Email Marketing: It’s a Question Of Value

Updated May 2026

In the old days of marketing, the very fact that a business had a prospect’s email address equated to having their attention.

Not any more.

If your email inbox is anything like mine, barely a day goes by when you don’t receive some kind of marketing circular. We’re all constantly bombarded with so-called “special” offers, company or product news updates, and sales pitches, all continually fighting for our attention. As a result we’re increasingly desensitized to most of the marketing messages that businesses send out. More often than not we look at the name of the email sender, perhaps read the subject headline, only to bin the email. The bin itself has got much smarter, too. Modern inbox apps now show AI-generated previews that summarize the email before we open it, which means we’re often deciding whether something’s worth our attention based on a machine’s reading of the message rather than our own. The marketing team may well have agonized for a week over what subject line to use. But what the recipient actually reads is a sentence written by an algorithm describing what the algorithm thinks the email is about.

The bombardment has worsened since generative AI handed every marketing department and every outright scammer the same cheap way to manufacture infinite variations of the same pitch, with the result that more than half of all email spam is now AI-generated, and a sizeable chunk of it slips past the filters because it reads as plausible English rather than the broken, all-caps Nigerian-prince stuff we trained ourselves to ignore twenty years ago. Volume has gone up and the grammar has improved. But what hasn’t shifted at all is the intent behind any of it.

Playing the numbers game

Marketers look at the problem as being one of numbers. Increase the number of people receiving the communication to increase the number of take-ups of the offer. They talk about “open rates”, meaning the number of people who view (or “open”) the email as a percentage of the number of emails that are sent out.

Since we’re all getting increasingly fed up receiving such mails, it won’t come as a surprise to anyone that real “engagement” with marketing email has been in continual decline. What’s maybe more surprising is that the headline open-rate number no longer shows it, because Apple’s Mail Privacy Protection now pre-fetches images on roughly half of all opens, inflating reported open rates by anywhere from 15% to 35%. Your campaign report may well show a 42% open rate in that sexy-looking browser dashboard, but the actual number of human beings who looked at the email might be closer to 12%. A metric the entire industry built its validation around turned into sludge, yet many people calling themselves “marketers” are still presenting it to their boards as if it meant something.

There’s a knock-on effect of all of this, in that subject-line A/B testing now measures how well our chosen email subject line triggers Apple’s proxy server, rather than how well it caught a human’s attention. Consequently, campaigns get optimized against a target that no longer correlates with what the campaign was supposed to achieve. List segments get rewarded or pruned on the basis of who Apple Mail happens to send through its privacy layer.

Meanwhile the AI-driven volume keeps climbing, and the deliverability picture looks like what you’d find at the bottom of a birdcage. Inbox providers know perfectly well that half their inbound traffic is machine-generated marketing slop, so they’ve responded the way you’d expect them to, with filters that get stricter by the quarter. But those filters are themselves now AI-driven, which has set up an arms race nobody from the marketing department seems to have noticed they’re losing. Models on the sender side generate content designed to look like real human writing. Models on the receiver side classify content based on whether it looks like real human writing. As quickly as a sender adjusts their parameters, receivers counter with ever more devious solutions, because they have orders of magnitude more training data. The net result is that inbox placement for high-volume senders dropped over 22% in the first quarter of 2025 alone, and roughly one in six legitimate marketing emails now never reaches the recipient’s inbox at all. So your sender list grows, open rates look healthier than they are in reality, and all the while a substantial percentage of the audience you think you reached never actually receives the email.

Email marketers will try to con us into believing that such shenanigans is normal. But I’d contend that it’s a sign that corporate marketing, as least with email, has comprehensively lost the attention of the people it’s trying to communicate with, and is now talking past a wall of filters into the abyss.

An abuse of trust

I think that many people on email distribution lists feel they’ve been conned. They feel they were forced into giving their contact details in order to receive something of a perceived value, and what they got in return was a relationship of a wholly different shape than the one they signed up for.

Now they’re receiving overly-frequent, vapid, untargeted communications that they respond to by ignoring. Fancy graphics, killer copywriting, and an ever-more compelling value offering now only go part of the way, because the underlying problem is no longer about factors like design, presentation, or even content. It’s the cumulative weight of a mountain of excrement landing in the same inbox at the same time, with so much of it generated by a machine that’s been pointed at the recipient’s first name in a CRM field and prompted to “personalize at scale.”

The continued collapse of email effectiveness due to lazy over-communication isn’t going away, given that 70% of consumers have unsubscribed from at least three brands in the past three months because of excessive messaging, and 78% say they’d prefer fewer, more targeted communications over the current volume. Every list-build exercise that focuses on subscriber growth as the headline metric is feeding a pool of people who are already most of the way to the unsubscribe button.

AI is both the victim and the aggressor in this story. The tools that allow a sales team to send 10,000 individually-named cold emails before lunch are the same tools that have trained recipients to smell the insincerity at 10 paces and bin it without reading, because nothing reads more obviously machine-written than a paragraph that opens with their first name and pretends to know what their company does. The result is that human-written outreach gets caught in the same downward trust spiral as the bot-written stuff, because by the time the recipient is reading the subject line, they’ve already decided the whole category isn’t worth their time. We all get tarred with the same brush.

So what’s the solution? Should we all stop offering something in return for a customer’s email address? That’s looking at the problem objectively backwards. Offering value in return for someone’s email address (and their consent to receiving information) isn’t the issue. What we then do with that address is where the wheels fall off.

The problem is that email lists compel us to categorize and segregate customers and prospects into nameless, faceless entities, while what we should be doing is using technologies that recognize there’s a person, an individual, at the end of every one of those email addresses. AI, instead of helping, has made the problem worse by giving marketers a scalable way to fake the appearance of individual attention without doing any of the work that real attention would require. But putting someone’s first name in the subject line of an email is not a relationship, any more than an automated “I noticed you recently visited our pricing page” is. Buyers are smart, and can smell the fakery because they’ve been told the same thing by 40 other senders this week, in subtly varied wording produced by what is almost certainly the same underlying model, and the cumulative effect is a kind of trust collapse that’s much harder to recover from than the old batch-and-blast routine was. Back when mass marketing was obviously mass marketing, sellers and buyers both knew where they stood. But when mass marketing dresses itself up in the costume of personalized and individual attention, the recipient who eventually clocks the trick treats every subsequent attempt at outreach as suspect, including the ones that may have been sincere.

The work of treating the recipient as a specific person with a specific situation is harder than it sounds, partly because it doesn’t scale the way the rest of the marketing stack now tries to do, and partly because it requires knowing things about people that probably can’t be added to a CRM field. It also means sending fewer emails rather than more, which is the one thing marketing dashboards are explicitly not set up to reward.

Sure, you need to create and show the value of your offering. But if you really want to attract and maintain a customer’s attention, then they need to feel that you value them too. This is probably truer now more than ever, because the cost of getting it wrong has gone up while the technological excuses for getting it wrong have multiplied. The premise of email marketing, based on the excuse that having someone’s address gave you the right to their attention, was always wearing a bit thin. Recipients have wised up to the trick and the filters have caught up to the volume, which has left only one thing that actually works, and it’s what we all originally argued for. Treating the person at the other end as a person worth not wasting.