---
title: "Channel mix cost per reply: email vs LinkedIn vs phone vs mail"
description: "A worked model for comparing cost per reply across email, LinkedIn, phone, and direct mail on the same unit economics."
date: "2026-08-15"
tags: "channel mix, outbound strategy, cost per reply, multichannel"
readTime: "19 min read"
slug: "channel-mix-cost-per-reply"
canonical: "https://firstsales.io/blog/channel-mix-cost-per-reply/"
---

# Channel mix cost per reply: email vs LinkedIn vs phone vs mail

**TL;DR:** Most teams pick channels by gut feeling, not by unit cost. Cost per reply is the fair comparison across email, LinkedIn, phone, and direct mail, because it accounts for both the price of the touch and how often it actually gets a response. Email usually wins on raw cost per reply. Phone wins on cost per meeting once you factor in how fast a call converts a reply into a booked slot. Direct mail is expensive per touch but can beat everything on a short list of high value accounts. The right mix depends on deal size, list size, and how saturated each channel already is for your buyer.

---

## Table of contents

- [Why cost per send is the wrong number](#why-cost-per-send-is-the-wrong-number)
- [The unit that actually matters](#the-unit-that-actually-matters)
- [A worked example across four channels](#a-worked-example-across-four-channels)
- [Why email still wins on raw cost per reply](#why-email-still-wins-on-raw-cost-per-reply)
- [Where LinkedIn changes the math](#where-linkedin-changes-the-math)
- [Phone: expensive per attempt, cheap per meeting](#phone-expensive-per-attempt-cheap-per-meeting)
- [Direct mail: the outlier that sometimes wins](#direct-mail-the-outlier-that-sometimes-wins)
- [Building your own model](#building-your-own-model)
- [Mixing channels changes the numbers again](#mixing-channels-changes-the-numbers-again)
- [Deal size changes which channel wins](#deal-size-changes-which-channel-wins)
- [Where saturation quietly breaks your model](#where-saturation-quietly-breaks-your-model)
- [Mistakes teams make with this model](#mistakes-teams-make-with-this-model)
- [A simple decision framework](#a-simple-decision-framework)
- [FAQ](#faq)
- [Conclusion](#conclusion)

Most channel debates in sales meetings are arguments about feelings.

Someone says LinkedIn is dead. Someone else says cold calling still works if you're good at it.

Nobody brings a number.

The fix is not complicated. It just requires tracking one thing most teams skip: what a reply actually costs, channel by channel, on the same unit.

## Why cost per send is the wrong number

Cost per send is the number most tools show you by default.

Email costs almost nothing to send. LinkedIn connection requests are free up to a monthly cap. A cold call costs a few minutes of a rep's time. Direct mail costs real dollars before it even lands.

Ranked by cost per send, email wins every time, mail loses every time, and the comparison is meaningless.

It ignores the only thing that matters commercially: how many of those sends turn into an actual reply from a real person.

A channel with a low cost per send and a near-zero reply rate can cost more per reply than a channel that costs ten times as much to touch someone but converts at ten times the rate.

Cost per send tells you what you spent. Cost per reply tells you what you got for it.

## The unit that actually matters

Cost per reply is one formula.

Total cost of the channel for a period, divided by the number of replies that channel produced in that period.

That's it. The complexity is entirely in getting the inputs right, not the math.

Total cost includes tool subscriptions, per-send fees, data costs, and rep time valued at a real hourly rate, not zero.

Rep time is the input teams most often leave out, and it's usually the single biggest cost on phone and LinkedIn.

A rep costing $35 an hour who spends six hours a week on LinkedIn outreach is spending $210 a week on that channel whether or not any tool invoice says so.

## A worked example across four channels

This is a worked example with stated assumptions, not measured data from any specific company. Use it as a template and swap in your own numbers.

Assume a team of one rep, 40 hours a week, targeting a mid-market B2B buyer, over a four week period.

**Email**

- 800 sends per week, 3,200 for the period
- Tool and data cost: $400 for the period
- Rep time: 3 hours a week reviewing and personalizing at $35/hour = $420
- Reply rate: 3.4% (roughly the 2026 platform-wide average)
- Replies: 109
- Total cost: $820
- Cost per reply: $7.52

**LinkedIn**

- 100 connection requests a week, 400 for the period
- Tool cost: $150 for the period
- Rep time: 8 hours a week (requests, follow ups, DMs) at $35/hour = $1,120
- Reply rate on accepted connections plus a follow up message: roughly 12%
- Replies: 48
- Total cost: $1,270
- Cost per reply: $26.46

**Phone**

- 60 dials a day, 5 days a week, 1,200 for the period
- No per-dial tool cost beyond a dialer subscription: $200 for the period
- Rep time: 15 hours a week dialing at $35/hour = $2,100
- Connect rate: roughly 8% of dials reach a live person (industry dialer data puts cold connect rates in the 5-10% range)
- Connects: 96
- Total cost: $2,300
- Cost per connect: $23.96

Phone doesn't produce a "reply" in the email sense. It produces a connect, which is the closer equivalent. Treat connect rate as the phone version of reply rate for this comparison.

**Direct mail**

- 50 pieces sent for the period, targeted at a short list of named accounts
- Print, postage, and a small gift item: $18 per piece = $900
- Rep time: 4 hours total prepping and personalizing at $35/hour = $140
- Response rate on a well targeted mail piece: roughly 4-9% depending on the offer (direct mail industry benchmarks for targeted B2B campaigns)
- Responses: 3 (using the low end of 6%)
- Total cost: $1,040
- Cost per response: $346.67

```mermaid
graph LR
    A[Email: 3200 sends] -->|3.4% reply| B[109 replies, $7.52 each]
    C[LinkedIn: 400 requests] -->|12% reply| D[48 replies, $26.46 each]
    E[Phone: 1200 dials] -->|8% connect| F[96 connects, $23.96 each]
    G[Mail: 50 pieces] -->|6% response| H[3 responses, $346.67 each]
```

![Bar chart comparing channel mix cost per reply across email, LinkedIn, phone, and mail](/images/blog/channel-mix-cost-per-reply/inline-1.webp)

## Why email still wins on raw cost per reply

In this worked example, email wins by a wide margin on raw cost per reply.

That result is not a fluke of the assumptions. It's structural.

Email scales without a linear increase in rep time. Doubling send volume does not double the hours a rep spends, because most of the labor sits in list building and copy, not per-send effort.

LinkedIn and phone are both bounded by rep hours in a way email is not. A rep can only dial so many numbers or send so many personalized connection requests in a day before quality drops.

That's the entire argument for treating [AI drafting with human approval](https://firstsales.io) as an email lever rather than a LinkedIn or phone lever first. It attacks the cost side of the one channel where cost per reply was already the strongest, and pushes it lower without touching reply rate.

![FirstSales campaign sequence screen showing a multichannel touch plan](/images/blog/shared/app-campaign-sequence.webp)

## Where LinkedIn changes the math

LinkedIn's cost per reply looks worse in the raw comparison, but the comparison hides something important: what happens after the reply.

A LinkedIn reply often arrives with more context than a cold email reply. The person has seen a profile, maybe a few posts, before responding.

That context can shorten the sales cycle after the reply, which the cost per reply number does not capture at all.

LinkedIn inboxes are also markedly less saturated than email inboxes for most buyer personas right now, which is the main reason its reply rate in this model sits at 12% against email's 3.4%.

That gap will not last. As more teams pile into LinkedIn outreach, its reply rate moves toward email's, the same way email's own reply rate fell from 5.1% in 2024 to about 3.43% in 2026 as volume climbed industry wide, a trend covered in more depth in our [cold email reply rate benchmarks](/blog/cold-email-reply-rate-benchmarks-2026).

Track your own LinkedIn reply rate monthly. If it's holding steady while your email reply rate erodes, that's real signal to shift budget, not just a hunch.

## Phone: expensive per attempt, cheap per meeting

Phone looks mediocre on cost per connect in the worked example, worse than email, close to LinkedIn.

The number that actually matters for phone is cost per meeting, not cost per connect, because a live conversation converts to a booked meeting at a far higher rate than a text reply does.

If roughly a third of connects turn into a booked meeting on the call itself, and only a fraction of email or LinkedIn replies convert to a meeting without additional back and forth, phone's cost per meeting can beat both channels even though its cost per connect looks worse.

This is the trap in comparing channels on reply alone. Reply is not revenue. Meeting is closer to revenue.

Run the same worked model one step further for your own numbers: track what percentage of each channel's replies turn into a booked meeting within 14 days, then divide total cost by meetings instead of replies.

That second table usually reorders the channel ranking, sometimes completely. Our full breakdown of [cost per meeting outbound](/blog/cost-per-meeting-outbound) walks through that second calculation on its own.

## Direct mail: the outlier that sometimes wins

Direct mail looks catastrophic on cost per response in the worked example, over 40 times more expensive per response than email.

At high volume, that math never works. Nobody should send 3,000 pieces of direct mail a month at $18 a piece hoping for a 6% response rate.

But direct mail was never meant to compete on volume. It's an [account-based tactic](/blog/account-tiering-outbound) for a short list where the deal size justifies a much higher cost per touch.

If the target account is worth $80,000 in annual contract value, spending $18 to stand out in a mailroom that receives zero other physical mail from vendors is not expensive. It's a rounding error against the deal.

The same logic explains why [B2B direct mail is working again](/blog/b2b-direct-mail-outbound) specifically for enterprise accounts and specifically now, when every other channel funnels into a screen the buyer has learned to ignore.

Run direct mail's numbers on a per-target-account basis, never on a per-1,000-sends basis. The moment you scale it past a curated list, the math that made it work disappears.

## Building your own model

The worked example above uses invented assumptions to show the mechanics. Your numbers will differ, sometimes by a lot.

Here is the build order that produces a usable model instead of a spreadsheet nobody trusts.

Start with actual replies from the last 30 to 60 days per channel, pulled from your CRM or sequencing tool, not estimated.

Add actual cost per channel for the same period: tool fees, data costs, ad spend if relevant, and rep hours multiplied by a real loaded hourly cost, not just base salary divided by hours.

Divide total cost by total replies for each channel. That's your baseline.

Rerun it monthly. Channel costs and reply rates both drift, usually down, as a channel gets more saturated or a vendor raises prices.

| Input | ✓ Include | ✗ Skip |
|---|---|---|
| Rep hours on the channel | Real hourly rate x hours spent | Treating rep time as free |
| Tool and data subscriptions | Full period cost, prorated | Only the sticker price of one plan tier |
| Replies counted | Genuine human responses | Auto-replies, bounces, out of office |
| Time window | 30-60 days of real data | A single good week |
| Channel comparison unit | Cost per reply, then cost per meeting | Cost per send |

## Mixing channels changes the numbers again

Channels rarely run in isolation, and the cost per reply model gets more interesting once you look at sequences that combine them.

A prospect who gets an email, then a LinkedIn connection request, then a follow up call replies at a different rate than the same prospect getting only one of those three touches.

[Multichannel sequences](/blog/email-linkedin-multichannel-outreach) generally lift total response rate over any single channel run alone, but attributing the reply to one specific channel gets murky fast.

The practical fix most teams use is last-touch attribution for cost per reply purposes: whichever channel's touch immediately preceded the reply gets credited, even though earlier touches almost certainly contributed.

That's an imperfect model, but it beats not measuring at all, and it's consistent enough month to month to catch real trend shifts.

![Diagram of a decision funnel showing deal size tiers routed to different outbound channels](/images/blog/channel-mix-cost-per-reply/inline-2.webp)

## Deal size changes which channel wins

There is no single winning channel independent of deal size. The worked example above assumes a mid-market deal.

For small deal sizes, under roughly $5,000 in annual contract value, email's cost advantage usually dominates outright. The math for phone or mail rarely closes at that price point.

For [enterprise deals](/blog/enterprise-vs-smb-outbound-effort), the calculus flips. A $150,000 deal can absorb a far higher cost per reply on any channel, and the channels that produce more context per reply, LinkedIn, phone, mail, start to look proportionally cheap.

The mistake is applying one channel mix uniformly across a pipeline with mixed deal sizes. Segment the list by deal size tier first, then assign channel budget per tier, not per total headcount.

## Where saturation quietly breaks your model

Every channel's reply rate is a moving target, and the direction is almost always down as more senders pile in.

Email's fall from 5.1% to roughly 3.43% platform wide between 2024 and 2026 happened because volume grew faster than inbox tolerance did.

The same dynamic is starting on LinkedIn, more slowly, because [connection request limits](/blog/linkedin-connection-request-limits-2026) cap how fast any one sender can flood the channel, even if the whole platform is trending toward more outbound volume overall.

Cost per reply models built on last year's reply rates will overstate a channel's current value. Refresh the reply rate input every month, not once a quarter.

If a channel's reply rate has dropped more than 20% quarter over quarter while cost held flat, that channel's real cost per reply is already worse than your last model shows.

## Mistakes teams make with this model

The most common mistake is comparing raw reply rates across channels without normalizing for cost, which is exactly the cost-per-send trap this whole framework exists to avoid.

The second is ignoring rep time on channels that feel free because no invoice shows up for them. LinkedIn and phone both hide their real cost this way.

The third is running the model once and treating the result as permanent. Channel saturation moves the numbers every quarter, sometimes faster.

The fourth is optimizing purely for cost per reply and ignoring cost per meeting. A cheap reply that never converts to a conversation isn't actually cheap.

The fifth is applying an average cost per reply blended across deal sizes, which hides the fact that the winning channel is different for a $3,000 deal than a $150,000 one.

## A simple decision framework

If you only have time to build one version of this model, build it in this order.

Pull 60 days of real reply data per channel from your CRM.

Calculate actual cost per channel, rep time included at a real hourly rate.

Divide cost by replies. Rank channels.

Then run the same division against meetings booked, not replies, and compare the two rankings.

If the rankings agree, you have a clear channel to lean into. If they diverge, as they often do with phone, the meeting-level number should carry more weight than the reply-level number, because meetings are closer to revenue.

Rerun the whole exercise every 30 to 60 days. Treat the output as a budget input, not a permanent verdict.

## FAQ

### What is cost per reply in outbound sales?

It's total channel cost, including tool fees and rep time, divided by the number of genuine replies that channel produced over a set period. It's a fairer comparison than cost per send because it accounts for how well a channel actually converts.

### Why is email usually cheapest per reply?

Email scales without a matching increase in rep hours, so cost stays close to flat while volume grows. LinkedIn and phone are both bounded by how many personalized touches a rep can physically do in a day.

### Is LinkedIn worth the higher cost per reply?

Often yes, because LinkedIn replies tend to arrive with more buyer context than a cold email reply, which can shorten the sales cycle in ways the cost per reply number alone does not capture.

### How do I count rep time in a cost per reply model?

Multiply hours spent on that channel by a real loaded hourly cost, not base salary alone. Skipping rep time understates the true cost of LinkedIn and phone specifically, since both are labor heavy compared to email.

### What is a reasonable cold email reply rate in 2026?

Platform wide average sits around 3.43% as of 2026, down from about 5.1% in 2024. Systematised, well targeted campaigns land in the 10-18% range, and signal-based sends can reach 5-18%.

### Should cost per reply or cost per meeting drive channel decisions?

Cost per meeting is closer to revenue and should carry more weight when the two rankings disagree. Phone in particular often looks worse on cost per reply and better on cost per meeting.

### Why does direct mail look so expensive in this model?

Because it's priced per piece, often $10 to $25 including postage and any gift, against a response rate in the single digits. It only makes financial sense on a short, high value account list, not at volume.

### Does channel mix matter more for enterprise or SMB deals?

More for enterprise. Small deals rarely justify the cost of phone or mail. Larger deals can absorb a higher cost per reply on any channel, which changes which channel wins the comparison.

### How often should I recalculate cost per reply?

Every 30 to 60 days. Reply rates drift as channels get more saturated, and a model built on stale data will overstate a channel's current value.

### What counts as a genuine reply for this calculation?

A real response from a human, not an auto-reply, bounce, or out-of-office message. Filter those out before running the math or the reply rate will be inflated.

### Can I use open rate instead of reply rate in this model?

No. Apple Mail Privacy Protection inflates open rates by pre-fetching images, making open rate an unreliable and largely dead metric for email specifically. Reply rate is the honest signal.

### Is phone outreach still worth it given low connect rates?

Yes for many teams, because connects that do happen convert to meetings at a much higher rate than a text-based reply does. The cost per meeting, not cost per connect, is the number to check before ruling it out.

### How does AI drafting change the cost per reply math for email?

It reduces the rep-time portion of email's cost by cutting the minutes spent per draft, which lowers cost per reply without needing reply rate to improve at all. It does not change LinkedIn or phone's cost structure.

### What's a realistic LinkedIn connection request reply rate?

In this worked example, 12% on requests that get accepted and receive a follow up message, though this varies by industry, persona seniority, and how personalized the request is.

### Should small teams bother building a cost per reply model?

Yes, even a rough version. A one rep team spending 8 hours a week on LinkedIn with weak results is often better served putting those hours into a cheaper channel, and the only way to see that clearly is the math, not a feeling.

### Does channel saturation affect every industry the same way?

No. Niche B2B categories with fewer total senders competing for attention see slower saturation than broad, heavily targeted categories like SaaS or recruiting.

### How do multichannel sequences complicate this model?

They make attribution harder, since a reply might follow an email, a LinkedIn touch, and a call in sequence. Last-touch attribution is an imperfect but workable fix.

### What's the biggest blind spot in most teams' channel cost models?

Uncounted rep time. Channels that feel free because no line item shows up for them, LinkedIn and phone especially, are usually the most expensive per reply once labor is priced honestly.

### Can this model help decide budget allocation across channels?

Yes, that's its main use. Rank channels by cost per meeting, not cost per reply alone, then shift budget toward the channel producing the cheapest meetings for your current deal size tier.

## Conclusion

Cost per reply is not a complicated formula. It's total cost divided by real replies, and most teams have never actually run it channel by channel.

Email usually wins on raw cost per reply because it scales without matching rep-hour growth. That's structural, not a temporary trend.

Phone and LinkedIn often look worse on cost per reply and better on cost per meeting, which is the number that's actually closer to revenue.

Direct mail only makes sense on a short, high value account list, never at volume.

Build the model with your own real numbers, refresh it every 30 to 60 days, and let deal size decide which channel gets the budget rather than a single blended average across your whole pipeline.

Teams running [FirstSales](https://firstsales.io) for the email leg of their mix tend to use exactly this kind of model to decide where the freed up rep hours from AI-assisted drafting should go next, phone, LinkedIn, or a curated mail list, rather than assuming the answer is always more email.

The channel that wins changes as reply rates shift and budgets move. The unit you measure it with should not.