This post may contain affiliate links. If you buy through one, we may earn a commission at no extra cost to you. See our Affiliate Disclosure for details.
A newsletter goes out to 1,400 people on a Thursday. By Friday the dashboard says 61% opened it. Two people clicked the link, nobody replied, and four unsubscribed. Which of those numbers describes what actually happened? Not the big one. Email open rates have been quietly broken since 2021, and a one-person newsletter that still steers by them is steering by a gauge that reads high no matter what the engine is doing.
Treat the open rate as a rough trend line at best and never as a per-send verdict. Apple’s Mail Privacy Protection preloads tracking pixels whether or not anyone reads the email, so a large share of “opens” are the device, not the person. Judge each send on four things your platform already records: unique clicks on the one link you cared about, replies, unsubscribes, and the spam-complaint rate reported in Google Postmaster Tools. If the email exists to sell something, add a UTM-tagged link and read the result in your analytics tool rather than in the email dashboard.
What actually broke in 2021
An open is measured by a one-pixel image in the email. When a mail client downloads that image, the sender's platform records an open. That mechanism was always leaky — plenty of clients blocked images by default, so real readers went uncounted — but it leaked in one direction, and people learned to live with an undercount.
Then Apple shipped Mail Privacy Protection with iOS 15 in the autumn of 2021. Apple's own privacy page describes the feature plainly: it hides the reader's IP address and loads remote content privately in the background. In practice, Apple Mail fetches the tracking pixel on the device's schedule, not the reader's. The email gets "opened" in the sender's dashboard whether it was read carefully, glanced at, or deleted unread from the lock screen.
Gmail runs images through its own proxy too, which strips location and device data even when the open itself is genuine. And corporate mail security tools scan incoming messages before delivery, fetching images and following links to check them — which registers as an open, and sometimes a click, from an address that hasn't seen the email yet.
The result isn't a slightly noisy number. It's a number where you can't tell which portion is human, and that portion changes with the mix of mail clients on your list. A list heavy with iPhone users looks healthier than a list of Outlook users, for reasons that have nothing to do with the writing.
Clicks are better, with one asterisk
A click needs a finger on a link, so it survives the pixel problem. The asterisk is those same security scanners: Microsoft's Safe Links and similar products open URLs on the recipient's behalf to check for malware, and that can show up as a click seconds after delivery, often on every link in the email at once.
Two habits make click data trustworthy enough for a small list.
Count unique clicks on the main link, not total clicks across the email. A newsletter with eight links spreads attention thin; the one link the send was built around is the one whose number tells you whether the email worked.
Then look at the timing. Human clicks trail in over hours and days. A cluster of clicks on every link in the first sixty seconds, from addresses that never click otherwise, is a scanner. Most platforms filter the obvious cases now, and none catch all of them.
A click needs a person. An open only needs a device. Photo via Pexels.
If a click is supposed to lead somewhere that matters — a booking page, a product, a signup — tag the link with UTM parameters using Google's Campaign URL Builder and read the sessions in your analytics tool. That cross-check settles the scanner question, because a scanner fetching a URL doesn't behave like a person landing on a page and scrolling.
The metric nobody puts on a dashboard
Replies. A reply is the only signal in email that cannot be faked by a client, a proxy, or a scanner, and it's the one most newsletter dashboards don't show at all, because replies land in your inbox instead of the platform.
For a list of a few hundred or a few thousand, replies also happen to be the most useful feedback you'll get. Three people writing back to say "this is exactly my situation" tells you more about the next month's topics than any percentage. Mailbox providers notice replies too — a recipient answering a sender is a strong hint that the mail is wanted, which feeds back into where future sends land.
The practical version: send from an address you read, ask one real question every few sends, and keep a tally. A newsletter that gets replies is working. One that gets opens might be.
Two numbers that protect the list
Unsubscribes and spam complaints are the metrics people avoid looking at, and they're the ones with teeth.
Unsubscribes per send is a health check on relevance. A steady trickle is normal and healthy; someone who no longer wants the email is doing you a favour by leaving instead of ignoring you. A spike after one particular send is information about that send — usually the topic or the frequency, occasionally the subject line overpromised.
Spam complaints are the one metric with an external threshold attached. Google's sender guidelines ask bulk senders to keep the complaint rate shown in Postmaster Tools below 0.3%, and to stay well under it. That figure sits in a free tool that most solo senders have never opened, which is a strange state of affairs given that it's the closest thing to an official grade your sending gets. Setting it up takes a DNS record and ten minutes. The wider deliverability fixes for a solo newsletter start there.
Net list growth rounds this out: signups minus unsubscribes minus bounces, per month. A list that gains forty and loses forty-five is shrinking, and the open rate will never tell you.
When opens still have a job
Not zero. The open rate is still usable as a trend line on the same list over months, because the inflation is roughly constant as long as the client mix doesn't shift. A send that scores far below your own running average probably did land somewhere bad — a subject line that tripped a filter, a deliverability problem, a broken send. Just don't read the absolute number as "61% of people saw this."
Where opens are now a poor input is anything automated that depends on them. Send-time optimisation learns from when people open; Apple's prefetch feeds it the device's schedule instead. Engagement segments that define "active" by opens will quietly count every iPhone subscriber as engaged forever. If you're building the kind of behaviour-based groups that an email list segmentation setup produces, define engagement by clicks and replies, and use opens only as a tiebreaker.
Four numbers per send, written down, beats a dashboard you've stopped believing. Photo by freestocks.org via Pexels.
A per-send scorecard that fits on one line
| Metric | Where it comes from | What a bad number looks like |
|---|---|---|
| Unique clicks on the main link | Email platform report | Far below your own average, or a burst in the first minute from every link |
| Replies | Your inbox | Zero, send after send, when you've asked a question |
| Unsubscribes | Email platform report | A spike tied to one send rather than a steady trickle |
| Spam complaint rate | Google Postmaster Tools | Approaching 0.3% on Gmail addresses |
| Sessions or conversions from the email | Analytics tool via UTM links | Clicks recorded, but no matching sessions on the page |
Every mainstream platform reports the first three. As of September 2026, which reports come with which plan varies — GetResponse, MailerLite, Brevo and the rest publish this on their own pricing pages, and it changes without announcement, so read the page rather than any blog's summary of it. Postmaster Tools is free from Google regardless of platform. The analytics column depends on whatever already runs on your site.
The whole scorecard takes about two minutes per send to fill in. That is the point of keeping it to five rows.
Do this for your last send
Open the report for the most recent newsletter you sent. Ignore the open rate entirely. Write down four things: unique clicks on the main link, how many replies you got, how many people unsubscribed, and — if you have it — the complaint rate for that day in Postmaster Tools. If you don't have Postmaster Tools yet, adding your domain is the more valuable ten minutes.
Then do the same for the send before it. Two rows is enough to see whether the numbers you've been ignoring tell a different story from the email open rates you've been reporting to yourself — and for most small lists, they do.
저는 한국에서 직원 없이 혼자서 두 개의 소규모 사업체를 운영하고 있습니다. 견적, 회계, 마케팅, 웹사이트 관리까지 모두 제가 직접 하고 있습니다. 개발자는 아니지만, 이 사이트의 자동화 시스템은 AI 코딩 도우미를 활용하여 밤늦게까지 오랜 시간 공들여 구축했고, 모든 콘텐츠 관리 및 게시도 제가 직접 담당하고 있습니다.