Cold Email Step One: Why 58% of Replies Come From Touch #1

Two of the most-cited outbound reports of 2026 landed within weeks of each other and said almost the same thing. Instantly's Cold Email Benchmark Report — built on billions of sends — found that the first email captures 58% of the replies with the remaining 42% being captured by follow-ups. Sales.co, analyzing 2M+ emails, went further: the first message drives 79.4% of all replies at a higher reply rate (1.83% vs 1.22%) and higher positive rate (8.90% vs 7.55%).

Most teams read those numbers and concluded "write better first emails." That's the wrong takeaway for anyone running a LinkedIn+email sequence in 2026.

The real question is structural. If touch #1 does 58–79% of the work, why is your team spending 80% of its build time on follow-ups 2 through 7? And in a multichannel world, what even is step one — the profile view, the connection request, the InMail, or the first email?

What "cold email step one" actually means in a multichannel cadence

Every single-channel study defines step one the same way: the first email in the sequence. That definition breaks the moment you add LinkedIn.

In a properly coordinated LinkedIn+email cadence, the email that pulls the 58% reply rate isn't cold anymore. It's a follow-up dressed as an opener. The typical pattern is LinkedIn for warm first contact (leveraging profile visibility and professional context), followed by email follow-up for persistence and scale once a relationship exists. Multichannel sequences outperform single-channel outreach by 2–3x on reply rates because prospects who recognize your name from LinkedIn open and respond to follow-up emails at much higher rates than fully cold email. LinkedIn first, email second is the winning order for B2B in 2026.

So the practitioner definition: cold email step one is whatever touch first puts your name in front of the prospect — even if it's a profile view or a connection request. The email inherits that context. If you skip the LinkedIn layer, your "step one" starts from zero and the 58% ceiling drops with it.

Your first email only captures 58% of replies when it's not really the first touch. That's the sleight of hand every benchmark report ignores.

Who this is for (and who it isn't)

Read this if you're:

• An SDR manager running 4–7 touch sequences and watching reply rates plateau
• A founder doing multichannel outbound solo and trying to compress the cadence
• An agency running LinkedIn+email for clients who keep asking why touch #5 isn't landing

Skip it if you're:

• Running pure cold email with no LinkedIn layer (this post assumes both channels)
• Sub-3% reply rate on touch #1 today — your problem is upstream (list quality, deliverability, ICP), not sequence design. Below 3% typically signals problems with list quality, deliverability, or messaging. Fix those first.
• Enterprise sales cycles with named-account plays where every touch is human-written

Why most teams over-engineer touches 2 through 7

Walk into any SDR floor and you'll see the same artifact: a whiteboard with a 7-step cadence, color-coded by channel, timed to the day. Someone spent a week designing it. Nobody's touched step 1 in three months.

The data says that's inverted. First email reply rate: 3.43% on average, generates 58% of all replies. First follow-up reply rate: 8.4% — the highest-performing individual step. Notice the framing: the first follow-up has the highest step-level reply rate, but step one still generates the majority of total replies because it goes to 100% of the list. Every subsequent touch reaches a shrinking pool.

The returns curve is brutal. The second email is the most under-worked asset in the average sequence: only 58% of replies come from the first send, the other 42% arrive later. But 2026 flipped the old "just add touches" advice. As inboxes tighten, aggressive sequences now damage sender reputation faster than they used to, and the reputational cost shows up as worse deliverability on the next campaign. Agency data shows returns bending after the third or fourth touch.

Translation: every touch after #4 has a negative expected value once you factor in domain reputation cost. The right sequence isn't longer — it's shorter, with a bigger investment in touch #1 and a graceful exit hardcoded in.

The rebuilt cadence: one killer opener, then aggressive graceful exit

Here's the design principle: if step one owns the outcome, treat it like a launch, not a template line. Then plan every subsequent touch as a controlled decay.

A reference 5-touch, 14-day cadence:

1. Day 0 — LinkedIn profile view + connection request with a 200-character note referencing a specific trigger (not "loved your post"). Target 40–60% acceptance.
2. Day 2 — Cold email touch #1. Under 80 words. Single binary CTA. References the LinkedIn context implicitly ("saw we just connected") only if accepted; otherwise cold but signal-anchored.
3. Day 5 — LinkedIn DM (if connection accepted) or email nudge with a different angle — same problem, new frame.
4. Day 9 — Email follow-up referencing a second data point. Not "just bumping this up."
5. Day 14 — Break-up email. Explicit, one sentence, no CTA. This one converts surprisingly well because it's a real close, not another ask.

That's it. Five touches, two channels, roughly 60% of the build effort concentrated on touch #1 (the connection note) and touch #2 (the first email). The data points to 4–7 touchpoints as the optimal sequence length. Under four touches and you are giving up before capturing the 42% of replies that come from follow-ups. Beyond seven, returns diminish unless each touch adds genuine new value.

The LinkedIn touch is what makes the 58% number transfer from Instantly's single-channel dataset to your multichannel one. It's also what protects domain reputation — you're not blasting a fifth email into an inbox that already ignored four.

For teams still hesitant about the LinkedIn-first order, our reply-rate breakdown of LinkedIn vs cold email in 2026 shows the channel gap in raw numbers.

What touch #1 has to look like structurally

Step one is not step two written earlier. It's a different object. Structurally, the elite openers share four traits — and the Instantly report is unusually specific about them.

Elite performers average fewer than 80 words per first-touch email. Brevity forces clarity. Every word must earn its place. Multiple CTAs dilute focus. Top performers use binary questions or simple requests that require minimal cognitive load: 'Does this make sense?' or 'Worth a quick call?' Lead with the problem, not your solution.

A practical checklist for touch #1:

• Under 80 words. Not 79. Actually count them.
• One CTA. Binary or ultra-low-friction. Never two.
• Problem-first. The first sentence names their situation, not your product.
• Trigger-anchored. A funding round, a hire, a post, a competitor switch — something that proves this email couldn't have been sent to anyone else.
• Subject line references a specific outcome or situation — not a category.

The follow-ups (steps 3–5) look completely different: longer is fine, secondary angles are fine, social proof belongs here — not in step one. The mistake most teams make is writing all five touches in the same voice and length. Step one is a hook. The rest are handles.

A/B data: what actually moves touch #1

When we look across the 2026 datasets, the leverage points on touch #1 are consistent, and they're not the ones most teams test.

Targeting depth beats copy edits. Small, targeted campaigns (under 50 recipients) average 5.8% reply rates. Large batch sends average 2.1%. That's not a minor difference. It's the difference between a pipeline that works and one that doesn't.

Signal-based personalization is the ceiling raiser. Advanced personalization (beyond first name) doubles reply rates. Signal-based personalization can multiply them by 3-5x. This is why the Claygency playbook has become the default agency stack — the signal layer is what feeds a step one that clears 8%.

Contact count per company matters more than copy. Belkins analyzed 16.5M cold outreach emails across 93 business domains and found that targeting 1-2 contacts per company produced a 7.8% reply rate. Blasting 10+ contacts at the same company? 3.8%. That's a 2x difference from targeting alone, before you write a single word.

Length compression is a free win. The optimal cold email length is 50–125 words. A study of 3 million cold emails found this range achieves a 2.4x higher reply rate than emails over 200 words.

What doesn't move the needle much on touch #1: subject line word count tests, emoji A/Bs, send-time optimization inside business hours. Test the upstream inputs — list, signal, angle — before you test copy at the margins.

The branching logic that makes it a sequence, not a blast

Running the same 5 touches on everyone is single-channel outreach done twice. Real multichannel means the sequence branches on prospect behavior.

Minimum viable branching:

• If LinkedIn invite accepted within 3 days → send a LinkedIn DM as touch #3 (skip the email nudge).
• If invite not accepted in 7 days → withdraw it, continue email-only, no future LinkedIn touches.
• If email opens but no reply after touch #2 → stay on email, switch angle for #3.
• If no email opens at all after touch #2 → deliverability alarm. Pull the domain from the campaign. This isn't a copy problem.
• If prospect replies negatively → kill the sequence. Don't nurture.

The reason branching matters isn't sophistication — it's protection. Every un-branched touch you send to a disengaged prospect degrades your sender reputation and your LinkedIn trust score. Second follow-up reply rate: an additional 3% — adds an incremental lift to the campaign. Third follow-up reply rate: not specified — returns begin to diminish at this stage. Fourth follow-up and beyond reply rate: not specified — the risk of spam complaints and unsubscribes triples.

And remember: Apple Mail Privacy Protection (MPP) pre-loads tracking pixels for Apple Mail users, which means a significant portion of reported "opens" never represent a real human reading your email. Branch on replies, not opens, wherever you can.

The graceful exit is the most under-designed touch

Step five (the break-up email) gets less attention than any other touch in most cadences, and it shouldn't. Done right, it recovers a meaningful chunk of the 42% that follow-ups own.

What works:

• One or two sentences, max.
• Explicit close: "I'll stop reaching out — happy to reconnect if timing shifts."
• No CTA. No calendar link. No "unless you'd like to chat."
• Sent from the same thread as touch #4, not a new subject line.

The reason it converts: it's the only touch in the sequence that isn't asking for something. Prospects who ignored four asks reply to the one that removes the ask. It's also honest — you are stopping, and the domain reputation math confirms you should.

For teams already at LinkedIn scale, the same principle applies to the connection request. If it's not accepted in 7 days, withdraw it. Pending invites drag your acceptance rate and — per our 20-invite rule breakdown — increase the odds of a restriction under LinkedIn's 2026 acceptance-rate gating.

• Instantly's 2026 data shows the first email captures 58% of the replies with the remaining 42% being captured by follow-ups; Sales.co's dataset puts the first-message share at 79.4% with follow-ups generating only 20.6% of replies at a lower reply rate (1.22% vs 1.83%). Either way, step one owns the outcome.
• In a LinkedIn+email cadence, "cold email step one" isn't the first email — it's the LinkedIn touch that makes the email land as a follow-up, not a cold intro.
• Run 4–5 touches, not 7. Returns bend negative after touch #4 and damage domain reputation.
• Concentrate 60% of build effort on touch #1: under 80 words, single binary CTA, problem-first, signal-anchored.
• Branch on behavior (invite accepted, email opened, negative reply) and design an explicit graceful-exit touch — it recovers the last measurable chunk of replies without burning reputation.