The state of AI in marketing: Q4 2026

Every day, the MarTech team receives reports and surveys that examine AI use in marketing and beyond.

As an example, we’re about to enter the first — or second, depending on which report you read — AI-powered holiday shopping season. And once again, AI agents are going to do some or all of the shopping, depending on whose analysis you’re reading.

As for marketers, the reports agree almost all marketers are using AI. They also agree the most common AI use case in marketing is content creation, although some marketers are building apps, powering personalization, and enhancing workflows.

The research also differs on the ROI marketers are seeing in their AI investment and how AI is impacting their workforce and hiring plans.

I do look at most of the research sent my way, I swear I do. But my sense of where AI stands here in Q4 2026 is based more on the conversations I’ve had with marketers and leaders in the martech space, on- and off-the-record. (You can catch many of the on-the-record conversations on our Conversations with MarTech podcast.)

Here’s my sense of where AI stands in marketing in Q4 2026.

The reports generally agree more than 90% of marketers say they’re using AI. I don’t doubt that for a second. Almost everyone now has an LLM of choice, or even two. Adoption of basic AI tools to help with basic tasks isn’t a problem and hasn’t been for quite some time.

But getting marketers out of their LLMs and getting AI into their workflows is the chasm too many marketers can’t seem to cross. It’s not always their fault. You can run your LLM and prompt away, but connecting to other apps is technically challenging for some and requires jumping through hoops.

What does your IT team allow? What resources are available if you need technical assistance? Are your teammates fluent enough in AI to use what you’ve built?

This is why so many people are (literally) cutting and pasting from LLMs. Cutting and pasting is easy. Integrations are hard. But cutting and pasting is not a workflow.

AI in every app means smarter, more powerful silos

Nearly every tool your marketing team uses has AI infused in some way, from elegant assistants and integrations to plain-old bolted-on AI that looks awkward at best.

It’s not just the martech apps. If you’re using Excel, Google Sheets, PowerPoint, or many other productivity tools, AI is there with you, ready to help. That is, help with that app. Not in the next app you’re going to use.

Embedding AI into every application doesn’t automatically make the applications work together.

Instead of using AI to connect the CRM and the marketing automation platform (MAP), and the productivity tools (see workflows above), we’re creating AI-powered silos for our teams.

This is what would scare me if I were a vendor. It costs real money to embed AI in everything, but the marketers can only use it in so many tools. If you’re a vendor paying the bill, you’d better hope your AI is the one being used. Marketers have plenty to choose from.

Are agents part of the answer? Sure. But almost every vendor now offers those too.

You don’t trust the data, so you don’t trust the output

Many marketers lack confidence in the quality of their data. They know AI will just amplify that bad data if it gets the chance. Poor data leads to poor outcomes. It delivers the wrong information to your customers and teammates.

Salesforce gets a lot of flak for its ecosystem, which has Data 360 as a data layer sitting beneath Agentforce and the CRM. But you know what? Salesforce is right.

Companies exploring Salesforce might feel locked into expensive contracts, or like they’re being sold more than they asked for, but data management and governance are hugely important. Costs and scope creep aside, Salesforce knows bad data can sink AI initiatives, and since they’ve bet the company on AI, they can’t let that happen.

You can expand your team’s use of AI with the data you have. But garbage in, garbage out.

AI is feeding marketers’ fixation on more

More content, more designs, more drafts, more emails… This AI thing is amazing. One person can fill up all of the channels with messages. Isn’t it great? Well, put on your customer hat for a second.

Several marketing leaders I’ve spoken to think marketers are overfocused on doing more. AI can easily create more, but it doesn’t think about resonance and the power of what’s being created.

Creating more marketing creates another problem as well.

The time fixing AI problems is not time well spent

On the content side in particular, many marketers report that checking AI’s prolific output is creating more work and failing to deliver the efficiency that was promised.

There are tools and guardrails marketers can use to help maintain their brand voice and their style in their AI-created assets. But as we discussed in the workflow section, a lot of marketers are still at the LLM stage, where they’re likely relying on prompts to guide asset creation. Prompts will help with ideation and an initial draft, but there’s more work to be done.

AI-generated work isn’t necessarily efficient work.

AI needs to give your customers what they want

Creating more is just one example where marketers see the potential of AI to help with their work, sometimes at the expense of the people on the receiving end. They don’t necessarily want more content. They don’t always want AI-assisted chat and help functions. They don’t always find those experiences helpful.

There are times when people want a quick answer, and AI is great. There are times when they need empathy, human touches, and understanding.

The question marketing teams need to ask is: Are you creating more because that’s what the customers want? Or are you creating more because AI makes it easy?

If you think proving the ROI of a brand campaign is hard, how are you demonstrating the ROI of your AI use in marketing to your leadership? Fewer marketers are confident they can prove ROI in 2026 than in 2025.

Whether it’s brand campaigns, event sponsorships, advertising, or something else, marketers are constantly being asked to prove their investments are delivering. Marketing attribution is messy, especially in B2B marketing.

As with advertising, if you want to hold on to the budget, you need to prove ROI. Now you have to prove it for AI too.

A significant problem facing marketing teams trying to measure the ROI of their AI investments is looking beyond the outputs, like content pieces created. They need to sufficiently account for rework, integration costs, customer outcomes, and business impact.

Let’s say your team does build something on top of an LLM and incorporates it into your personal or team workflows. It works great. Until it doesn’t.

AI models are changing rapidly. These model changes can introduce unpredictability into workflows. That leaves marketers trying to figure out why a tool that worked fine one day doesn’t work the next.

The amount of chaos this causes depends on how deeply you embedded that AI in the first place.

These eight observations tell us AI adoption in marketing is widespread, but meaningful AI transformation remains elusive for many organizations.

Marketing teams are embracing AI tools, but they haven’t sufficiently transformed their workflows, systems, and organizations to realize the technology’s potential.

This isn’t a marketing-specific problem. It’s a strategic problem many organizations face because AI adoption has outpaced organizational strategy.

The easy availability of LLMs helped with rapid adoption, but it introduced everyone to AI on an individual level. Integrating the AI platforms, the existing stack, and the data must be addressed at the organizational level, not by the marketing team alone.

Technology and business leaders need to collaborate to develop a strategy to move to the next level.

As for the AI adoption issues specific to marketing, they are familiar to experienced marketing leaders.

They include discussions of quality vs. quantity, workflow efficiency, customer experience, and business value. Marketers have wrestled with these issues for years. They shouldn’t expect AI to solve them.

MarTech is owned by Semrush. We remain committed to providing high-quality coverage of marketing topics. Unless otherwise noted, this page’s content was written by either an employee or a paid contractor of Semrush Inc.

Mike Pastore is the Head of Content & Media at Third Door Media, the publisher of the Martech and Search Engine Land websites and the producer of the SMX and MarTech Conferences. In nearly three decades in B2B marketing, Mike has worked as an editor, writer, and marketer. He first wrote about marketing in 1998 for internet.com (later Jupitermedia). He then worked with marketers at some of the best-known brands in B2B tech, creating content for marketing campaigns at both Jupitermedia and QuinStreet. Prior to joining Third Door Media as the Editorial Director of the MarTech website, he led demand generation at B2B media company TechnologyAdvice.

Google's "preferred sources" feature allows users to customize their search results by selecting news outlets they want to see more often in the "Top Stories" section.