dated signal. Signal-triggered openers lift reply rate 3–4×.

2. 2

   2. Single-channel cadence

   Eight emails in a row is a mailing list, not a cadence. Fix: layer phone, LinkedIn, and video. Multi-channel sequences convert 28% better than single-channel (Tendril).

3. 3

   3. "Just checking in" touches

   Touches without new information train the prospect to ignore you. Fix: every follow-up carries a new angle, a new proof point, or a new question — never a ping.

4. 4

   4. Same copy for every persona

   VP Sales and Head of RevOps get the same email — and neither one replies. Fix: one persona per cadence, rewrite the pain line for each, reuse the offer.

5. 5

   5. No break-up

   Running 15+ touches with no exit trains reps to spam and trains inboxes to filter. Fix: every cadence ends with a break-up and a 90-day cooldown. Break-up emails pull the highest reply rate in the sequence.

6. 6

   6. Sending at 10:00am, every time

   Mid-morning is the most crowded inbox window. Fix: test 7:15am local and 4:45pm local. Off-peak sends run 15–30% higher open rates in most 2025 benchmark cohorts.

The meta-mistake under all six: treating the cadence as a send list instead of a reply mechanism. Every decision in the cadence should answer the question "what makes this person more likely to reply?" — not "did we hit our activity number?" Activity-tracked cadences burn both the domain and the rep. Reply-tracked cadences compound.

The 3 metrics that tell you it is working

Cadence dashboards show fifty numbers. Three of them matter.

1. 1

   Reply rate, cadence-level

   Replies ÷ prospects who entered the cadence, measured at cadence end — not per-touch. Healthy B2B cold cadences hit 8–12%. Under 5% and either the trigger is wrong, the first email is generic, or the ICP is off. Above 15% and you should raise volume — you are leaving pipeline on the floor.

2. 2

   Meeting book rate

   Booked meetings ÷ prospects entered. Aim for 2.5–3.5%. If reply rate is fine (8%+) but booking rate is low (under 1.5%), the ask is wrong — usually too soft, too late in the email, or with too many time options. "Thursday 3pm?" outperforms "let me know what works" by 30–40% in most cohorts.

3. 3

   Meeting-to-pipeline ratio

   Qualified opportunities ÷ meetings held. One-in-four is the floor. Below that and the ICP or the trigger is targeting people who are willing to take the meeting but not positioned to buy. This is the number that catches cadences that "work" on the reply metric but never produce pipeline — the quietest failure mode in outbound.

Track these weekly, per cadence. If you have three cadences running, you have three sets of three numbers. Nine data points tells you everything — which cadence is working, which motion is broken, which ICP is tapped out. Dashboards with fifty metrics bury these three and convince reps the cadence needs a new subject line when it actually needs a new trigger.

How Gangly builds cadences that fire on signals

Gangly builds the cadence as part of a connected rep workflow — signal detection, outreach drafting, call prep, live coaching, and post-call notes all in one sequence, not five tabs. The cadence inside Gangly fires on signals, not on calendar days:

• Signal Detection watches connected sources — CRM activity, LinkedIn job changes, company news — and surfaces accounts the moment a real trigger lands. The cadence does not start on Monday. It starts on the signal.
• Outreach Writer drafts every touch against the signal that triggered it — opener, hook, proof, and ask, in the rep's voice. The rep reviews and sends. Personalization scales without losing the trigger.
• Workflow Sequencer carries the cadence across channels — email, LinkedIn, call, video — and tracks where each account is in its sequence. Touch 5 does not get missed because the rep had a Friday.

The rep still runs the deal. Gangly handles the sequencing, the signal watch, and the drafting. The cadence fires on behavior; the rep sends every touch. See how per-seat pricing maps to your team or start a 14-day trial below.

Related reading: why your cold emails go to spam covers the deliverability layer every cadence lives or dies on, the 7 buying signals every rep should watch is the trigger menu for signal-driven cadences, and the 5-part follow-up email breaks down the copy shape every touch 2 and beyond should use.

Frequently asked questions

Tags: sales cadence · outbound sales · cold email · multichannel outreach · sales sequence · B2B sales · signal-led selling