Quick Summary
Email marketing ROI is (revenue attributed to email minus total campaign cost) divided by total campaign cost, times 100. If you want the number to mean anything, use gross profit rather than revenue, and put every cost in the denominator: copywriting, design, platform fees, overhead, the lot.
- The profit-based version is the one worth reporting, because it accounts for margin.
- Price labor at a loaded hourly rate, and pick one overhead percentage you apply to everything.
- Spread one-time costs, like a template build, across the months they serve.
- Pick an attribution window, write it down, and stop changing it.
- A holdout group is the only way to find out how much of that revenue email actually caused.
What Email Marketing ROI Measures
Email marketing ROI tells you how much profit each dollar of email spend brings back. Open rate and click-through rate do not tell you that, which is worth saying out loud because they get reported as if they do. They are early signals about a send, not a return.
A campaign can post lovely opens and still lose money, because people opened it and then didn't buy anything. When a budget conversation is happening, this is the number that should be on the screen.
There are two versions of it. Revenue-based ROI puts total revenue in the formula, which is easy and ignores margin completely. Profit-based ROI uses gross profit instead, and for most businesses that is the figure that reflects reality.
Use profit-based ROI when money is being allocated. Revenue-based ROI is fine for a quick sanity check between two campaigns.
The Email Marketing ROI Formula
The core formula is:
ROI (%) = (Revenue attributed to email − Total campaign cost) ÷ Total campaign cost × 100
So a campaign that generated $12,000 on $3,000 of total cost comes out at ($12,000 − $3,000) ÷ $3,000 × 100 = 300%.
Swap gross profit in for revenue whenever money is being allocated, and especially when margins vary across products. A 40% margin product and a 70% margin product counting the same is how a campaign ends up looking better than it was.
Email Marketing Cost Analysis: What to Include
Most bad ROI numbers are costing problems before they are attribution problems. A campaign looks profitable when you count the platform fee and nothing else, then stops looking profitable the moment someone's hours go into the denominator. Work from a written checklist rather than from what you happen to remember.
The full cost checklist
| Cost category | What to count | How to price it |
|---|---|---|
| Platform subscription | Monthly or annual plan fee, prorated to the campaign period | Invoice amount divided by campaigns sent in the period |
| Sending infrastructure | Custom sending domain setup, any dedicated IP fees if you pay for one | Recurring fees for the period; one-time fees amortized over 12 months |
| Copywriting | Hours spent drafting, editing, and reviewing | Hours times a loaded hourly rate |
| Design and production | Images, video, HTML build, template edits | Hours times rate, plus any stock or asset licensing |
| List growth | Paid acquisition, lead magnets, landing page costs, contest prizes | Total spend for the period |
| List hygiene | Verification or cleaning tools, suppression list maintenance | Tool fee plus staff hours |
| Deliverability work | Authentication setup, DNS changes, monitoring time | Hours times rate, amortized if one-time |
| QA and testing | Seed testing, link checks, A/B test setup and analysis | Hours times rate |
| Reporting | Time pulling numbers, building the monthly review | Hours times rate |
| Discounts and incentives | Code redemptions, free shipping, bundled offers | Total discount dollars redeemed, with revenue kept at list price |
| Overhead allocation | A share of tools, management, and shared services | A fixed percentage of direct cost, applied consistently |
Pricing staff time honestly
Use a loaded hourly rate, not salary divided by 2,080. Loaded means salary plus payroll taxes and benefits. A multiplier between 1.25 and 1.4 on base hourly pay is close enough; it is approximate, and it is still far closer than the raw salary figure.
If someone splits their week across channels, estimate the email share and apply it. A guess of 20% beats the zero you are currently using.
Amortizing one-time costs
A custom sending domain, a template rebuild, a landing page: none of these are monthly costs. Spread them across the months they will serve. A $600 template used for a year is $50 a month. Dump the whole $600 on one campaign and you have made that campaign look bad and every campaign after it look better than it is.
Allocating overhead
Shared costs such as a design tool subscription or a manager's time don't belong to any one campaign, which is not the same as them being free. Pick an overhead percentage of direct cost, something in the 10% to 20% range, and apply it to everything. The exact figure matters less than using the same one every month.
A worked cost build
Discount codes are an easy line to get wrong. A 20% code on a 50% margin product gives up 20% of the list price, which is 40% of the profit on that order. Every discount dollar comes straight out of profit, so count the full amount redeemed, and keep revenue at list price so the discount is only counted once. Order exports often show revenue after discounts, so add the discounts back first. Here is a month of a weekly newsletter costed out, with illustrative numbers:
| Line item | Monthly amount |
|---|---|
| Platform plan | $79 |
| Copywriting, 6 hours at $60 loaded | $360 |
| Design, 2 hours at $65 loaded | $130 |
| QA and reporting, 3 hours at $55 loaded | $165 |
| List growth spend | $150 |
| Discounts redeemed | $200 |
| Overhead at 15% of direct cost | $162.60 |
| Total campaign cost | $1,246.60 |
$1,246.60 is your denominator. The $79 plan fee is the number that often gets used instead, and the gap between those two numbers is wide enough to flip a budget decision from yes to no. Keep the checklist in a spreadsheet, one column per month, and fill it in while the work is fresh.
How to Track Email Campaign Revenue
Attribution is the hard part, and there is no clean answer. Email almost never closes a sale by itself. Someone opens three emails, clicks one, then buys a week later off a search ad. Whatever you attribute to email is a product of the rules you chose, so choose them before you look at any numbers, and then leave them alone. Picking the model, the conversion window and the UTM convention is covered step by step in how to set up an email attribution model.
Pick an attribution window and defend it
The window is the period after a send during which a conversion counts toward that email. Seven days, fourteen, thirty, measured from a click or from an open. Each choice gives you a different answer:
| Window | What it captures | What it misses |
|---|---|---|
| 1 day after click | Immediate, high-intent purchases | Considered purchases, B2B cycles |
| 7 days after click | Most consumer impulse and short consideration | Longer research cycles |
| 30 days after click | Longer consideration, subscription trials | Overlaps with other channels, inflates credit |
| Open-based | Readers who saw but did not click | Understates when images are blocked |
Click-based windows are the safer default, because a click is something a person actually did. Opens depend on a pixel that plenty of inboxes block or prefetch, so the open count is already an estimate sitting underneath your revenue estimate.
Write the window into the reporting doc. If it changes between months your trend line is fiction, because a jump in ROI might only be a wider net.
Separate correlation from causation
Attribution tells you who bought after getting an email. It does not tell you the email caused it. Some of those people were going to buy regardless. A holdout test is the way to find out how many:
- Split the list at random into a test group and a holdout group.
- Send to the test group only.
- Once the window closes, compare conversion rates between the groups.
- The difference is your incremental lift. Multiply it by the number of people you sent to and by average order value at list price for incremental revenue. Run the ROI formula on that figure, using gross profit if that is the version you report, to see what email actually earned.
Hold back around 10% and you will usually see a signal on a big send. On a small list the difference can disappear into noise, and when that happens the result is a hint, not a finding. If you want the same control-group logic applied stage by stage, there's a walkthrough of measuring lifecycle email.
Handle delayed and cross-device conversions
Two things quietly eat email revenue before it reaches your report.
The first is delay. A B2B lead books a meeting 45 days after the first touch, your window closed at 30, and that revenue simply never shows up. Either stretch the window for lead generation work or report pipeline value on its own. The second is devices. Someone reads on a phone at lunch and buys on a laptop that evening, and cookie-based tracking loses the thread. Tagged codes and logged-in order data claw back some of it, never all.
The cleaner report has two numbers side by side: attributed revenue inside the window, and a separate pipeline figure for things that closed later. Merging them into one ROI number buries two different sets of timing assumptions in one decimal.
Use three tagging layers
No single tracking method catches everything, so stack them:
| Method | Catches | Fails when |
|---|---|---|
| UTM parameters | Clicks that lead to a tracked session | Reader blocks tracking, or types the URL |
| Per-campaign discount codes | Purchases that skip the link | Reader buys something else or uses no code |
| Logged-in or order-source data | Purchases tied to a known account | Guest checkout, privacy restrictions |
Then compare the totals. When UTM revenue and code revenue diverge badly, the usual culprit is lost sessions rather than a weak campaign, though it is worth checking both before you conclude that.
A validation workflow
Before anyone acts on a monthly figure, pull three exports and make them agree: the platform's click report for the period, analytics sessions filtered to your UTM values, and order or deal data with source and date fields. Reconcile the counts, and chase anything more than about 10% apart. Record what you settled on, the window you used, and the holdout result if you ran one, because in six months nobody will remember.
Double counting is a common failure. Two emails in one sequence both claiming the same purchase under a 30-day window, and suddenly the revenue exists twice.
For teams running broadcasts and one-to-one sequences side by side, SEMAOS keeps both in one tool. Platform sends track clicks automatically; a sequence sent from a connected Gmail or Outlook mailbox has no click data unless you turn on click tracking and verify a tracking domain. Pull revenue from your own order or deal records.
Worked Calculation Examples
Take a small online store, again with illustrative numbers and only three cost lines so the arithmetic stays visible:
| Item | Amount |
|---|---|
| Email platform (monthly) | $79 |
| Copywriting and design time | $400 |
| Discounts redeemed | $200 |
| Total campaign cost | $679 |
| Revenue attributed to email, at list price | $3,400 |
| Gross profit (at 50% margin) | $1,700 |
Revenue-based ROI: ($3,400 − $679) ÷ $679 × 100 ≈ 401%
Profit-based ROI: ($1,700 − $679) ÷ $679 × 100 ≈ 150%
Same campaign, and the second number is well under half of the first. 150% is still a good month. It is just a less flattering slide.
ROI by Campaign Type and Business Model
Returns vary a lot by campaign type. Welcome sequences tend to do well, mostly because the reader just raised their hand and still remembers who you are. Newsletters earn less per send and cost very little, which is a different kind of good.
What you measure depends on the business. E-commerce comes down to revenue per email and average order value. B2B services means counting booked meetings or qualified leads and putting a value on each, which is its own argument internally. Subscription products need trial starts tracked through to paid conversion over a longer period, often a quarter.
Compare campaigns within a business model, never across. Ranking a lead-gen email against a flash sale on the same scale produces a number that looks rigorous and means nothing.
Improving Email Campaign Performance
With a baseline in hand you can start moving it. If part of the question is how much of the sales motion should be automated at all, the manual vs automated sales workflow guide covers that. Otherwise, the levers that actually move revenue or cost:
- Segment so the offer matches what the subscriber has done
- Change one variable per broadcast test, however tempting it is to change three
- Use engagement scores to work the most engaged contacts first
SEMAOS handles broadcasts and one-to-one sequences in the same place, with broadcast A/B testing and a nightly 0 to 100 engagement score per contact on paid plans and the trial. For ranking contacts by intent, see how to use email engagement scores to prioritize leads.
Setting Up Your Monthly ROI Review
Same calculation, same day every month. The review itself is short:
- Total all campaign costs for the month
- Pull revenue at list price, attributed within your fixed window
- Calculate both revenue-based and profit-based ROI
Doing it monthly is what catches a problem while it is still small. When a campaign falls from 300% to 120%, you want to know why before the next budget goes out, not after.
FAQ
What is the email marketing ROI formula?
ROI (%) = (Revenue attributed to email − Total campaign cost) ÷ Total campaign cost × 100. A campaign that generated $12,000 on $3,000 of total cost returns 300%. Use gross profit instead of revenue for the version you report.
What costs should I include in email marketing ROI?
Platform fees prorated to the period, sending infrastructure, copywriting and design hours at a loaded rate, list growth and hygiene, deliverability work, QA, reporting time, the full value of discounts redeemed (with revenue kept at list price so they count once), and a consistent overhead allocation of roughly 10% to 20% of direct cost.
How do I price staff time for an email campaign?
Use a loaded hourly rate rather than salary divided by 2,080. Multiply base hourly pay by about 1.25 to 1.4 to cover payroll taxes and benefits. If someone splits their week across channels, estimate the email share and apply it.
How do I know if email actually caused the revenue?
Run a holdout test. Split the list at random, send to the test group only, and compare conversion rates once the attribution window closes. The difference is incremental lift; multiply it by the number of people you sent to and by average order value at list price for incremental revenue. A roughly 10% holdout usually shows a signal on a large send.