A majority of marketers—63%—use generative AI tools, according to AI platform Jasper’s survey of 503 marketing leaders, published in its State of AI in Marketing 2025 report. What’s more, 79% of marketers plan to either begin using the technology or expand their usage this year.
Yet only 49% of those using genAI are measuring the return on their investments. Even among teams of more than 1,000 marketers, only 62% measured ROI; among teams with fewer than 25 marketers, that fell to 38%.
This sort of failure to quantify results would normally be anathema to marketers. But many businesses, motivated by a need to keep up with marketplace trends, are implementing AI before they’ve determined exactly what they want from it. “People feel like they have to get out from behind,” says Loreal Lynch, Jasper’s Chief Marketing Officer. “They rush getting these use cases set up and don’t think about the business outcomes and how they’re going to measure them.”
This, in turn, causes many genAI projects stall. A team might launch a pilot program, but because they can’t quantify the business impact, they’re unable to make a convincing case to leadership for continuing. “You need to be very thoughtful and think about your business outcomes first, and then back your project into it,” Lynch says.
Key AI Metrics to Track
Of those marketers who do measure ROI, 58% track revenue outcomes per marketing expenses, making that the most common metric. The second most commonly tracked metric was output per marketing resource (43%), followed closely by number of marketing resources per dollar of revenue (42%) and revenue outcomes per marketing resources (40%).
And while the majority of marketers might not be able to prove AI’s impact on the bottom line just yet, they are reporting less-tangible benefits: More than three-quarters (78%) are enjoying greater job satisfaction following the implementation of AI.
AI Facilitates Creativity at Scale
AI is arguably most touted for its ability to improve predictive models and analysis and to automate workflows. Yet only 23% of the marketers surveyed are using it for predictive analysis and 26% for workflow automation. Instead, it is most commonly being used for content creation (57%) and idea generation (55%).
One reason for this, says Lynch, is that generating content is one of the simplest applications for marketers to implement. However, “the more mature and larger organizations aren’t using it for single pieces of content but for workflow and multiple pieces of content,” she adds. Designers and writers develop and refine the initial creative, and from that AI creates myriad assets for multiple channels and markets. “One piece of content can easily scale into 1,000 pieces. AI helps facilitate the content supply chain,” but does not replace the human in the loop.
Forty-five percent of marketers now use genAI for SEO, which Lynch says provides another example of why people remain essential to marketing. “The human does the keyword research. The AI agent could let you know you dropped in ranking and make recommendations,” she explains. “AI prompts you instead of you prompting it.”
Among other key findings:
- Data privacy and security concerns were the top reason respondents gave for not adopting AI, with 21% citing them. Among companies with more than $1 billion in annual revenue, however, brand governance was the number-one reason, mentioned by 33%; 31% cited concerns about output quality.
- Less than half—46%—of respondents said their company has a documented AI policy and guidelines in place. That includes 56% of the largest companies but only 33% of the smaller ones.
- 65% of CMOs expect AI to “greatly” affect team roles and structure, compared with 28% of field marketers. Overall, 48% of marketers expect AI to have a major impact on their team, and another 39% anticipate moderate changes.
- Another area of disconnect between the C-suite and the rest of the marketing team: 26% of CMOs rate their organization’s AI maturity level as “very advanced,” compared with just 3% of directors and 8% of managers.