Email marketing isn't dying. But it is being rewritten (by machines).
Glasp’s Note: At Glasp, we value the close personal relationships we’ve built with thinkers, creators, and doers who fuel curiosity and growth. This week, we spotlight Des Brown—a dynamic digital marketing force and founder of Email Expert Africa. He’s also the co-founder of CrossLetter, a platform for cross-promoting, finding email sponsorships and discovering newsletters. You may remember Des from his Glasp Talk appearance, where he shared deep insights into digital strategy and email excellence. We’re honored to keep learning from him and excited to share his thoughtful reflections with our community.
There’s a question I keep hearing from marketers, newsletter creators, and founders alike:
“Is email marketing still worth it when AI is doing… practically everything?”
Look, it’s a fair question, and I think many folks are pondering this themselves.
The short answer is yes. Emphatically! But the longer answer is a little more interesting, because the way email works is changing faster than most senders realise (or are able to get comfortable with).
And this has nothing to do with email being broken. It’s built around the premise that the inbox itself is getting quite a bit smarter.
Allow me to explain (and grab a cup of coffee because I’m taking a deep dive into this one).
Your inbox has a new gatekeeper
There’s an uncomfortable future ahead for email senders, and there are elements of this already happening. And it begins with the fact that your subscribers may wind up never actually reading your email. Well, not the way you wrote it, anyway.
Inboxes are already generating AI-powered summaries that appear before an email is opened. Apple, for example, is replacing your carefully crafted preheader text with a machine-written preview. Gmail’s Gemini integration does something similar post-open, summarising your content into a few bullet points before a reader scrolls to your words.
Let that sink in for a minute. In the short-attention-span economy, this is going to mean folks aren’t even skimming anymore. They’re getting things told to them by AI, and choosing what to dive into from there.
Yes, your subject line still matters. And yes, your content still matters (even more so). But there’s now an intermediary AI layer sitting between your send and your subscriber’s attention. And that AI is making editorial decisions about your email on your behalf.
The almost scary part is that this isn’t even a future scenario. It’s happening now. Gmail made its AI summary features free to all personal account users in the US earlier this year, and Apple Intelligence has been summarising emails since late 2024 on supported devices. If your audience skews toward Gmail or Apple Mail (and for most of us, close to 90% of inbox usage comes from these two clients), this directly affects how your emails are consumed.
So what do you do about it?
The answer is what’s going to separate email senders that send from those that perform, and it begins with a relatively simple concept: You write for two readers, both the human and the machine.
Writing for AI (without losing your soul)
This doesn’t mean stuffing your emails with keywords or writing like a robot. In fact, it almost means the opposite.
AI summaries extract meaning from structure and clarity, so if your email rambles, the summary will too…or worse, it’ll pull the wrong takeaway entirely. Now, if your email is clear, scannable, and has one obvious point (likely per section), the AI in the inbox is far more likely to surface the thing you want it to surface.
And if you want to see what that looks like in practice:
Lead with the value. You don’t want to be burying your most important point below an insanely long intro. If AI is summarising your first 100 words, you’ll want to make those words count. Treat the opening of every email like a headline and a subheadline had a proverbial baby: punchy, clear, and (hopefully) impossible to misinterpret.
One idea per section. This has always been good email practice (the Rule of One, as I call it), but it’s now almost structurally critical. AI models parse sections, so if each section has a distinct, clear point, the summary will likely reflect that. Mix three ideas into one paragraph, and the machines will pick whichever they fancy, meaning less control over what you want your readers to see.
Use descriptive headers. Not clever ones. Descriptive ones. “The Secret Sauce“ tells an AI very little unless it’s extremely contextually supported. “How to Improve Open Rates With Segmentation” tells it all it needs to summarise that section well. This is a trade-off, I’ll admit, and for creative writers, it’s weighing up personality vs. parsability. But the best email writers find a way to do both, and it can be done well.
Should you care that LLMs might cite your email content?
This one’s interesting, and it feeds on both the email and SEO worlds. You may have heard the term AEO (Answer Engine Optimisation). Essentially, it’s the practice of structuring content so that AI tools like ChatGPT, Perplexity, and Google’s AI Overviews can find it, understand it, and cite it when answering user questions.
Now, most of this conversation lives in the SEO world, true, but if you’re publishing your newsletter on the web (think Substack, Beehiiv, Ghost, or even your own site), your email content is web content. And that means AI search tools can, and likely will, pull from it.
So, the question begs: Should you optimise for this? And, like any good email advice, it depends.
And here’s what it depends on.
If you’re writing actionable, original content with clear structure and genuine expertise, you’re already doing 80% of the work. AI models favour content that’s well-sourced, clearly structured, and factually dense. Sound familiar? That’s just good writing practice and the stuff you should be doing for email anyway.
Where it gets tactical? Well, my advice is often that you’ll need to make sure your web-published newsletters have descriptive titles, clear meta descriptions, and aren’t locked behind a paywall that blocks AI crawlers (unless this is your strategy, of course, but that’s a discussion for another day).
Structure your posts with headers that mirror the questions your readers actually ask. If your newsletter answers “What’s the best email platform for creators?” in a detailed, opinionated way, there’s a real chance an AI search tool surfaces that answer.
This isn’t necessarily about gaming a system either. You want to create content and structure advice that’s genuinely useful (making that usefulness discoverable). Which, if we’re being honest, has always been the email marketer’s job.
AI-Generated content (The current elephant in every inbox)
I have a few strong opinions on this one, because we’re seeing a turning point in the value of content of late. AI should help you write emails. AI should not write your emails.
And that distinction matters more than you think, because your audience is, and will, notice.
Yes, the tools are extraordinary (especially in getting things done and fitting that into the limited time we all have). I use AI regularly. For brainstorming, researching, restructuring drafts, testing different angles on a paragraph, and mainly for understanding the intricacies of my audience behaviour.
It’s brilliant at accelerating the process and removing that blank-page paralysis we all sometimes face as writers and creatives. But if you hand over the entire creative process to a machine, your emails will sound like everyone else’s emails, and then the point of sending something falls away. In a channel where trust and voice are absolutely everything, sounding generic is almost worse than sounding imperfect.
But the balance is tricky, isn’t it? Your readers can now spot AI-generated copy far more easily than ever, precisely because they’re exposed to so much of it (cue the “delves”, “it’s this, it isn’t this, the poor “em-dashes”, you know the drill.
Phrases, cadence, the relentless positivity…it all starts blurring lines. And then, when everything sounds the same, nothing stands out. You see where this is going?
To give this an analogy, you want to use AI as a co-pilot (yep, there’s even AI called that, thanks Microsoft). You should ALWAYS remain the captain.
Let it generate options. Let it tighten your prose. Hell, let it even write something in different angles to help you define some creative approaches if you need to, but make sure the final product sounds like you. It’s your experiences, your observations, your voice that your subscribers signed up for, and it’s the one thing AI genuinely cannot replicate (at least, not until they plug it into our brains, which, hey, may still be coming).
Which email platforms enable you for AI well?
Because I work with many email senders looking to grow in the right direction, I get asked this almost constantly. Truthfully, my answer is always the same: the best platform is the one that fits your stage, your end-goal, and your budget.
That said, a few stand out for different reasons:
Beehiiv and Substack have become a go-to for newsletter creators.
Both have strong growth tools (referral programmes, recommendations, boosts, etc), clean editors, built-in monetisation features, and audience growth opportunities..
I’m intrigued to see where their AI-tools begin transforming things, as this is where the majority of newsletter-specific senders are finding the most traction right now.
Stripo is another surprising email platform with strong AI-led tools. Now, Stripo isn’t being used to send and manage emails, but their AI-assisted modular and HTML building tools are exceptional for time-saving and pushing email boundaries, and if you’re in the more visual email space, a solid consideration for email design generation.
Brevo and ActiveCampaign remain solid for businesses running broader email marketing campaigns (think e-commerce, B2B lead nurturing, and transactional emails alongside newsletter content).
They deserve a mention here because they’re also doing a heap to ensure email senders are enabled with strong AI-led tools around email metric analysis, campaign composition and strategy optimisation across various elements within their platforms.
The common thread here, though? Whatever you choose, you’ll want to make sure it supports the fundamentals before any fancy AI features: clean list management, solid deliverability, segmentation capabilities, and analytics you’ll actually use.
No platform (or its AI tools) will save a bad strategy, and a good strategy can work on almost any platform.
What’s actually working right now
Rather than simply theorising, here’s what I’m seeing work in practice among the email senders I’m working and collaborating with, and who are adapting well to the AI era that we’re moving into:
Hyper-relevant segmentation. AI tools inside platforms like Beehiiv are making it easier to segment audiences based on engagement patterns (and not just demographics, which is what we’ve all been doing for ages). Senders who use this are seeing stronger click-through rates and more engaged audiences because they’re matching content to behaviour, not merely on what they assume audiences want to engage with.
Cross-promotion and discovery networks. One of the most effective growth strategies I’ve seen for audience growth is newsletter cross-promotion. Essentially, finding complementary newsletters and sharing audiences is almost a staple for most growing newsletters, so finding solid cross-promotion tools and building on those is now becoming a strong recommendation.
Plain-text and hybrid formats. There’s a growing shift back toward simpler email design. We’re talking fewer images, more text, and even more conversational tone (likely, to stand out amongst the AI-sludge). This isn’t merely a style preference; simpler emails tend to render more consistently across AI-summarised views and are less likely to have their key message buried by a machine-generated preview that pulls from an image alt tag instead of a headline.
Consistent, audience-first content. This one never changes. It will never change (well, for good email senders anyway). The newsletters thriving in the AI era are the ones that always treated their audience as people first and metrics second. They write with empathy, specificity, and some genuine expertise, because no amount of AI can substitute for that.
What to do with all of this
Email marketing isn’t being replaced by AI. You can stop worrying about that.
Email is, however, being mediated by it, and this is out of your control.
Inboxes are moving from neutral containers to interpretative layers. AI decides what to surface, what to summarise, and increasingly so, what to actually prioritise, meaning email senders need to think about their structure, clarity, and discoverability in ways they haven’t (a.k.a, shifting your mindset from what it was five years ago).
The most important thing for any email sender right now is to realise that the fundamentals haven’t changed. Write something worth reading. Send it to people who want it. Make the value obvious. And, folks, be human. That’s why your audience is there.
That’s also why knowledge-sharing communities matter more than ever. When AI can generate a passable email in seconds, the voices that cut through are the ones rooted in real experience, contain genuine curiosity, and have a willingness to learn publicly.
Communities built around that kind of exchange (where practitioners share what’s actually working, not just what sounds good) become a beacon of light in a tumultuous digital ocean filled with tons of AI noise.
AI raises the bar on all of those things. And honestly? That’s not a bad thing. The senders who were already doing this well are the ones who’ll continue to thrive. The ones relying on sheer volume, generic (read “bland”) content, and spray-and-pray tactics are going to find the AI era increasingly unforgiving.
The inbox is telling you what it wants. My question, dear reader, is will you be listening, or will you pass that on to AI too?
Des Brown is an email marketing expert and the creator of Email Advice in Your Inbox, a newsletter for email professionals, marketers, and creators. He’s also the co-founder of CrossLetter, a platform for cross-promoting, finding email sponsorships and discovering newsletters. Connect with Des on LinkedIn.
📣 Community Updates by Glasp
🟥 Glasp Talk with Des Brown:
In this episode of Glasp Talk, Des shares how to achieve 50%+ open rates — from nailing deliverability (SPF, DKIM, DMARC) and crafting personalized emails, to re-engaging inactive subscribers and using AI for smarter segmentation. If you want to build an email list people actually read, this one’s for you.
🟩 Updated PDF upload limit:
You can now upload PDFs up to 500MB per file to Glasp — perfect for lengthy research papers, full ebooks, and dense reports. Start highlighting without worrying about file size! Check the tutorial below.
🟨 New Tutorial - Bulk Download YouTube Transcripts:
Did you know you can bulk download transcripts from your entire YouTube channel with Glasp? Connect your channel, select the videos you need, and export everything at once in .txt, .md, .csv, .json, or .srt — no copy-pasting required. Check the tutorial below.
Would you like to take Glasp on the go?
With the Glasp mobile app, you can highlight and organize your favorite content anytime, anywhere. Stay productive on the move and never miss an insightful quote.
Partner With Glasp
We currently offer newsletter sponsorships. If you have a product, event, or service you’d like to share with our community of learning enthusiasts, sponsor an edition of our newsletter to reach engaged readers.