Cold Email ROI: How to Calculate It, and What Actually Moves It

Cold email ROI = (gross profit from closed deals − fully loaded campaign cost) ÷ fully loaded campaign cost. In most B2B cases the cost side is small and nearly fixed. Mailboxes, data, tooling, and operator time typically land in the hundreds of dollars per closed deal, which means ROI is decided almost entirely by two things: your average contract value and whether your emails reach the inbox at all.

That sentence is the whole article in compressed form. The rest shows you the arithmetic, where the numbers come from, which inputs are worth fighting for, and the specific ways cold email ROI math misleads people who are otherwise doing good work.

What is the formula for cold email ROI?

Use two formulas, not one. The first is the honest ROI calculation:

• ROI = (Gross profit from closed-won deals − Total campaign cost) ÷ Total campaign cost
• Gross profit per deal = ACV × gross margin. Use gross profit, not revenue. Revenue ROI flatters software companies and wildly flatters agencies and resellers.

The second is the reverse funnel, which is what you actually plan with:

• Deals needed → ÷ close rate from meeting held → meetings held
• Meetings held → ÷ show rate → meetings booked
• Meetings booked → ÷ positive-reply-to-meeting rate → positive replies
• Positive replies → ÷ positive reply rate per prospect → prospects contacted
• Prospects contacted → × touches per sequence → total emails sent
• Total emails sent → ÷ (sends per mailbox per day × working days) → mailboxes required

Note the unit on reply rate: per prospect, not per email sent. A 3-step sequence to one person is one chance at a reply, not three. Mixing these up is the single most common reason a plan that looked fine on a spreadsheet under-delivers by roughly the number of steps in the sequence.

What does cold email actually cost per month?

There are five real line items. Everything else is noise.

• Mailboxes and domains. The per-seat cost of authorized Google Workspace or Microsoft 365 mailboxes, plus domain registration. Inboxlogy provisions these from $2.80/mailbox/mo with $0 setup, billed monthly, with SPF, DKIM, and DMARC configured automatically. Domains run roughly $10–15/year each.
• Sending tool. Instantly, Smartlead, ReachInbox, or similar. Per-seat or per-contact. Check current pricing rather than trusting a number in a blog post.
• Data and verification. Contact sourcing plus email verification, charged per contact or per credit.
• Operator time. The person who builds lists, writes sequences, triages replies, and watches deliverability. This is almost always the largest line item, and it is the one most teams leave out.
• Closer time. Calls held, prep, follow-up, proposals. Include it at a loaded hourly rate or your ROI number is fiction.

A useful way to normalize all of this is cost per send: take your monthly total for tooling, mailboxes, data, and operator salary, then divide by monthly send volume. Most functioning programs land somewhere between a few cents and about $0.15 per send, dominated by labor. Infrastructure is usually under 5% of it.

How do I model cold email ROI end to end?

A complete worked example follows. Treat every input as a placeholder for your own number. The point is the structure and the sensitivity, not these specific values.

Inputs: ACV $12,000; gross margin 80% (gross profit $9,600/deal); close rate from meeting held 20%; show rate 75%; positive-reply-to-meeting 40%; positive reply rate 1.5% of prospects contacted; 3 touches per prospect; 30 sends/mailbox/day across 22 working days (660 sends/mailbox/month); data at $0.06/verified contact; operator cost $0.10/send; tooling $0.005/send; closer time 10 hours/deal at $75/hour loaded.

Working backwards from one closed deal:

• 1 deal ÷ 20% = 5 meetings held
• 5 ÷ 75% = 6.7 meetings booked
• 6.7 ÷ 40% = 16.7 positive replies
• 16.7 ÷ 1.5% = 1,111 prospects contacted
• 1,111 × 3 = 3,333 emails sent
• 3,333 ÷ 660 = 5.1 mailbox-months

Cost per deal: mailboxes 5.1 × $2.80 = $14; data 1,111 × $0.06 = $67; operator 3,333 × $0.10 = $333; tooling 3,333 × $0.005 = $17; closer 10 × $75 = $750. Total ≈ $1,181.

ROI = ($9,600 − $1,181) ÷ $1,181 ≈ 7.1x, or about 713%. Cost per acquisition is $1,181 against $9,600 in gross profit.

Two things jump out of this model. First, infrastructure is $14 of a $1,181 cost, roughly 1%. Second, break-even ACV is tiny: at these conversion rates you need about $1,476 in revenue per deal at 80% margin to break even. Cold email is structurally profitable for mid-ACV B2B. It still loses money for plenty of teams, and the reason is usually one layer up the funnel.

What happens to cold email ROI when deliverability is bad?

Rerun the identical model with one change: positive reply rate drops from 1.5% to 0.3% because most mail is landing in spam. Everything downstream of the inbox is unchanged. Same copy, same offer, same closers.

• Prospects contacted per deal: 16.7 ÷ 0.3% = 5,556 (5x more)
• Emails sent: 16,667; mailbox-months: 25.3 → $71
• Data: $333; operator: $1,667; tooling: $83; closer: $750
• Total ≈ $2,904 per deal. ROI ≈ 2.3x. Break-even revenue per deal jumps to about $3,630.

Inbox placement moved ROI from 7.1x to 2.3x and raised the minimum viable ACV by 2.5x, without touching a single word of the campaign. Placement is a multiplier that sits in front of every other number in the funnel. It is also why sub-$2,000 ACV offers tend to fail at cold email. They only work at high reply rates, so they have no tolerance for deliverability problems.

The checks that protect the multiplier

• Google's bulk sender guidance is explicit: keep the spam complaint rate in Postmaster Tools below 0.10%, and never at or above 0.30%. Watch this number weekly. It is the closest thing to a public scoreboard you get.
• Keep bounce rate low. Most practitioners treat sustained bounce above 2–3% as a reputation risk and pause to re-verify the list.
• Never send cold volume from your primary company domain. Use separate sending domains so a reputation problem can't reach your invoices, password resets, and support mail.
• Authenticate everything. SPF, DKIM, and a DMARC record are table stakes for Google and Microsoft. Misconfiguration here is the most common silent cause of a collapsed reply rate.
• Include a working unsubscribe path, ideally a List-Unsubscribe header. Complaints are more expensive than opt-outs.
• Keep per-mailbox daily volume conservative, commonly 20–50/day, and scale with mailbox count rather than by pushing individual mailboxes harder.

Why does infrastructure quality affect ROI if it's only 1% of cost?

Its cost is not where its impact is. Infrastructure is cheap, and it governs the expensive variable. Three failure modes show up directly in the ROI calculation:

• Unauthorized or resold mailboxes get suspended. When a tenant disappears mid-campaign, you lose the sequences in flight, the domain reputation you spent weeks building, and the ramp time to rebuild. That cost lands as weeks of zero pipeline, not as a line on an invoice. Inboxlogy provisions authorized Google Workspace and Microsoft 365 mailboxes specifically to remove this category of risk.
• Shared sending IPs mean inherited reputation. Your placement depends on strangers' behavior. Dedicated US or EU IPs make reputation something you control and can diagnose, and EU-region sending matters when your prospects or your data policy are European.
• No ownership means no exit and no visibility. If you don't hold admin access to the tenant, you can't audit DNS, can't export, can't migrate without starting reputation from zero, and can't debug a placement problem below the sending-tool layer. Inboxlogy gives 100% ownership plus admin access, with a full API for programmatic provisioning when you're scaling mailbox counts regularly.

One clarification that matters for planning: Inboxlogy provides infrastructure, not warmup. Warmup runs in whatever sending tool you connect, whether that's Instantly, Smartlead, or ReachInbox, using their warmup networks and schedules. Budget for both layers, and don't assume buying mailboxes means warmup is handled.

How do I measure cold email ROI when the sales cycle outlasts the campaign?

This is where most reporting breaks. If your sales cycle is 60–90 days, a monthly ROI report compares this month's cost against deals sourced three months ago, which makes ROI look terrible while you're scaling and great while you're shrinking. Fix it with three habits:

• Cohort by send month. Tag every prospect with the month they were first contacted, and attribute deals back to that cohort. Report ROI per cohort as it matures. This is the only number that isn't an artifact of your spend curve.
• Track leading indicators weekly, ROI quarterly. Weekly: spam rate, bounce rate, positive reply rate per prospect, meetings booked per 1,000 prospects contacted. Those move fast enough to act on. ROI doesn't.
• Use pipeline-weighted ROI as the interim number. Multiply open pipeline from cold email by your historical stage-weighted close rate, add closed-won gross profit, then divide by cost. State plainly that it's a forecast.

One cheap discipline pays for itself: record the cost per meeting held for every cohort. It's available weeks before revenue is, it's hard to game, and comparing it to your blended CAC from other channels tells you whether to add mailboxes or stop.

What's a realistic payback period on cold email?

Payback has two components people conflate. The ramp is mechanical: new domains and mailboxes need warmup before meaningful volume, so realistically you're 3–4 weeks from provisioning to full send capacity, plus your sales cycle before cash arrives. The financial payback then depends on whether you bill upfront or monthly. A $12,000 ACV collected annually pays back a $1,181 CAC immediately. The same contract billed at $1,000/month on 80% margin pays back in about 1.5 months of gross profit.

The practical implication: judge a new cold email program at 90 days minimum, and never on the first 30. Month one is all cost and no closed revenue by construction. If someone shows you a 30-day ROI figure on a 70-day sales cycle, they're measuring something else.

How do I tell if my cold email is losing money?

Run these five diagnostics in order. Each one isolates a different layer, so you fix the right thing.

• Placement, not opens. Seed a handful of real Gmail and Outlook accounts into a live campaign and look at where the mail actually lands. Open rates are unreliable, since image proxying and prefetching corrupt them. Use seed placement plus Postmaster Tools instead.
• Positive reply rate per prospect. If placement is fine and this is still near zero, the problem is targeting or offer, not deliverability. More volume will only cost more.
• Positive reply to meeting booked. Interest that never converts to calendar time usually means slow or weak reply handling. This is a process fix and it's nearly free.
• Show rate and close rate from cold meetings. Compare against your inbound meetings. A large gap means you're booking the wrong people, which is a list problem disguised as a sales problem.
• Cost per meeting held versus your other channels. Fully loaded, including operator and closer time. If cold email is more expensive per meeting than paid acquisition and the meetings convert worse, the honest answer is to stop, not to send more.

What are the common ways cold email ROI math lies?

• Using revenue instead of gross profit. Inflates ROI by whatever your cost of delivery is. Brutal for services businesses.
• Omitting labor. In the worked example above, operator and closer time were $1,083 of $1,181. Excluding them turns a 7x into a 96x and makes every decision downstream wrong.
• Reply rate measured per email instead of per prospect. Divides your apparent conversion by the number of sequence steps and inflates the volume you think you need.
• Counting all replies as positive. "Not interested," "wrong person," and out-of-office are not pipeline. Track positive replies separately or the funnel is meaningless.
• Attributing deals that would have closed anyway. If a prospect was already in an active inbound motion, a cold touch didn't source it.
• Averaging across segments. One profitable segment and three unprofitable ones can average to "fine." Compute ROI per ICP segment and the decision usually becomes obvious.
• Ignoring the cost of a blown domain. Reputation damage shows up as a reply-rate decline weeks later, never as a line item, which is exactly why it goes unbudgeted.

Frequently asked questions

What is a good ROI for cold email?

There's no credible universal benchmark, because ROI scales with your ACV and margin. Build the reverse funnel with your own numbers and compare the result to your other channels on fully loaded cost per meeting held and CAC against gross profit per deal. If cold email beats your blended CAC and the meetings convert comparably, it's working, whether that reads as 3x or 20x.

How many mailboxes do I need, and what will they cost?

Divide your required monthly sends by (sends per mailbox per day × working days). At a conservative 30/day over 22 days, each mailbox carries about 660 sends/month, so 10,000 monthly sends needs roughly 15 mailboxes. At Inboxlogy's $2.80/mailbox/mo that's about $42/month with $0 setup. Mailbox cost almost never determines whether a program is profitable.

Does buying better infrastructure actually improve reply rates?

It removes ceilings rather than creating replies. Authorized mailboxes, dedicated IPs, and correct SPF/DKIM/DMARC prevent the placement failures that cap your reply rate regardless of how good the copy is. Infrastructure can't make a bad offer convert, but bad infrastructure will reliably stop a good one from being read.

How long before I can judge whether cold email is worth it?

Allow 3–4 weeks of ramp while domains warm up in your sending tool, then one full sales cycle. Judge at 90 days using cohort-based attribution. Watch spam rate, bounce rate, and positive reply rate per prospect weekly in the meantime. If placement is broken, you'll see it in days, and you shouldn't wait for the quarter to end to fix it.