How marketers use autonomous AI (and where they still want control)

Artificial intelligence has made incredible strides in what it can accomplish, particularly in enhancing marketing and sales team processes. But does that mean AI is ready to run entire campaigns autonomously? According to our new survey of marketers, not by a long shot.

The 2025 AI and Marketing Performance Index explores how high-performing marketing teams are using AI to accelerate growth and where they see the greatest value for it. One thing is clear: Humans are essential governors of AI systems, and there are just a few areas we trust AI to act freely. 

Let’s jump into some of the report findings, including how marketers are using AI, the types of decisions AI can make, and why humans need to be in the loop so you can leverage AI effectively.  

Marketing agility drives revenue

If you feel pressure to accomplish more as a marketer in less time (and possibly with fewer resources), you’re not alone. The majority of marketers (89%) feel significant pressure to produce more and deliver better outcomes. 

When it comes to producing those results, our research also reinforces why agile marketing is so important. Organizations with “fairly fast” or “extremely fast” marketing cycles (54%) were more likely to have significant revenue growth in 2024.

The highest-performing teams — those who experienced significant revenue growth at their company in the last year — are typically able to launch campaigns very quickly:

• 70% of high-growth teams can launch a campaign in under 30 days

• 41% can do so in under 15 days

• 14% can do so in under 7 days

Slower-growth teams, by contrast, commonly take 16-30 days to launch a campaign (38%), or more than 31 days (18%). 

The customer experience difference becomes more pronounced between organizations that effectively activate their first-party data and those that struggle to execute informed and timely campaigns. Without the right blend of technology and processes, some teams will fail to keep up and gradually lose customers to competitors. 

Common hurdles for marketing teams (and how AI helps)

So what’s holding teams back? The top five challenges preventing marketing from growing faster or achieving better results are:

• Delayed approvals and decision-making (40%)

• Limited resources/budget (38%)

• Difficulty identifying/engaging the right audience (30%)

• Manual processes/limited automation (29%)

• Reliance on input from other teams (29%)

AI can assist many of these areas, like extending team resources, identifying audience segments, and automating redundant and time-intensive tasks. By handling these areas, AI frees up time for marketers to tackle the operational backlogs that require human-to-human collaboration.

How marketers use AI for better campaigns

37% of marketers say integrating AI into team processes is a top priority over the next year to improve their marketing performance. 66% even believe that AI-powered technology could accelerate their marketing growth by 10-50%. 

These estimates aren’t far-fetched. Compound marketing driven by agentic AI can achieve 1% growth daily, but we must build that momentum step by step. 

Teams are currently using AI to improve and accelerate activities like:

• Predicting customer behavior (39%)

• Content generation or ideation (38%)

• Generating insights and recommendations (36%)

• Personalizing content at scale (36%)

• Optimizing campaign journeys (31%)

These uses complement team priorities, including accelerating response to shifting customer behavior (31%), using real-time data to adjust campaigns quickly (30%), and speeding up campaign execution (29%).

Each of these use cases is a valuable starting point for AI given its ability to monitor trends and analyze data instantaneously. AI creates highly accurate customer personas for campaigns, which improves the success of all outreach. Agentic AI, in particular, can quickly learn from its outreach to continually iterate its approach. This sparks a flywheel of ongoing improvement at a scale humans could never accomplish alone. 

How marketers want to use AI

Some organizations are focused on just one or two AI use cases as they learn the technology. Others are using AI to support every step of their campaign development. 

When evaluating priorities for their AI implementations, marketers see several areas as ripe for potential. Coincidentally, these areas also overlap with the current process bottlenecks they face:

• 66% want to use AI for data analysis and insights, which is a bottleneck for 53% of teams.

• 52% want to use AI for campaign development, a bottleneck for 47% of teams.

• 44% want to use AI for audience segmentation, a bottleneck for 40% of teams.

These three areas form a complete marketing cycle. AI will segment audiences, suggest the ideal campaign channel and message, and analyze campaign results to fuel improvements. But just because AI can complete entire cycles doesn’t mean it should do so without human oversight.

When to let AI make marketing decisions (and when to rein it in)

Teams can use AI as a creative partner and greatly accelerate their campaign development, but they aren’t willing to completely hand over control to AI. 

• 86% of marketers think AI is better with human intervention.

• 46% believe AI should make low-risk decisions with human oversight.

• 31% believe AI should only provide insights while humans make all final decisions.

• Only 4% believe AI should be in full control of executing marketing efforts. 

We agree that AI needs human oversight. Agentic AI is a powerful marketing partner but it can’t replace human expertise and creativity. 

Organizations can get started using AI with low risk for marketing needs like:

• Audience segmentation - Allow AI to review your customer personas and suggest segments for upcoming campaigns. AI uses complete, real-time data from a data cloud to ensure your personas reflect current behaviors and accurate information. 

• Journey orchestration - AI can help curate personalized omnichannel experiences that appeal to individual buyers. AI can decide the right channels to approach customers on, so teams don’t have to guess which channels will be most effective.

• Campaign analysis - AI can review campaign results and suggest improvements for the next iteration. AI is especially valuable for conducting campaign tests to quickly learn what resonates with individual buyers. 

Other areas require heavier oversight when using AI. Marketers ranked the following areas as the highest for needing human intervention:

• Predicting customer behavior - Despite being one of the most common use cases for AI in marketing, humans are essential for predicting customer behavior and dissecting omnichannel journeys. AI delivers essential data insights to inform human decisions, but AI can’t predict behavior alone. 

• Personalizing content at scale - Personalization is a major priority for marketing teams and AI can help us accomplish 1:1 personalization at scale. However, humans need final approval on all content and to identify personalization opportunities where AI can help. 

• Content creation - Most content created by AI systems requires a thorough review to ensure it accurately reflects your brand's voice. AI can help marketers overcome writer’s block and accelerate content creation, but teams still need a human review of any content generated by AI.

Using AI to unlock compound marketing growth

Agentic AI will forever shape how marketers deliver fully personalized journeys and continually improve their customer experiences. Although many companies are still early in their AI journey, those that use AI effectively can start a cycle of rapid innovation.

Marketers should actively explore areas where AI can save their time and deliver insights to inform campaigns. By starting in low-risk areas like audience segmentation, marketers can get comfortable using AI in more areas. The key is to enforce the necessary governance to ensure everything AI creates fulfills the brand’s vision and promise to customers. 

GrowthLoop’s Compound Marketing Engine includes several powerful AI agents to empower marketers to immediately activate the potential of their first-party data. If you’re ready to equip your marketing teams with agentic AI, learn more and book a demo today.