Email Open Rate: What It Measures and Why It's Misleading

Email open rate is the percentage of delivered emails that triggered a tracking pixel load, calculated as unique opens divided by delivered emails times 100. It is not a measure of how many people actually read your email, and since Apple started pre-loading images for Mail Privacy Protection users in 2021, a large share of "opens" happen without a human ever looking at the message.

That gap between what open rate claims to measure and what it actually measures is why so many SaaS teams misread their own numbers, celebrating a 40% open rate on a list that is half asleep, or panicking over a "drop" that is really just an image-blocking setting.

Key takeaways

• Open rate = unique opens / delivered emails x 100. It is triggered by a tracking pixel loading, not by a person reading the email.
• Apple Mail Privacy Protection (MPP) pre-fetches images for opted-in users, which can push open rates toward 100% for that segment regardless of real engagement.
• Mailchimp's own benchmark data puts the average open rate across all industries at 21.33%, but treat any single "average" as directional, not a target, since inflation varies by how Apple-heavy your list is.
• Click-through rate and click-to-open rate are more resistant to pixel-based distortion, though CTOR still leans on open counts as its denominator.
• The fix is not to ignore open rate. It is to separate genuine opens from automated ones and to pair open data with click and conversion metrics before you make a send decision.
• A resend-to-non-openers campaign still works, but only if your platform can tell a real non-opener from an Apple MPP false negative.

What email open rate actually measures

Every marketing email contains a tiny, invisible 1x1 pixel image hosted on the sending platform's servers. When the recipient's email client renders that image, the platform logs an "open." That is the entire mechanism. It says nothing about whether the person read a word, scrolled past it in a preview pane, or had their client fetch it automatically in the background.

The standard formula:

Open Rate = (Unique Opens / Emails Delivered) x 100

Delivered emails means total sent minus hard bounces. Some platforms report "total opens" (counting the same person opening twice) separately from "unique opens" (one person counted once no matter how many times they reopen). Always compare unique open rate to unique open rate. Mixing total and unique numbers is the single most common way two teams argue over "different" benchmarks that are actually the same list.

Why the pixel fails in both directions

The pixel over-counts in some cases and under-counts in others:

• Over-counts: Apple Mail Privacy Protection, corporate security scanners (Mimecast, Proofpoint, Microsoft Defender) that pre-fetch links and images to check for malware, and any email client that preloads images by default.
• Under-counts: Recipients who have images turned off (common on older Outlook desktop defaults), plain-text email clients, and Gmail's behavior of clipping messages over roughly 102KB, where the tracking pixel at the bottom of a long email may never load unless the recipient clicks "View entire message."

Neither failure is rare. If your list has any meaningful share of Apple Mail users (frequently 40 to 60% for US consumer or founder-heavy B2B lists), your reported open rate is almost certainly inflated. If your list skews enterprise or security-conscious, both inflation and deflation can be happening in the same send.

Why Apple Mail Privacy Protection broke the metric

Apple rolled out Mail Privacy Protection in September 2021 with iOS 15, iPadOS 15, and macOS Monterey. When a user turns it on (Apple made it opt-out, not opt-in, so most Apple Mail users have it enabled by default), Apple does not deliver the email directly. Instead, Apple routes it through a proxy server that pre-loads all images, including the tracking pixel, whether or not the person ever opens the message.

The practical effect: for any recipient using Apple Mail with MPP enabled, your platform can log an "open" the moment the email hits the proxy, sometimes before it even reaches the inbox. Multiple industry analyses documented open rates for Apple Mail segments approaching 100% after MPP rollout, a number that reflects pixel pre-fetching, not reading behavior.

This has two downstream effects worth knowing before you build any process around opens:

1. Segmentation by engagement breaks. "Most engaged" and "least engaged" segments built on open history become unreliable, because a large share of your "engaged" contacts may never have looked at a single email.
2. Resend-to-non-openers gets noisier. A resend campaign that targets people who did not open the first send will, by definition, skip anyone MPP marked as an opener even if they never saw it. This does not make resending useless, it just means the audience is smaller and less precise than it looks.

What is a good open rate for email in 2026

There is no single correct number, and anyone who gives you one flat target without asking about your industry and your list's Apple Mail share is oversimplifying. Published benchmark data varies by source and methodology, but Mailchimp's own benchmarks report (calculated across billions of emails sent through its platform) puts the average open rate across all industries at 21.33%, with meaningful spread by sector, government and nonprofit senders trending highest, retail and ecommerce trending lowest.

Use benchmarks as a rough compass, not a scoreboard:

A worked example

Say you send to 10,000 contacts, 9,700 are delivered, and your platform reports 3,200 unique opens. Raw open rate: 3,200 / 9,700 x 100 = 33%. That looks strong against a 21.33% industry average. But if 55% of that list is on Apple Mail with MPP enabled, a large share of those 3,200 opens likely happened without a human present. The number is not wrong, it is just answering a narrower question than "how many people engaged with this email" than most people assume.

What to track instead (or alongside)

Open rate is not worthless, it is incomplete. Pair it with metrics that are harder to fake with a pixel pre-fetch:

• Click-through rate (CTR): total clicks divided by emails delivered. Not affected by MPP, since a click requires an actual person to interact with a link.
• Click-to-open rate (CTOR): unique clicks divided by unique opens, times 100. It tells you how compelling your content was to the people who did open it, but remember its denominator (opens) is still MPP-distorted, so CTOR can look artificially low on Apple-heavy lists even when content quality is fine.
• Conversion rate: the percentage of recipients who completed the action you wanted (signed up, upgraded, booked a call). This is the number that ties email back to revenue.
• True open rate (human vs scanner): some platforms can distinguish a human open from a known scanner or proxy pattern (Apple MPP, Mimecast, Proofpoint, Microsoft Defender) and report both a "true" open rate and a scanner percentage side by side, so you are not guessing how much of your number is inflated.
• Unsubscribe and complaint rate: low opens paired with rising unsubscribes usually means a content or frequency problem, not a tracking problem.

Example: reading a real report

If a broadcast report shows:

Delivered: 9,700
Opens (unique): 3,200 (33.0%)
Scanner opens: 1,050 (10.8% of delivered)
True opens: 2,150 (22.2% of delivered)
Clicks (unique): 410 (4.2% of delivered)
CTOR (true): 19.1%

That is a materially different story than "33% open rate" alone. A 19.1% click-to-open rate against true opens tells you the content resonated with people who genuinely saw it, a stronger signal than the headline open number.

How this changes what you do with the number

• Stop using open rate alone to decide subject line winners on small lists. If your A/B test sample is Apple-heavy, both variants may show similarly inflated opens regardless of which subject line is actually better. Where possible, judge subject line tests on click-through, not open, especially on lists that skew consumer.
• Still send resends to non-openers, but expect a smaller true audience than the raw number suggests. A resend to people who appear not to have opened often lifts opens among people who did not see the first send, even accounting for MPP noise in who counts as a "non-opener."
• Do not fire automation on open events alone. If a sequence step is gated on "contact opened the last email," a false-positive scanner open can advance someone through a lifecycle sequence before they have actually engaged. Gate meaningful automation on clicks or product events instead, and reserve open-based triggers for low-stakes personalization.
• Investigate deliverability before you investigate copy. A collapsing open rate that is not explained by MPP is more often a deliverability or sender reputation problem (emails landing in spam or being throttled) than a subject-line problem. Fix inbox placement first.

A note on sending infrastructure and true opens

Part of why open rate reporting varies so much between platforms comes down to what is under the hood. If you send through your own Amazon SES setup, you get the raw delivery, bounce, and complaint events straight from AWS, but distinguishing a human open from a scanner open still requires the sending platform layered on top to do that analysis, not SES itself. This is one reason SaaS teams comparison-shopping email tools should ask a vendor directly whether they separate true opens from scanner opens, rather than assuming every platform's "open rate" number means the same thing.

Meisa's broadcast reporting does this split explicitly: every broadcast shows a true open rate alongside a scanner percentage, so a founder can see how much of a headline number is Apple MPP or corporate scanner noise before deciding whether a subject line worked or a resend-to-non-openers send is worth queuing.

If you are choosing a platform partly on the strength of its reporting, ask to see a live broadcast report before you commit, not just a demo screenshot. You can try Meisa to see how a true-open breakdown looks against your own list.

FAQ

What is a good open rate for email?

There is no universal number, but published benchmark data (Mailchimp's platform-wide report) puts the all-industry average around 21.33%. Government, nonprofit, and education senders tend to run well above that; retail and ecommerce tend to run below it. Treat any benchmark as a rough comparison point, not a pass/fail target, and always check what share of your list is on Apple Mail before comparing your number to an industry average.

Why did my open rate suddenly jump without any change in my emails?

The most common cause is a shift in your list's Apple Mail Privacy Protection adoption, either because more of your subscribers upgraded their OS or opted into MPP, or because your list composition changed (a new segment or import that is more Apple-heavy). A sudden jump with no content or list change is a tracking artifact, not new engagement, until proven otherwise.

Does Apple Mail Privacy Protection affect click rate too?

No. Click tracking requires the recipient to actually click a link, which MPP does not simulate. Click-through rate and raw click counts remain reliable even on Apple-heavy lists. Click-to-open rate is the exception, since its denominator (opens) is still affected by MPP, so CTOR can look artificially depressed even when clicks themselves are healthy.

Should I stop tracking open rate altogether?

No. Open rate is still useful for spotting deliverability problems (a collapsing open rate across your whole list, not explained by MPP, often points to inbox placement issues) and for rough week-over-week trend-watching. The mistake is treating it as a precise measure of content engagement or as the sole metric for A/B test decisions. Pair it with click-through rate, conversion rate, and, where available, a true-open-vs-scanner breakdown.

How is open rate different from click-to-open rate?

Open rate is opens divided by delivered emails. Click-to-open rate (CTOR) is clicks divided by opens. Open rate tells you roughly how many people's clients loaded your email; CTOR tells you what share of those loads turned into an actual click. CTOR is a better read on content quality, but it inherits any distortion already baked into the open count it is dividing by.