It is September 2026, 2027 is round the corner (*GULP*) and the last eighteen months have moved faster than any period I can remember in this industry. AI everywhere. Every tool rebuilt, rebranded or re-pitched with some layer of it bolted on. Platforms going headless (wtf even is headless, just got my imagining people with no heads). Enforcement tightening. Whole job descriptions being rewritten.

For the record, I am not a fan of replacing everything with AI, in that respect, I am anti AI. 

And email marketing, the actual fundamentals of it, has not moved an inch.

Which sounds like a criticism of the industry and is not. It is the single most useful thing I can tell you, because the reason nothing fundamental has changed is that email's fundamentals were never technological in the first place. They are human. Permission, attention, memory, habit, trust. And humans do not ship a new version every eighteen months.

So this is my read on the year, what I think 2027 brings, and the thing I most want you to take into it. Fair warning that it is long and opinionated, and that I am going to say some things about AI that will annoy a few people.

The belief system 

Why email keeps getting left behind

Before any of the 2026 stuff, we need to talk about the thing sitting underneath all of it, because you cannot understand why email gets neglected without it.

Email has always been the back-burner channel. It is rare, properly rare, to walk into a business that is AHEAD with email. Some organisations have invested properly, with real tooling, a real function and skilled people running it, and some of the biggest brands in the world are among the worst offenders on this. Enormous budgets, beautiful brand work, and an email programme that has been running on the same assumptions since about 2017.

The reason is not laziness, and it is not budget. It is that we have been taught the wrong thing about what email is.

The two beliefs, and they are the same belief

• Belief one: email is a revenue-generating channel. So it gets judged on revenue per send, it gets pushed to produce, and when it stops producing on demand it gets pushed harder. Which generates the volume, the complaints and the reputation damage that make it produce less.

• Belief two: email does not work. Usually arrived at after belief one has been applied for two years. So it gets deprioritised, under-resourced and left alone, which guarantees it will not work, which confirms the belief.

Two opposite-looking positions arriving at the same destination, and both of them come from the same original error, which is mis-categorising what the channel is for.

You cannot fix a channel you have mis-categorised. You will keep applying the wrong interventions, measuring the wrong things, and drawing confident conclusions from numbers that were never answering your question.

The measurement mismatch, which is the real mechanism

Now the bit I would most like you to sit with, because I think it explains more about the state of email than anything else in this post.

Email compounds!! OVER time. My email list didn't make me money for ages. Now it brings in decent revenue for me every month.

Every send builds or erodes a little bit of recognition, a little bit of trust, a little bit of association with a subject. The value accumulates across months and years, and the return on any individual email is mostly not collected by that email. It is collected later, by a different email, or by a sales conversation, or by somebody typing your name into Google eighteen months after they first heard of you.

And we measure it per campaign. Per month. Per quarter at best.

So the measurement window is shorter than the mechanism. Which means your reporting is structurally incapable of showing the value of the thing it is reporting on, no matter how good your attribution is.

Follow that through and the whole loop falls out of it:

1. Email compounds over months and years.

2. It gets measured in weeks.

3. The reporting shows a small, unimpressive number.

4. Email gets deprioritised, because the number is unimpressive.

5. Less investment, less attention, worse execution.

6. It performs worse, which produces an even smaller number.

7. Which confirms that email does not work.

Nobody in that chain did anything stupid. Every step is a reasonable response to the step before it. The error was made right at the top, when somebody decided what window to measure in, and everything after that is arithmetic.

Free Email Marketing Health Check

Audit your entire ecosystem in under 30 minutes

Answer a set of straightforward questions across every area of email marketing and get a personalised score, traffic-light priorities, and clear actions and improvements you can make today.

Take the Health Check →

So what is email for?

Same job as every other marketing channel, and we seem to have forgotten it. Build demand, and build trust!! 

That is it.

Email marketing exists to build demand, build trust, and increase the likelihood of a conversion at some point across somebody's journey with you. Not this week or in a week, at some point, when they need you, your product or your service - which also could be never and that is totally okay! 

Can email be attributed back to a sale, a lead, a conversion? Absolutely, and it happens with my emails all the time. But that is the visible tip of it. The rest is the compounding, and the compounding is where the money is.

Revenue is a lagging indicator of trust. If you treat the lagging indicator as the objective, you will spend your entire year optimising the thing that reports rather than the thing that works.

2026

The year people woke up, and the year it got harder

The wake-up (wakey wakey)

It has been my busiest year by a distance. Deliverability problems, email strategy work, workshops going out of the door faster than I can build them. And the conversations split neatly into two types.

• “I think we are missing a trick with email.” Businesses realising the channel they have neglected for five years might be doing more than it is, and wanting to know what good looks like.

• “We used to get leads from email and now we get nothing.” Or the other version, which is we changed email provider and everything has gone down the pan. Both are usually the same underlying problem wearing different clothes.

Both groups arriving at the same moment is not a coincidence. The businesses that were quietly getting away with poor practice have stopped getting away with it, and the ones who never took the channel seriously have noticed the people who did are pulling ahead.

Deliverability got materially harder (like CRAZY HARDER) 

The tolerance disappeared. Enforcement across the major providers converged and hardened, Google retired the reputation dashboards that gave people the illusion of visibility, and the practices that used to be absorbed quietly now produce rejections.

Across all providers you should expect somewhere around fifteen percent of a typical send to land in spam, and that is NORMAL rather than a crisis. Every business lands in spam to some degree. What matters is whether you are consistently there and whether it is persistent.

My expectation is that the number moves to somewhere between seventeen and twenty one percent next year, and I will come back to why.

AI arrived everywhere, including places nobody needed it

You want my personal stance, so you are getting it. 

I am not an AI advocate. I use it, and I use it to my advantage, and I am extremely intentional about where. I am also very mindful that I still have to use my own brain, because the moment I stop doing that, I stop being any good at this and get brain rot. 

Do I want it to replace everything I do? Do I want to sit at the top of an agency and an email consultancy that are both being run by AI? BIG FAT NO THANK YOU. Managing AI does not bring me joy; it is not what I set out to do, and personal connection is the thing I care about in this work.

Here is what I hate, since we are being straight with each other...

Everybody is being pushed into using AI in ways they do not understand. Every tool, product and service has had some layer of it bolted on so the marketing can say it is AI-powered. Klaviyo has gone headless. You can now sit inside a platform and let it do the lot for you.

What gets removed in that transaction is critical thinking and strategic thinking, and those are the only two things in this job that were ever worth paying for.

The trust deficit (the real story of the year)

What happened in 2026 was this: Content became free to produce.

Anybody can generate a competent blog post, a competent email, a competent sequence, in minutes. Which means competence stopped being a differentiator, because a differentiator has to be scarce and competence no longer is.

So the volume went up enormously, and the scarce thing moved. It used to be information it is now trust! 

We are in a trust deficit as a marketing economy. Information overload, indistinguishable content, and a reader who has learned that most of what arrives was produced without much thought. Which makes the brands and people who have built real trust disproportionately valuable, and it makes building it harder and slower than it was.

Cold email died in this, by the way. You cannot build demand from cold email any more and you cannot generate trust from it either, because there is no prior relationship for either to attach to. The leads dried up, and the tactics did not cause it, the trust deficit did.

And AI started reading your email before the human does

The change I think is most underrated. There is now an AI layer sitting between your email and your reader in the largest inbox in the world, summarising, prioritising and deciding what deserves attention.

Which means landing in the inbox stopped being the finish line. You can pass every authentication check, clear every threshold, land in Primary, and still be summarised into three lines that somebody else wrote, ranked below four other senders, or quietly demoted for being low value.

You are now writing for two readers, and only one of them has feelings.

What did not change

The fundamentals are human, and humans do not get a software update

Every year the industry chases the new thing, and every year the answer turns out to be the same boring set of things nobody wants to hear. There is a reason for that, and it is not that email people are unimaginative.

It is that everything that decides whether email works is a fact about people, and people are running the same operating system they were running in 2015, 2005 and 1995.

• Permission still determines everything downstream. How somebody got onto your list is still the strongest predictor of how your email performs, and no tool has ever changed that.

• Attention is still allocated the same way. The inbox is a task environment, not a browsing one. People triage rather than evaluate, and the question is whether something needs them, not whether it is any good.

• Habituation still works exactly as it did. Repetition builds recognition and erodes attention at the same time, from the same input, which is why sending more eventually stops working no matter what generated the copy.

• Associations still form through pairing. Every email quietly teaches somebody what your name means, and after a year your brand carries whatever feeling it has been arriving with.

• Trust still builds slowly and breaks quickly. Negative events carry more weight than equivalent positive ones, which is as true of a broken promise in an email as it is of anything else.

• Recognition is still processed before content. Your sender name still does more work than your subject line, and a name somebody has a relationship with does the most work of all.

Not one of those is a technology problem, so not one of them has a technology solution. Which is why the businesses chasing tooling keep arriving back at the same place, slightly poorer.

The inbox is still where work happens

In B2B, the inbox is where you execute. It does not matter how much Slack or Teams a business runs on, you will still have an email inbox, because that is where stakeholders get managed, tasks get tracked, diaries get arranged, decisions get recorded and things get replied to. People go in there to DO something, which completely changes how your marketing email is received.

On the consumer side, it is habit and retrieval, plus a low-level worry about missing something important. Not missing a sale. Missing something that matters.

Either way you are not competing with the other brands in the inbox for attention. You are competing with that person's to-do list, and the to-do list is winning.

2027

What I think happens next

Predictions are opinions with a date attached, so take these as mine and hold me to them.

The correction

Email marketing is far simpler than the industry wants you to believe

There is an enormous commercial incentive to make this look complicated. Complexity sells software, sells agencies, sells certifications and sells consulting days. And most of the time, the thing that works is embarrassingly plain.

A client story, because it makes the point better than I can

I work with a B2B client, and have done for a long time. They came to me as a HubSpot client originally, and email came with it, which is the two sides of what I do.

The marketer there did not have much time. Not a big team, not a big budget, no capacity to build an elaborate lifecycle programme. So we agreed on one thing, and one thing only. One newsletter, every single week, done properly.

That single weekly newsletter now generates hundreds of thousands of pounds worth of leads a month.
No intricate journeys. No predictive scoring. No twelve-step nurture with branching logic nobody can maintain. One email, every week, to the right people, with something worth reading in it, from a business that had done its foundations properly.

What back to basics means

1. Foundations first. Authentication, domain structure, stream separation, data quality. None of it is interesting and all of it is load-bearing.

2. Deliverability as an ongoing discipline. Not a project you completed in 2023. A condition shaped by your behaviour over time, monitored in a way somebody in your business understands.

3. Know who is on your list and how they got there. Because permission determines everything downstream and most businesses properly cannot tell you.

4. Define engagement for your own business. Rather than inheriting ninety days from a blog post that was describing somebody else entirely.

5. Send consistently, to people who want it. Consistency is a reputation signal as well as a relationship one, and too little is as much of a problem as too much.

6. Give more than you ask. Measured across the programme rather than assumed, because almost nobody who thinks their ratio is fine has checked it.

7. Learn the psychology. How attention, habituation, association and trust work, because that is the layer underneath every tactic and it does not go out of date.

8. Then, and only then, add the clever bits. Segmentation, derived data, prediction, automation. All useful, all built on top of the above, and all worthless without it.
   
   
   

Going into 2027

What I would do between now and January

• Find out where you are landing. Placement per provider, measured as a trend rather than a single reading. Most businesses have never looked, and you cannot plan around a number you have never seen.

• Change what you report on. Engaged list size, revenue or pipeline per thousand sent, complaint rate as a headline. If your reporting only shows a short window, it will keep telling your business that email does not work.

• Have the belief conversation with whoever holds the budget. Email builds demand and trust and increases the likelihood of conversion over time. Getting that agreed is worth more than any campaign you will run next year.

• Audit where AI has quietly replaced thinking. Not where it is saving time, which is fine. Where it is making decisions nobody in your team could explain or defend.

• Pick one thing and do it properly. One flow, one newsletter, one segment. Consistency beats ambition, and the compounding only starts once something is running.

• Invest in understanding rather than tooling. The gap between people who understand this and people who press buttons is going to be the most valuable gap in marketing next year.
  
  

The conclusion

Everything around email changed this year, and email did not. That is not the industry failing to innovate, it is the subject being what it has always been: a set of facts about how people behave when something lands in a space they consider personal.

AI has made producing email trivial and made doing it well harder, because the volume went up, the competence floor went up, and the only things left to compete on are the ones a model cannot generate for you. Judgement, taste, restraint and trust.

Which is why I keep coming back to the same unglamorous advice. Respect the inbox. Know who is on your list. Send consistently, to people who want it. Give more than you ask. Understand the person on the other end well enough to write as them. And measure it over a window long enough to see what it is doing.

None of that is new. None of it will be new in 2027 either. That is the whole point, and it is the most valuable thing I know.

Foundations first, and the foundations start with deliverability

If a fifth of your email is not arriving, nothing else on your 2027 plan matters very much. My free 60 minute Email Deliverability Training covers the absolutes, the metrics worth tracking, what normal looks like and how the audit process works.

 

The Email Deliverability Certification Programme goes further and gets you certified, using the same framework and templates I use with clients. Self-paced, CPD certified, and your employer should be paying for it.

Join the free deliverability training →