2026 Trends: The agentic boom is underway

“I'd like to thank my agent”

The agentic revolution is upon us.

For anyone with imposter syndrome, LinkedIn has been an unforgiving place of late. Self-proclaimed ‘growth hackers’ boast of ruthlessly orchestrated content or sales workflows that are both efficient and effective (“just comment ‘agent’ if you want to know how I do it”).

However, in reality, most companies haven’t got stuck in... yet.

According to Chiefmartec and MartechTribe’s November 2025 survey (see the Future of Martech report), just under a quarter of respondents have deployed agents into production. Another third were in the limited production phase, and another third in the experimentation phase.

Content and customer service as proving grounds

When it comes to use cases, content production (69%) and customer service (54%) were overwhelmingly the most popular. But many of the other use cases in the survey had been adopted by over a fifth of respondents.

Most of us will have at least experimented with agents for summarising or optimising copy, and others will have gone further with workflows that help with content adaptation and posting. The same applies to sales development (AI SDRs), where B2B marketers are automating ICP identification and outreach. Agentforce, for example, has been in the market since late 2024 and its lead nurturing agent can send customised outreach emails using data on record, follow up with nudges, answer questions from prospects (including off-topic) and hand them off to sales reps.

What’s exciting about agents is not just their capacity to take away admin (after all, automation has been around for a long while) but their contextual understanding of a business.

Knowledge management

With the right grounding, agents can superpower knowledge management, including:

• the summarisation of CRM history around customer enquiries;
• the creation of response libraries, and a feedback loop that improves communication and recommendations;
• and potentially the summarisation of everything going on in the organisation, from meetings to processes.

The agentic caveat is data, a familiar refrain for marketers. In the aforementioned survey, 56% of respondents said poor data quality was a significant implementation challenge. Ensuring the right agents have access to the right data is where some of those legacy issues rear their heads.

No replacement for SaaS platforms

Though agentic tools are attractive because they often live adjacent to the stack (and marketers can set them up quickly), it’s clear that they don’t currently replace SaaS platform functionality.

Scott Brinker and Fran Riemersma, authors of the Future of Martech report, highlight that a large majority of agentic use cases (85%) enhance existing martech functionality. Two out of five use cases offer new functionality altogether, and only 30% said they had replaced existing martech features (NB. the figures overlap because some agents were found in multiple categories – adding and enhancing, or enhancing and replacing, for example).

These are probably comforting statistics for those who are yet to explore agentic. They allow the marketer to think laterally, and they don’t encourage root and branch replatforming in martech (the sort that gets expensive and time consuming).

Agents of the customer

Before you get too comfortable though, notice we haven’t touched on customer agents. These new intermediaries may upset the apple cart for all of us soon enough if they allow consumers to delegate research and purchasing to a trusty assistant.

Notably, the past few months have seen Google launch its Agent Payments Protocol and Universal Commerce Protocol. These are designed to create a common language between agents and systems in a world where shoppers need to be shown to have purchase intent when they tell an agent to go off and buy something.

There are major doubts, though, about whether shoppers will be happy making considered purchases outside of a fine-tuned ecommerce UI. As such, we haven’t plumped for ‘chat to checkout’ behaviour as a 2026 trend for marketers, but we note that AI companies are already testing ads – ChatGPT is doing so in the US, and Google has introduced Direct Offers in AI Mode.

Time will tell whether we can delegate queuing for gig tickets, researching airline ticket prices, or finding the best price for consumer goods to our agent intermediaries.

This is an excerpt from CTI Digital's report, 2026 Trends: Marketing in the Age of AI.