Issue #242 All About Email

Hi, {{first_name|friend}}. 👋

Welcome to Issue #242 of All About Email!

Last week was the Email Markup Consortium's 2026 Accessibility Report, with a 99.88% failure rate that the industry really can't keep ignoring (but probably will 😢).

This week, I want to pick up a thread that's been running through several recent issues and take a closer look at it. Because if opens are unreliable and clicks are increasingly complicated, the question is: “What do you actually measure?”

Let’s go! 👇

🧠 This isn't a new problem. But in 2026, though, we’re dealing with three distinct layers of decline occurring simultaneously, and together they have a real impact that warrants closer exploration.

If you've been reading this newsletter for a while, you know this one. Apple Mail Privacy Protection, which we covered in detail back in Issue #190, pre-loads tracking pixels before a human ever reads the email.

💡 A 2024 Validity study found open rates climbing 18–32% above verified engagement benchmarks for senders with Apple Mail-dominant audiences, and higher still for those where Apple Mail accounts for the majority of opens

Open rates still have directional value. But as a precise measure of who actually read your email? They haven't been that for over four years.

💡 In Issue #237, I spent a whole issue unpacking why real, engaged subscribers sometimes never click anything. But there's an inverse problem worth highlighting too.

Security gateways, antivirus software, and corporate IT systems routinely pre-scan every URL in an email by firing the tracking redirect before any human interaction.

That inflates your click data with bot-generated signals, making engagement look better than it actually is.

🚨 And as I discovered the hard way with my re-engagement automation, some of your most loyal readers may never click a thing, while some of your click "data" never involved a human at all.

High click numbers and low click numbers are both harder to interpret than they used to be. 🫠

🤔 This is the newest layer, and the one I think deserves the most attention right now.

As we covered in Issue #240, Gmail's AI Inbox, Apple Intelligence, and Gemini Daily Brief are now summarising, triaging, and in some cases acting on emails before a human consciously reads them. Gmail can mark an email as "done" without the subscriber opening it at all.

🚨 An email intercepted, summarised, and actioned by AI doesn't register as an open or a click. But it was received. It was read. It may have prompted a response. And none of that appears in your dashboard.

🧠 Here's the thing: opens and clicks were never perfect. We used them because they were accessible, automated, and good enough for most purposes.

The problem is that we built a lot on top of them:

💡 This isn't about abandoning metrics entirely.

It's about understanding what each number can and can't tell you, and shifting from passive tracking to active signals.

🚨 That's the real distinction worth making:

✅ The hardest signal to fake, and one of the most underused metrics in email marketing.

A reply requires deliberate human action. No bot fires one, no AI pre-loads one, no privacy tool blocks it. A reply is unambiguously real (…at least for now, who knows what will be automated with AI in the future 🤷‍♂️).

For that reason alone, it's worth tracking even if your platform doesn't surface it prominently.

💡 For anyone selling anything through email, products, services, consultancy, or sponsorships, this is the number that connects your email programme to the actual business.

Not a vanity metric. A smaller, more engaged list with high revenue per subscriber beats a large list with impressive open rates every single time.

🤔 If you're a newsletter publisher, this might be sponsorship revenue attributed per send. If you're selling a product, it's tracked via email conversions. The mechanism varies, the principle doesn't.

🚨 This framing has been gaining traction in the industry, and I think it's a useful one.

Rather than chasing positive signals that might be partially false, look at the negative ones, which are considerably harder to inflate.

🧠 The concept was first coined by Melinda Krueger in 2005 as a smarter way to look at unsubscribe rate. In its original form, it's a ratio: unsubscribes divided by unique clicks, which tells you how many people clicked on your email solely to leave.

The 2026 interpretation, from a Validity report covered by MarTech, broadens that into a composite view:

💡 What matters is tracking it consistently over time rather than in isolation. A single high-unsubscribe send might mean nothing. A composite that's been quietly climbing across multiple issues is telling you something real.

🧠 This matters because mailbox providers increasingly treat negative signals as stronger indicators than positive ones when assessing sender reputation. If your disaffection composite is flat or declining over time, your programme is likely healthy, regardless of what your open rate says.

🚨 One important caveat to note: unsubscribes are not entirely immune to automation.

Security scanners and antivirus tools are documented to pre-click links in emails, including unsubscribe links, which can generate false unsubscribe events, particularly for B2B audiences behind corporate security gateways.

Spam complaints made within the mailbox interface are a cleaner signal, since they require deliberate action in the email client rather than clicking a URL.

If your unsubscribe rate spikes suddenly across a single send to a B2B segment, it's worth ruling out bot activity before drawing conclusions.

💡 B2B is not really my wheelhouse, so if I’ve made any mistakes here, let me know, {{first_name|friend}}.

🧠 Gross new subscribers minus churn, tracked over time.

A list that's growing after accounting for people leaving tells you something meaningful about the health of your email programme overall.

A shrinking list, despite strong "open rates", is trying to tell you something your dashboard might be hiding.

💡 Neither the click number nor the open number is clean in isolation. But the ratio between them can be really helpful.

If someone both opened and clicked, at least one of those events is more likely to have involved a human.

CTOR isn't a perfect signal, but it's better than raw click rate alone, and it's a useful way to evaluate content effectiveness within a campaign without being thrown off by inflated open rates.

✅ A subscriber who answers a question has given you something explicit, intentional, and the zero-party data we need.

The "hit reply" prompt at the end of every issue is a version of this. It's also, as it turns out, one of the most useful things I do in this newsletter, {{first_name|friend}}.

🎉 One of my RSS subscribers sent me two lovely replies to last week’s newsletter, and I’ve included one of his links in this week’s “News & Tips” section.

🚨 I want to be clear about something. None of this means opens and clicks are worthless, or that you should rip your reporting apart tomorrow.

It means using them for what they're actually good for, not for what we've been pretending they are.

🤔 What they're not useful for: precise conclusions about individual subscriber behaviour, automated decisions about who to suppress, or reporting that implies a level of certainty the data doesn't support.

🚨 The aim is not to create a flawless metric. Instead, we want a more accurate view that comes from considering several signals, rather than relying on a single, convenient number.

That's it for this week.

As always, {{first_name|friend}}, let me know your thoughts. 👋

Every week, as I write this newsletter, I'll share the track of the moment to create an unbelievably eclectic playlist just for your inbox.

🚨 If you’re interested in sponsoring the “All About Email” newsletter, you can find all the details in this Google Doc.

This week's excellent and insightful email news & tips:

** 🎩 Hat tip to James, one of my subscribers via RSS, for suggesting this one.

If you have any questions about this email or email marketing, please reply, and I will get back to you as soon as possible.

I hope you have a great week! 👋