Inez Zimakowski.
Inez Zimakowski, Head of Digital & Media at Connecting Plots Group
Once the standard metric for measuring the success of an email campaign, Open Rates have taken a few hits over the past five years, becoming the ultimate vanity metric.
Marketers love seeing a 40% open rate on a slide, but hate when it leads to 0% revenue growth. The same technology that improves user experience and enhances email security now distorts how opens are recorded, making the 40% datapoint a big fat lie.
Privacy updates, AI-driven summaries, and increasingly aggressive security filters have reshaped how emails are processed long before a human ever sees them.
Apple kicked things off with privacy updates
Of course, we have the Apple inflation effect that kicked off with the introduction of Apple’s Mail Privacy Protection in 2021. This is when Apple Mail started pre-loading email content (including tracking pixels) on their own proxy servers before the user even looks at it. That means if a large portion of a brand’s database has their email hooked up to the mail app on their iPhone, that pre-load will skew open rates.
AI ghosts are opening emails without genuine interaction
With the explosion of AI, we now have AI summaries to factor in. With Gemini and Apple Intelligence on the rise, the inbox has changed, and again, open rate accuracy takes the hit.
AIs now scan a person’s emails to generate a summary or ‘promotional snapshot’. This scanning looks at the first 100 characters of the email body (over click-baity subject lines) and often requires the email to be opened by a bot to process the text.
For user experience, this is great; they get a general idea of your offer or content (which is especially handy for discount codes), but they never have to click into your email.
Security filters are giving false positives and false negatives
If you’re B2B, good luck getting through all the upgraded security filters.
Systems like Mimecast and Proofpoint are getting more and more aggressive, and these filters automatically click links and open images to check for phishing or malware. What’s the impact? Up to 60% of recorded engagements are from bots, and that A/B test you just ran? Wasted.
On the flip side, some organisations have image-blocking that prevents opens being recorded. Many users (and widely-used clients like Outlook) have ‘Block images by default’ turned on. Since tracking pixels are tiny 1x1 images, if the images don't load, the open isn't recorded.
In the end, we have false positives and false negatives, rendering the metric quite genuinely useless. It’s not just wrong; it’s inconsistently wrong. No proxies for you!
The way forward: Replace open rate with a new metric
Now that I’ve doom-and-gloomed you, here’s an easy-to-implement new metric to replace open rate. I like to think of it as an Enhanced Engagement Rate, or more specifically: Active Response Rate.
Typically, engagement rate for email is the classic CTOR (clicks / opens). But if opens are redundant, this bad boy is too.
Instead, use total clicks / total number of email sends. It’s a simple swap that allows you to track engagement over time and can be implemented retrospectively to create a pre-Apple, pre-AI baseline. Of course, revenue per email, unsubscribe rate, and conversion rate are all still relevant; just swap opens for sends. Percentages will go down as the ratio increases, but you’ll finally have consistency in measurement.
Sources: Apple MPP: Litmus, "2024 State of Email" (Confirms market share and pre-loading mechanics). AI Summaries: MediaCat UK, "How AI is changing the Inbox" (Discusses the rise of "Zero-click" email consumption). Bot Activity: Stoneshot, "The Truth About Bot Clicks in B2B" (Provides the ~60% figure for high-security environments). Image Blocking: Campaign Monitor, "The Impact of Image Blocking on Tracking".