How to Calculate Unsubscribe Rate the Right Way

To calculate unsubscribe rate, divide the number of unsubscribes from an email by the number of delivered emails, then multiply by 100. Forty-five unsubscribes from 18,000 delivered emails is a 0.25% unsubscribe rate. Use delivered emails rather than total sends, because a bounced message never had the chance to produce an opt-out.

How to calculate unsubscribe rate

The formula for almost every report you'll build:

Unsubscribe rate = (number of unsubscribes ÷ number of delivered emails) × 100

Unsubscribes means unique recipients who opted out of that message or series. Delivered means sent minus bounces. If you'd rather skip the arithmetic, the unsubscribe rate calculator takes both numbers and returns the rate.

A worked case: 25,000 sent, 1,000 bounced, so 24,000 delivered, and 72 people opted out.

72 ÷ 24,000 × 100 = 0.30%

In a spreadsheet that's =Unsubscribes / Delivered * 100, or just =Unsubscribes / Delivered if the cell is already formatted as a percentage. Whatever precision you pick, keep it fixed across dashboards; switching between 0.28% and 0.3% between reports makes small movements look like changes when they're rounding.

The denominator is where teams get it wrong

Three versions circulate, and they don't agree with each other.

Delivered is the right default. Sent-based math flatters you exactly when you least deserve it, because a stale list with a high bounce rate inflates the denominator with addresses that were never reachable. Opens and clicks are diagnostic views, useful for asking why engaged readers left, and they should never stand in for the headline number.

Take the extreme case. You sent 100,000, 20,000 bounced, and 400 people unsubscribed. Sent-based, that's 0.40%. Delivered-based, it's 0.50%. The second number is the true one, and the first one hides a list-quality problem behind a list-quality problem. If your delivery rate is the thing that's slipping, fix that before you read anything into opt-outs at all; hard versus soft bounces explains what you're actually looking at.

What the number is telling you

Unsubscribe rate answers one question: of the people who could receive and react to this email, how many decided they were done with you?

That makes it a diagnostic for expectation mismatch more than anything else. A high rate usually means the content didn't match what the signup promised, the cadence outran what the audience agreed to, the list source was weak, an automation fired at the wrong moment, or the subject line sold something the body didn't deliver.

Unsubscribes are not automatically bad, and treating them as failures leads people somewhere worse. Someone who will never buy, never clicks, and would rather not hear from you is better off leaving cleanly than reaching for the spam button; one costs you a contact, the other costs you reputation with every mailbox provider watching. Gmail and Yahoo now require bulk senders to support easy, one-click opt-out, so the hiding strategy is closed off anyway; the deliverability guide covers how complaints feed back into placement.

Zero unsubscribes isn't the target. A healthy pattern relative to list source, campaign type, and revenue is.

Worked examples across three programs

Newsletter

A SaaS company sends a monthly product newsletter: 40,000 sent, 400 total bounces, 39,600 delivered, 118 unsubscribes.

118 ÷ 39,600 × 100 = 0.298%, which rounds to 0.30%.

Whether that's acceptable depends on the last six to twelve newsletters, not on a benchmark. The next move is segmentation by acquisition source; if webinar leads opted out at 0.12% and old content-syndication leads at 0.75%, the newsletter isn't the problem, the source is.

E-commerce promotion

An online store runs a flash sale: 120,000 sent, 2,400 bounced, 117,600 delivered, 470 unsubscribes.

470 ÷ 117,600 × 100 = 0.40%

Higher than the newsletter, and that's normal; discount-driven mail creates more opt-outs, particularly on lists that get discounts often. Read it against revenue per recipient and against lifecycle stage before you react. Recent buyers unsubscribing at 0.08% while lapsed prospects hit 0.95% is a suppression decision, not a copy decision, and the e-commerce email guide covers how to split those audiences.

Welcome automation

A founder runs a five-email welcome sequence and email three looks wrong: 5,200 delivered, 52 unsubscribes.

52 ÷ 5,200 × 100 = 1.00%

One percent inside a welcome flow is a loud warning, because these are the newest and theoretically warmest subscribers you have. Check the promise on the signup form, the gaps between the first three emails, and whether email three starts selling before the sequence has earned it. Splitting the flow into an educational branch and a sales-led branch is the cleanest test, and the A/B test calculator will tell you whether the difference you see is real or noise.

What counts as a good rate

Most healthy opt-in marketing lands well under 0.5%, and anything sitting near or above 1.0% deserves a look. Beyond that, general benchmarks are orientation rather than targets; your own baseline is the number that should drive decisions, though the industry benchmarks are useful for a sanity check when you're starting from nothing.

Build the baseline from your own history. Pull the last twenty to fifty campaigns, exclude the odd ones (compliance notices, outage announcements, deliberate list cleanups), calculate each on delivered, then find the median and the 75th and 90th percentiles. Flag anything past the 75th for review and treat anything past the 90th as an incident.

A real account might look like this: median newsletter 0.18%, 75th percentile 0.32%, 90th percentile 0.55%. In that account a 0.40% campaign is worth a conversation and a 0.75% campaign is an outlier that needs an explanation.

Keep separate baselines per program type, because they aren't comparable. A win-back campaign to a dormant segment will always produce more opt-outs than a customer-only release note, and an outbound sequence to business contacts produces far more than either; an agency seeing 2.00% on cold follow-up isn't failing by newsletter standards, it's operating in a different context entirely. The right question for that program is whether the opt-outs are clean, complaints are low, and replies justify the send. There's more on blending the two motions in outbound and inbound as a unified email strategy.

One caveat on attribution. Unsubscribes are more concrete than opens, but they're not perfectly counted: people reply asking to be removed, use the mailbox provider's native unsubscribe button, or opt out after a message was forwarded to them. Platforms record those events differently, so compare like with like inside one tool rather than across two.

Why a rate spikes

Spikes have causes, and guessing is expensive. Work through five questions in order.

Did the audience expect this? Expectation mismatch drives more opt-outs than any other single factor. Somebody who downloaded a pricing guide and started receiving daily product pitches is behaving rationally. Check the form copy, whether the lead magnet mentioned ongoing email, whether the contacts were imported from an event or an old CRM, and whether the topic matched the original interest. When one source keeps producing spikes, isolate it and either slow the cadence or ask for a fresh opt-in.

Did frequency change? A list that tolerates one weekly email can reject four in five days, and launches, holiday sales, and quarter-end pushes are where this happens. Calculate by campaign and by rolling period, then compare opt-out rates across frequency bands: subscribers who got one email that week, two to three, or four and up. If the heavy band shows higher opt-outs and a falling click rate, cap sends or give people a preference center instead of a binary.

Did the subject line overpromise? A curiosity-heavy line can lift opens and unsubscribes at the same time when the body doesn't pay off the hook. The subject line tester helps you compare options, but resist optimizing for opens alone; a line that gets fewer opens, more qualified clicks, and fewer opt-outs is the better line even though it looks worse in the first column of your report.

Did deliverability change? Sometimes the rate moves because placement moved. More messages reaching the inbox instead of spam means more people seeing the email, which lifts clicks and opt-outs together. That's not the email getting worse. Read delivered volume, bounce rate, spam complaint rate, and revenue as one picture before you rewrite anything.

Did the rendering break? An email that's unreadable on a phone, or that buries the content under a hero image, loses people who would otherwise have stayed. Check the render before you blame the audience; the responsive email tester covers mobile behavior.

Turning the number into a decision

The percentage on its own is inert. What matters is what you do with it.

Never stop at the campaign average, because averages hide the actual story. Break the rate down by acquisition source, signup date, engagement level, customer versus prospect, lifecycle stage, and frequency exposure. A 0.35% campaign average can be 0.05% among customers and 1.4% among imported trade-show contacts, and those two numbers call for opposite decisions.

Read it alongside intent, not in isolation. An email at 0.45% opt-out with a 9% click rate and real sales is a strong campaign; an email at 0.15% with almost no clicks is being ignored, which is worse and looks better. Put unsubscribe rate next to click rate, conversion rate, complaints, and list growth rate on the same view. That last pairing matters more than people expect: a 0.25% opt-out rate against 0.10% net growth per campaign means the reachable audience is shrinking, and no amount of campaign tuning fixes an acquisition problem. If clicks are the weak side, how to improve click through rates is the better place to spend the effort.

Set thresholds per program so alerts mean something. Warning and critical bands might sit at 0.35% and 0.60% for newsletters, 0.50% and 0.90% for promotions, 0.50% and 1.00% for welcome flows, and 1.00% and 2.00% for re-engagement. Those are illustrative; derive yours from the percentiles above. When something crosses a line, log the segment, subject line, offer, delivered count, the three rates, your hypothesis, and what you changed. Six months of that beats six months of opinions.

Offer a middle option. Plenty of subscribers don't want to leave, they want less: a monthly digest instead of a weekly send, product updates only, or a 30-day pause. A preference center converts some of those people back into subscribers instead of losing them, and it costs less than winning a replacement.

What you should not do is make leaving harder. Hiding the link suppresses visible unsubscribes for about a week and raises complaints permanently, which is a bad trade at any volume. Keep the raw unsubscribe events for troubleshooting even though you report on unique contacts; repeated clicks from the same person usually mean a broken preference page rather than an emphatic subscriber, and every confirmed opt-out belongs on the suppression list before the next import can undo it.

Purely transactional messages such as receipts and password resets generally don't need a marketing unsubscribe link, since they carry only service information. Add promotional content to them and the rules change, so get advice for your market before you blur that line.