There is one marketing channel that many SMEs consider to be outdated, yet which continues to offer the best return on investment: email. In 2026, the AI-powered email marketing It is no longer simply a matter of «sending out a newsletter on Thursdays»; it has evolved into a system that automatically segments recipients, drafts subject lines, determines the best time to send the newsletter to each individual, and predicts who is likely to make a purchase and who is about to unsubscribe.
And the figures back this up. According to Omnisend, which analysed more than 27,000 million emails sent via its platform in 2025, the messages automated They accounted for just 2% of emails sent, but generated around 30% of email revenue. To put it another way: it is the small, automated part of email that generates the most revenue. That is where artificial intelligence is making a difference.
Why email is still winning (even though it may seem otherwise)
Every few months, someone declares that email is dead. And yet, year after year, it remains one of the most profitable channels in existence. According to the report by Litmus, the average return on email marketing stands at around $36 for every dollar invested. This is an average across sectors — it should be taken as a rough guide rather than a guarantee — but it gives an idea of why brands are so keen on it.
The underlying reason is simple: the email is yours. You don’t have to rely on a social media algorithm or an advertising auction to reach your audience. It’s a channel built on first-party data, precisely the asset that has gained the most value in the age of privacy. And when you add AI to the mix, that channel stops being just a newsletter and becomes a revenue generator that works whilst you sleep.
What exactly changes when you incorporate AI into email?
Artificial intelligence doesn’t «write the email for you» at the touch of a button. What it does is relieve you of repetitive decisions and those that require analysing large amounts of data at once. In practice, this is what it offers today:
- Behavioural segmentation. Instead of a single list, the AI groups your contacts based on what they open, what they buy and when they stop engaging. Each group receives a different message.
- Topics and texts. The built-in assistants on the main platforms suggest various versions of the subject line and body text, and learn which ones work best with your specific audience.
- Optimal dispatch time per person. You don’t send emails to everyone at 10.00; the AI calculates the time at which each subscriber usually opens their emails.
- Forecast of staff absences. Identify who is losing interest before they unsubscribe, so you can launch a reactivation campaign in good time.
None of these tasks are new. What is new is that you no longer need a team of analysts to carry them out: they are included in tools that an SME can afford.
The figure to keep in mind: 2% in shipments, 30% in revenue
Let’s go back to the Omnisend figure, because it sums up the whole strategy. According to that same analysis, automated emails — those that are triggered automatically in response to a customer action — delivered around 16 times more value per email than manually scheduled emails. It’s not magic: it’s simply that they arrive at the exact moment when the recipient’s interest is at its peak.
These are the three automated workflows that almost any business should have up and running before considering anything more sophisticated:
- Welcome. The first email after signing up or making a first purchase. That’s when you have the most attention.
- Abandoned shopping trolley. A reminder for anyone who added items to their basket but didn’t complete their purchase. In e-commerce, this is usually the most profitable customer journey.
- After-sales. Follow-up, recommendations and requests for reviews. This is where repeat business is built, which is where the real profit lies.
AI is integrated into each one, fine-tuning the content, timing and target audience. You set them up once; they keep working for months.
Watch out for openings: they no longer mean what you thought they did
Here’s the awkward bit. If you keep measuring the success of your campaigns by the opening rate, you’re looking at a broken indicator. Ever since Apple introduced its Email Privacy Protection, the system preloads the images in emails — including the tracking pixel — even if the user does not actually open the message. As a result, the open rates appear inflated.
And this is no minor detail: a very high proportion of email opens now come from Apple Mail, according to various email delivery platforms. With so many «false positives», comparing your open rate against last year’s figures or against a supposed industry benchmark becomes unreliable.
What should you look at instead? Metrics that reflect genuine intent: clicks, conversions and revenue per email sent. It’s the same shift that other areas of marketing are already experiencing, where the The last click is no longer king when it comes to measurement. If a tool promises results based solely on page views, be wary.
How to get started with your SME without getting overwhelmed
You don’t need a list of 100,000 contacts or a master’s degree in automation. A realistic roadmap would look like this:
- Organise your own database. Clear consent, minimal fields (what they bought, when, what they’re interested in) and GDPR compliance from day one.
- Enable two or three workflows. Welcome, abandoned basket and post-purchase. That’s all for now.
- Let AI personalise it. Use the built-in features of your platform to segment and refine your topics; you don’t need any fancy tools.
- Measure by revenue, not by store openings. Set yourself a clear target: what percentage of your turnover do you want to come from email?.
- Keep the human eye in mind. The AI makes suggestions; a person then checks the tone and branding, and ensures no nonsense slips through before it’s sent.
That last point is not just for show. The very logic of AI-powered hyper-personalisation Even magic can come across as impersonal or go wrong if no one is keeping an eye on it. The combination that works is automation plus judgement – not just one or the other.
Frequently asked questions about AI-powered email marketing
Will email marketing still work in 2026?
Yes. It remains one of the channels with the highest return on investment, according to data from Litmus, and AI-powered automated workflows account for a disproportionate share of revenue, according to Omnisend. What has changed is how it is measured and personalised, not its relevance.
Do I need a huge list for the AI to be useful?
No. AI adds value even with modest lists, because it segments and chooses the best time to send messages better than any manual rule. The quality and accuracy of the data are more important than its size.
Does AI write emails for me?
It drafts them and suggests different approaches, but the branding criteria and final review must be carried out by a human. It works best as a co-pilot rather than on autopilot.
And what about WhatsApp and other channels?
They coexist. Email is the channel you own and can use in the long term; messaging, as we saw when discussing AI-powered conversational marketing, deals with the immediate situation. A sensible strategy does not choose between them: it coordinates them.
If you take just one idea away from all this, let it be 2% and 30%. Most of the value of email lies not in the campaign you send to your entire list, but in the automated messages that reach the right person at the right time. Artificial intelligence hasn’t come along to reinvent email; it’s here to ensure you finally make the most of it. At Vandelay, we help SMEs in Barcelona set up exactly that: workflows that run themselves, measured by what really matters – revenue.
Sources
- Omnisend, Email Marketing Report (figures for 2025): omnisend.com
- Litmus, Email Marketing ROI: litmus.com
- Constant Contact, Apple Mail Privacy Protection: constantcontact.com
English (UK) 
Español