In September 2021, Apple Mail Privacy Protection (MPP) changed the meaning of an email open.
Before MPP: an open event meant a human had opened the email and the tracking pixel fired. After MPP: Apple preloads email content including tracking pixels for users who have enabled privacy protection, which is now the majority of Apple Mail users. An "open" can now be recorded by an Apple server, not a human.
This has a specific and underappreciated consequence for list hygiene strategy. If your programme uses open rate engagement as a proxy for list health to identify active contacts, to trigger re-engagement campaigns, or to decide who to include in sends, your signals are partially corrupted. Contacts who appear active based on open data may not be. Contacts who appear inactive may be reading every email through a non-Apple client that does not fire a tracking pixel.
What Apple MPP Did to Email Open Data
Apple Mail Privacy Protection works by routing email content through Apple's proxy servers before delivering it to the device. The proxy servers preload the email, including tracking pixels, ls regardless of whether the user actually opens and reads it.
The result: your ESP records an open event when Apple's server loads the pixel. The actual human may never have looked at the email. They may have deleted it without reading. They may have never seen it at all if it was filtered before full delivery.
Scale of the Impact
Apple Mail has a significant share of email client usage. As of mid-2026, Apple Mail (combining iPhone, iPad, and Mac clients) accounts for approximately 35–50% of email opens in most B2C programmes, with a lower but still significant share in B2B programmes where corporate email clients are more common.
In consumer programmes where Apple devices are dominant, MPP can inflate open rates by 20–35 percentage points. What appears to be a 45% open rate may have a true human-open rate closer to 20%.
For ESP platforms that have not specifically accounted for MPP either by filtering known Apple proxy IP ranges or by explicitly flagging MPP opens separately, the inflation is invisible in standard reporting.
How MPP Creates a List Hygiene Problem
The standard list hygiene model uses engagement signals to identify active versus inactive contacts. The typical logic is:
Opened or clicked in the past 90 days = active, keep sending
No open or click in 90–180 days = declining engagement, reduce frequency
No open or click in 180–365 days = candidate for re-engagement campaign
No open or click in 365+ days = sunset, suppress
This logic breaks when open signals are MPP-inflated. Contacts who have not genuinely engaged in 18 months may appear to have opened every campaign because Apple's proxy servers have been preloading their email. These contacts would never appear in the re-engagement or sunset bucket under open-based logic.
The Zombie Engagement Problem
A contact enrolled via Apple Mail who has not actively read your email in two years but whose opens are consistently recorded by MPP will never show up on your inactive list. They will be included in every campaign send. They generate no real engagement. They produce no conversions. They consume ESP budget. And if their address eventually becomes invalid, either through address abandonment or account closure, they generate a hard bounce that damages your sender reputation.
This is the zombie engagement problem: contacts that look active in your data but have no genuine relationship with your brand.
The Segments Most Affected by MPP Signal Corruption
Consumer Programmes with High Apple Device Penetration
B2C programmes in retail, lifestyle, fitness, and consumer technology typically have the highest Apple Mail usage rates among subscribers. These programmes are most exposed to MPP inflation. In some consumer programmes, 60–70% of apparent opens may be MPP-generated.
US and UK-Focused Programmes
Apple device market share is highest in the US and UK. Programmes primarily targeting these markets see the most MPP impact. Programmes targeting markets with lower Apple device penetration, such as parts of Asia, Eastern Europe, and Latin America, are less affected.
Programmes That Have Not Updated Their ESP Open Attribution
Not all ESPs have separated MPP opens from genuine opens in their reporting. If your programme's open data does not distinguish between MPP proxy opens and human opens, your engagement segmentation is based on mixed signal quality.
How to Identify MPP-Inflated Open Data in Your Programme
Check Your ESP's MPP Reporting
Most major ESPs Klaviyo, Mailchimp, Salesforce Marketing Cloud, HubSpot, Braze now provide separate tracking for MPP-suspected opens. In Klaviyo, for example, you can see "Machine Open" events distinct from "Email Opened" events. Review your programme's MPP proportion in your ESP's reporting.
If MPP opens represent more than 25% of your total open events, your engagement segmentation is meaningfully affected.
Build a Click-Only Engagement View
Create a report segment showing contacts who have clicked, not just opened, in the past 12 months. Compare this to your open-based engagement view for the same period. The difference between these two populations represents the contacts that appear engaged based on opens but have no confirmed click-through intent signal.
The click-based view is your more reliable picture of genuinely active contacts. Open-based views include MPP inflation.
Run Email Verification on Your "Inactive" Segment
With MPP-inflated opens, your inactive segment contacts with no opens and no clicks may actually be smaller than it should be. Contacts who appear active due to MPP will not appear there. But those that do appear are genuinely at risk.
Run email verification on your click-inactive segment (contacts with no clicks in 12 months, regardless of open data). The invalid rate in this segment tells you more about true list health than the open-rate-defined inactive segment.
Rebuilding Suppression Logic Without Open Rate Dependency
The solution is not to ignore open data entirely; it still contains useful signal, even if inflated. The solution is to use click data as the primary engagement signal and treat open data as a secondary, noise-adjusted signal.
Click-First Suppression Framework
Define contact activity tiers based primarily on click behaviour:
Active: Clicked in the past 90 days
Cooling: Clicked 91–180 days ago (reduce frequency)
Inactive: Clicked 181–365 days ago (re-engagement campaign trigger)
Dormant: No click in 365+ days (sunset candidate regardless of open data)
This framework suppresses contacts based on the most reliable engagement signal and avoids keeping zombie-engaged contacts active indefinitely based on MPP-inflated opens.
Supplemental Open Signal Usage
Use open data to adjust confidence within click tiers. A contact who has clicked twice and opened 15 times in the past 90 days is more engaged than a contact who has clicked twice with only 2 opens. But the primary suppression trigger should be click-based, not open-based.
Click-Based and Behaviour-Based Alternatives to Open Tracking
For programmes where clicks are infrequent newsletters with high content value but low CTA volume, for example,e click-only engagement tracking underrepresents genuine interest. In these cases, supplement with:
Website visit tracking: If a subscriber clicks through and visits your site, that is a confirmed engagement signal that does not depend on email open attribution. Track post-click site behaviour and incorporate it into engagement scoring.
Reply rate: Email replies are the strongest unambiguous engagement signal. A contact who replies to one email per quarter is demonstrably more engaged than a contact who appears to open every email but never clicks or replies. Segment your active contacts by reply history.
Purchase and transaction signals: For e-commerce and SaaS programmes, a purchase or active product use is a stronger engagement signal than an open event. Include transactional signals in your suppression logic.
Preference centre interaction: A contact who has updated their email preferences in the past year is demonstrably active and monitoring their inbox for your emails. Preference centre interactions are a reliable supplemental signal.
What This Means for Re-Engagement Campaigns?
The standard re-engagement campaign targets contacts who have not opened or clicked in a defined window. With MPP inflation, contacts who have not genuinely engaged will often not appear in this bucket yet; they will appear to be opening, even if they have not truly read the email in over a year.
This means re-engagement campaigns built on open-based inactive segments will be missing a significant proportion of the contacts who actually need re-engagement. Your re-engagement campaign is targeting the wrong people.
Rebuilding Your Re-Engagement Trigger
Rebuild your re-engagement trigger based on click inactivity rather than open inactivity. Target contacts who have not clicked in 180 days, regardless of their open data. This will produce a much larger re-engagement audience than the open-based definition because it will capture the MPP-inflated zombies who appear active on paper.
Before deploying the re-engagement campaign to this expanded audience, run email verification on the full segment. A significant proportion of the truly inactive contacts in this group will have decayed to invalid addresses in the intervening months. Remove invalids before sending; the re-engagement audience you actually need to reach is the valid, non-click-engaged population, not the bounced-address cohort.
Key Takeaways
Apple MPP preloads email content through proxy servers, recording open events without human action. This inflates open rates by 20–35 percentage points in consumer-focused programmes and corrupts engagement segmentation built on open data.
The primary list hygiene consequence is zombie engagement: contacts that appear active based on MPP-inflated opens but have no genuine relationship with the brand. These contacts accumulate without appearing in re-engagement or sunset buckets.
Identify your MPP exposure through your ESP's machine open reporting. Build a click-only engagement view and compare it to your open-based view to see the gap.
Rebuild suppression logic on click-first frameworks. Use open data as a supplementary signal only. Define activity tiers: active, cooling, inactive, dormant based on click history.
Re-engagement campaigns built on open-based inactive segments are missing the zombie engagement population. Re-build re-engagement triggers on click inactivity to capture the full disengaged audience. Then verify that audience before sending.
Supplement click tracking with website visit data, reply rates, transaction signals, and preference centre interactions for programmes where click frequency is inherently low.
Frequently Asked Questions
Is Apple MPP only a problem for B2C programmes?
Primarily yes, but not exclusively. B2B programmes with significant proportions of contacts using Apple Mail (common in creative industries, startups, and US-based professional services) also see MPP inflation. The impact is smaller in corporate environments where Microsoft Outlook dominates, but any programme with a 15%+ Apple Mail share should audit its open data for MPP influence.
Can I filter out MPP opens in my ESP reporting?
Most major ESPs now provide some level of MPP filtering either by flagging opens from known Apple proxy IP ranges or by surfacing "machine open" versus "human open" data. Check your ESP's specific reporting capabilities. If your ESP does not distinguish MPP opens from genuine opens, consider a third-party email analytics tool like Litmus or Email on Acid that provides client-level attribution.
How does MPP affect my unsubscribe rate?
It does not directly. Unsubscribe rates are based on clicks on the unsubscribe link, which MPP does not inflate. However, MPP may suppress unsubscribes from contacts who are not genuinely reading the email and therefore never see the unsubscribe option. These contacts stay on the list as apparent openers rather than converting to unsubscribes or being sunset. The result is a smaller-than-appropriate unsubscribe rate and a larger-than-appropriate active list size.
Conclusion
Apple MPP is not a temporary inconvenience. It represents a permanent shift in the signal quality of email open data, and it has been in place long enough that programmes which have not adapted their list hygiene logic are now operating with systematically incorrect engagement data.
The fix is not complicated. Move your primary engagement metric from opens to clicks. Rebuild your suppression, re-engagement, and sunset logic on click-based tiers. Use email verification to confirm list health for segments that open-based logic has been classifying incorrectly.
A clean list, in the post-MPP world, is one maintained on signals that mean what they appear to mean. Open rate is no longer that signal.
Start Verifying TODAY!
Top comments (0)