The received wisdom that email is a settled, legacy channel is not just wrong; it’s dangerous. The inbox is not merely a destination for newsletters and promotional blasts. It is becoming an operating system, orchestrated by increasingly sophisticated AI agents working on behalf of the user. For UK marketers who have spent a decade optimising for open rates and click-throughs, the ground is not just shifting. It has been entirely replaced.
By 2026, the concept of a singular "email marketing" discipline will be obsolete. It is being fractured and rebuilt by agentic artificial intelligence. These AI agents, embedded within clients like Gmail and Outlook or operating as third-party services, will manage, filter, and even act upon commercial messages with near-total autonomy. The brands that thrive will be those that learn to market to the machine as much as to the human behind it. This is not a distant sci-fi scenario; the foundational elements are already deployed and scaling.
The Agentic Layer: Goodbye, Campaigns; Hello, Conversations
The traditional email campaign is a monolithic, top-down broadcast. A marketer decides on a message, a segment, and a time, then hits send. The metrics of success are crude: opens, clicks, unsubscribes. This model is already failing. Data from the UK Direct Marketing Association (DMA) shows that while email still delivers a formidable ROI—an estimated £36 for every £1 spent—engagement rates are in steady decline. The average UK open rate hovers around 21%, meaning four out of five messages are, for all intents and purposes, digital landfill.
Agentic AI introduces a new intermediary layer between sender and receiver. Think of it as a personal data valet for the user. Google’s "Gmail Tabs" was the primitive ancestor; today’s AI agents are infinitely more powerful. They do not just categorise; they analyse intent, synthesise information, and execute tasks.
For years, marketers have pursued hyper-personalisation through granular segmentation. We built complex workflows based on user behaviour, purchase history, and demographics. Agentic AI renders this approach quaint. An AI agent doesn’t care about the "Weekend Shopper" segment you’ve assigned its user. It cares about fulfilling the user’s underlying intent.
For example, a user planning a weekend break to the Peak District doesn’t need ten separate emails from train lines, hotel chains, and restaurant booking services. Their agentic AI will ingest these disparate offers, synthesise them, and present a single, consolidated brief: "Here are three travel and accommodation packages for your Peak District trip, ranked by price and travel time. The best-rated restaurant near your preferred hotel has a table at 7 pm. Shall I book it?"
The marketing objective, therefore, shifts from winning the click to providing the most valuable, machine-readable data for the user’s agent to synthesise. The new battleground is for the AI’s recommendation. Brands like Trainline or Premier Inn will need to structure their email (or, more accurately, their data feeds) to be optimally parsed and favoured by these agents.
The Contrarian Take: Your Click-Through Rate is a Vanity Metric
For over a decade, marketers have been conditioned to worship the Click-Through Rate (CTR). A high CTR was a sign of a compelling offer and an engaged audience. In the agentic era, this is a dangerously misleading metric. An email that requires a user to click through, land on a page, and manually complete a task is a failure of design.
Success in 2026 will be measured by zero-click conversions. This is where the AI agent acts on the user