Open rate is the first number everyone checks. It’s on every campaign dashboard, the first stat in every outreach retrospective, and the metric most founders use to decide whether a subject line worked.
It’s also been broken since September 2021.
What Apple Did to Your Tracking Pixel
iOS 15 shipped with Mail Privacy Protection enabled by default. Here’s what it does: when an email arrives in Apple Mail, Apple’s proxy servers pre-fetch all remote content in the message — including your 1×1 tracking pixel — before the recipient opens the app. Your platform records an “open.” The person may never look at the email.
This isn’t an edge case. Apple Mail has roughly 55–60% of the mobile email client market. Add macOS Mail and iPad Mail, and you’re looking at a large share of your list being processed through Apple’s servers before any human eyeball touches the message.
Your tracking pixel fires. The open gets logged. The recipient is in a meeting, or asleep, or hasn’t touched their phone. Your dashboard shows engagement that doesn’t exist.
How Much of Your Open Rate Is Actually Noise?
It depends on your audience — but the numbers aren’t pretty.
Consumer-heavy lists (personal Gmail accounts used from iPhones, early-stage founders on Macs) can see inflation above 40%. B2B enterprise lists that run on Outlook are less affected, since Outlook doesn’t pre-fetch pixels by default. But a typical cold outreach list — solo founders, agency owners, small business contacts — skews heavily Apple.
Working estimate: 30–40% of your reported opens are Apple proxy opens, not real engagement.
If your last campaign showed a 48% open rate, the real number might be closer to 29–34%. That’s the difference between “this subject line crushed it” and “this subject line was mediocre and we got lucky on replies.”
Some tools have introduced “bot filtering” — logic that tries to distinguish proxy opens from human opens based on timing patterns, device signals, and IP ranges. These filters help but they’re estimates, not measurements. A filtered open rate is better than an unfiltered one. It’s still not accurate.
The Decisions You’re Getting Wrong Right Now
When your open rate is inflated, every decision downstream gets distorted.
Subject line testing. You run an A/B test. Version A gets 42% opens, Version B gets 35%. You declare A the winner and roll it out. But if 35% of both those numbers are phantom opens, the real spread might be within statistical noise. You made a campaign-level decision based on fake data.
Follow-up timing. Your sequence triggers a second email after “no open in 3 days.” But many of those contacts had their open recorded by Apple’s proxy on day one. You’re sending follow-ups to people who technically “opened” — which means your sequence isn’t working the way you designed it. Contacts who never engaged get a different message than what you intended.
Campaign evaluation. Two campaigns, different angles. Campaign A has higher opens, Campaign B has higher replies. You optimize for opens, you pick the wrong one. Replies are real. Most opens aren’t.
List suppression. You clean your list by suppressing contacts with zero opens in six months. Some of those contacts are on Outlook or Gmail web and legitimately never triggered Apple’s proxy. Their opens aren’t recorded not because they’re disengaged, but because their client doesn’t inflate the count. You’re suppressing real prospects because the absence of a fake signal looks like disengagement.
Three Signals You Can Actually Trust
None of these are affected by Apple Mail Privacy Protection. All of them require a deliberate human action.
Clicks
When someone clicks a link, they made a choice. They moved their cursor or finger to a specific element and pressed it. Apple’s proxy doesn’t do that. Clicks are the first clean signal in a cold sequence.
A low click rate on a “high open” campaign is one of the clearest signs of inflation. The clicks reveal what the opens were hiding.
Track clicks on every link in every email — your CTA, your calendar link, your website URL. Click rate is a better subject-line proxy than open rate, because it means the person at minimum read far enough to reach the link.
Replies
A reply is the purest signal in cold email. The prospect read enough of your message to respond, decided it was worth their time, and wrote words back. Reply rate is your north star — every other metric is a proxy for this one.
If you’re currently optimizing campaigns for open rate, stop. Optimize for reply rate. Rewrite your sequences with one goal: get a reply, not a recorded open. The framing change sounds subtle. The downstream impact on your copy and strategy is not.
IMAP Reply Detection
Less discussed but operationally important: when someone replies to your cold email, their response lands in your sending account’s inbox — not necessarily in your platform’s reply tracker.
By scanning that inbox via IMAP and matching In-Reply-To headers against message IDs you sent, you can catch every reply regardless of routing. This matters for senders who use their own Google Workspace accounts, since Gmail routes replies differently depending on whether you’re using Gmail SMTP directly or a third-party relay.
PitchGale scans every connected sending account this way. Replies that arrive in the account’s inbox but bypass the platform tracking layer still get detected and logged. No reply slips through.
How to Rebuild Your Strategy Around Real Signals
First, demote open rate. Stop using it as a primary performance indicator. At best, use it as a directional signal when comparing similar time windows and list segments — the inflation is somewhat consistent across periods, so relative changes might still carry weak signal. But don’t make copy or strategy decisions based on it.
Second, fix your follow-up conditions. Replace “no open in 3 days” with “no reply in 3 days.” This is a one-line change in your sequence logic that immediately makes your follow-up behavior more accurate.
Third, evaluate subject lines by reply rate, not open rate. Split your list, run the test, and look at which subject line generated more replies. It takes more volume to reach statistical significance, but you’re measuring something real.
Fourth, be careful with open-based list suppression. Use bounce data, unsubscribes, and reply behavior to manage list hygiene instead. A contact who never “opened” but whose email hasn’t hard-bounced is worth a re-engagement attempt before you suppress them.
Fifth, look at click-to-reply ratio. If someone clicked your calendar link but didn’t reply, something in the handoff broke. If your click rate is solid but reply rate is low, the problem is downstream of the email — the landing page, the calendar booking flow, the offer itself.
The Metric That Never Lies
Reply rate. That’s it.
A reply means a person read your message, found it relevant enough to respond, and spent time writing back. No proxy server does that. No bot does that. A reply is a human decision, and it’s the only metric in cold outreach that directly predicts revenue.
Build every part of your outreach system — your copy, your sequences, your subject line tests, your list hygiene — around maximizing genuine replies. Open rate is a noise machine. Stop optimizing for it.