Marketing teams have widely adopted artificial intelligence (AI), but many organisations are struggling to translate its use into meaningful business value because of fragmented workflows, data-quality concerns and difficulties measuring returns, according to a new analysis by MarTech.
In The state of AI in marketing: Q4 2026, published on October 12, Mike Pastore, MarTech’s head of content and media, argues that the widespread availability of AI tools has accelerated individual adoption without bringing about the organisational changes needed to realise their full potential.
“The reports generally agree more than 90% of marketers say they’re using AI,” Pastore writes, noting that content creation remains the most common application, although some marketers are also using the technology for personalisation, app development and workflow improvements.
However, he identifies a significant gap between using AI for individual tasks and integrating it into everyday marketing operations.
“But getting marketers out of their LLMs and getting AI into their workflows is the chasm too many marketers can’t seem to cross,” he writes, referring to large language models (LLMs).
The difficulty, according to Pastore, lies partly in connecting AI tools with other applications and overcoming technical constraints within organisations. While marketers can readily generate content using AI assistants, integrating those tools with customer relationship management systems, marketing automation platforms and other applications remains more challenging.
“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,” he writes.
AI tools are creating new silos
The report argues that embedding AI into individual applications does not necessarily improve how those applications work together. Marketing teams may have access to AI-enabled productivity tools, spreadsheets and marketing platforms, but these capabilities can remain isolated from one another.
“Embedding AI into every application doesn’t automatically make the applications work together,” Pastore writes.
Instead of connecting systems and enabling information to move across them, companies risk creating what he describes as “AI-powered silos”.
The report also highlights data quality as a fundamental obstacle to effective AI implementation. Pastore cautions that organisations lacking confidence in their data cannot expect AI-generated outputs to be reliable.
“Poor data leads to poor outcomes. It delivers the wrong information to your customers and teammates,” he writes.
He adds that companies need to address data management and governance alongside AI deployment, warning that “garbage in, garbage out”.
More content does not necessarily mean better marketing
Another concern is that AI’s ability to generate large volumes of content could encourage marketers to prioritise quantity over effectiveness.
“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,” Pastore writes.
Drawing on conversations with marketing leaders, he argues that the emphasis on producing more material can overlook whether that content resonates with audiences.
The growing volume of AI-generated output also creates a separate challenge: the time required to review, correct and adapt it.
“AI-generated work isn’t necessarily efficient work,” Pastore writes, pointing to reports from marketers that checking AI’s output can create additional work and undermine the efficiency gains the technology promises.
While prompts can help generate ideas and initial drafts, the report says, further work is often needed to ensure that content meets brand standards and serves its intended purpose.
Pastore also questions whether businesses are using AI to improve customer experiences or simply because the technology makes it easier to produce more content and automate interactions.
“There are times when people want a quick answer, and AI is great. There are times when they need empathy, human touches, and understanding,” he writes.
He urges marketing teams to examine whether their use of AI reflects actual customer needs rather than the availability of the technology.
Proving AI’s return on investment
Demonstrating return on investment (ROI) remains another major challenge. Pastore writes that fewer marketers are confident about proving ROI in 2026 than in 2025, although the article does not provide the underlying survey figures.
As marketing leaders face pressure to justify spending on advertising, campaigns and sponsorships, AI investments are increasingly subject to similar scrutiny.
The report argues that measuring the technology’s impact requires more than counting the number of content assets produced or tasks completed.
“ They need to sufficiently account for rework, integration costs, customer outcomes, and business impact,” Pastore writes, referring to the factors marketers must consider when evaluating AI investments.
The analysis also highlights the risks created by rapidly evolving AI models. Changes to the underlying models can make previously reliable tools behave unpredictably, disrupting workflows that organisations have built around them.
“The amount of chaos this causes depends on how deeply you embedded that AI in the first place,” Pastore writes.
Organisational change remains the challenge
Concluding that AI adoption has outpaced organisational strategy, Pastore argues that companies need to move beyond individual experimentation and address the systems, processes and governance required to use the technology effectively.
“Marketing teams are embracing AI tools, but they haven’t sufficiently transformed their workflows, systems, and organizations to realize the technology’s potential,” he writes.
He adds that the challenge extends beyond marketing departments, requiring coordination between technology and business leaders to integrate AI platforms, existing systems and organisational data.
“Technology and business leaders need to collaborate to develop a strategy to move to the next level,” Pastore writes.
Ultimately, the report suggests that AI will not automatically resolve long-standing marketing challenges around content quality, workflow efficiency, customer experience and business value.
“Marketers have wrestled with these issues for years. They shouldn’t expect AI to solve them.”
