AI agents are now part of real marketing work, but only a minority of teams let them run tasks on their own. Only 13% of marketing teams tap into self-directed agents, and 19.2% of marketers use agents to run campaigns end to end. Yet 75% of marketers already use AI in some form. So the gap between plain AI and true agents is still wide.
We compiled these numbers from primary surveys, press releases and company filings published this year. Then we ran our own math on the budget data and noted who was asked in each survey. So every figure here has a date and a named owner in the list at the end.
Our Summary Table of AI Agent Marketing Numbers
First, we put the full set in one table. We grouped each number by theme and kept the survey base next to it. So you can check how big each sample was before you quote it.
| Theme | Statistic | Who was asked | Published |
|---|---|---|---|
| Adoption | 75% of marketers have adopted AI | 4,450 marketing decision makers, four regions | Feb 2026 |
| Adoption | 13% of marketing teams use self-directed agents | Same global marketer survey | May 2026 |
| Adoption | 19.2% use AI agents to run campaigns end to end | 1,500+ B2B and B2C marketers | 2026 |
| Adoption | 21% of video ad buyers run agentic AI live | US digital video buyers | May 2026 |
| Awareness | 96% of ad buyers know about agentic ad buying | 205 US buy-side decision makers | Jan 2026 |
| Spend | 15.3% of marketing budgets go to AI | 401 CMOs in North America, the UK and Europe | May 2026 |
| Spend | Agentforce ARR passed $1.5 billion | Salesforce quarterly results | Aug 2026 |
| Results | Up to 8 hours a week saved with agents (expected) | Global marketer survey | May 2026 |
| Outlook | Automated marketing work to reach 36% by 2028 | 402 CMOs | May 2026 |
| Outlook | 60% of brands to use agentic AI for one-to-one contact by 2028 | Gartner forecast | Jan 2026 |
| Trust | 50% of consumers prefer brands that avoid GenAI content | 1,539 US consumers | Mar 2026 |
Also, note the mix of bases. Some numbers come from marketers, some from ad buyers, and some from CMOs. So a 13% figure and a 21% figure do not clash, because they measure different groups.
How Many Marketing Teams Run Agents Today
Plain AI use is close to universal. In a survey of more than 1,500 marketers, 86.4% of teams use AI in at least a few areas. In fact, only 1.7% neither use it nor plan to. A separate global survey of 4,450 marketing decision makers put AI adoption at 75%.
Agents are a different story. An AI agent does not just draft copy. Instead, it takes a goal, picks the steps and acts across tools. Only 13% of marketing teams tap into self-directed agents. Besides, 19.2% of marketers use agents to automate campaigns end to end.
Full rollout lags too. Just 39% of marketing teams have reached full AI integration. For the other 61%, the work is still in progress. So most teams use AI tools, but few have wired them into one system that can act alone.
Where Marketers Use AI Most Often
Content is still the main job for AI. For example, 42.5% of marketers use it extensively for content creation, and 38% use it now and then. Then come admin tasks, media creation and ad automation. We listed the extensive-use rates for each area.
| Area of marketing work | Use AI extensively | Use AI occasionally | Total (our math) |
|---|---|---|---|
| Content creation | 42.5% | 38.0% | 80.5% |
| Media creation | 37.2% | 37.7% | 74.9% |
| Administrative tasks | 35.6% | 40.5% | 76.1% |
| Learning and skills | 34.6% | 40.1% | 74.7% |
| Ad automation and optimization | 34.1% | 36.5% | 70.6% |
| Brainstorming | 33.9% | 39.8% | 73.7% |
| Planning and forecasting | 33.4% | 37.5% | 70.9% |
However, ad automation is where agents tend to show up first. Ad tasks have clear goals and quick feedback. That is why the ad-buying numbers in the next section run ahead of general marketing.
Ad Buyers Are Moving Faster on Agentic Buying
Ad buyers are ahead of other marketers. In a January study of 205 US buy-side decision makers, 96% were aware of agentic AI ad buying. Also, two-thirds plan to focus on agentic AI for buying and campaign work this year.
Next, we charted what those buyers use agents for, or expect to. Analysis and creative testing top the list. Direct deal execution sits at the bottom, at 45%.
Video buyers show how far use has spread. In May, 21% of digital video buyers said agentic AI was live, and 20% were testing it. Then 25% planned to use it, and 28% were looking into it. By our math, 41% are already running or testing agents.
Teams also split on how to get agents. Among the 161 buyers who knew about agentic ad buying, 37% chose a hybrid setup. Then 29% plan to buy or license a third-party tool, 20% lean on partners, and 10% build in-house. So few teams want to build their own.
What CMOs Spend on AI and What Our Budget Math Shows
CMOs put 15.3% of their marketing budgets into AI work this year. That came from a survey of 401 marketing leaders, and most of their firms earn over $1 billion a year. Besides, total marketing budgets rose to 7.8% of company revenue, up from 7.7% a year earlier.
Still, readiness is low. Only 30% of CMOs report mature AI skills and systems. Meanwhile, 70% say their internal processes are not ready to scale AI. And 56% say they lack the budget to deliver this year’s plan.
We turned those shares into dollars for a sample firm with $100 million in revenue. First, we multiplied revenue by the marketing share. Then we multiplied that budget by the AI share. The AI-ready group spends more on each line, so we ran it too.
| Firm type ($100M revenue) | Marketing budget share | Marketing budget | AI share of budget | AI spend (our math) | AI spend as % of revenue |
|---|---|---|---|---|---|
| Average firm | 7.8% | $7.8 million | 15.3% | $1.19 million | 1.19% |
| AI-ready firm | 8.9% | $8.9 million | 21.3% | $1.90 million | 1.90% |
So an AI-ready firm puts about 60% more money into AI than the average firm of the same size. In fact, the gap comes from two levers at once. These firms give marketing a bigger slice of revenue, and they give AI a bigger slice of marketing.
Vendor Revenue Tracks the Same Shift
Vendor numbers point the same way. Salesforce says Agentforce annual recurring revenue passed $1.5 billion in its quarter reported on 26 August 2026. It says that figure grew over 240% in a year. However, it now also counts Slackbot and other AI products, so it is wider than marketing alone.
By our math, 240% growth to $1.5 billion means about $440 million a year before. Also, Salesforce says its agents and Slack have delivered 7.0 billion “agentic work units” to date. We cover the company’s marketing push in our story on the Salesforce deal for Braze.
Time Saved and Other Results Marketers Report
Marketers expect to save up to eight hours a week with AI agents. That is a full working day. For broad AI use, the reported savings are larger. About one-third of marketers say AI saves their team 10 to 14 hours a week, and another third say over 15 hours.
Output goes up, but so does pressure. For example, 26.5% of marketers say AI raised their productivity a lot, and 66.2% say it rose slightly or moderately. Yet 83.5% say they must now produce more content because of AI.
High performers use agents in more places. Top teams are 1.5 times more likely to answer customers across channels with agentic AI. They are also 2.2 times more likely to have optimized for AI search. In addition, 88% of marketers have begun optimizing for AI answers, which is why tools like ChatGPT rank trackers now exist.
Even so, results are hard to prove without clean tracking. If you test an agent on paid campaigns, tag every link first with our UTM builder. Then compare cost per customer before and after, using the method in our guide to blended CAC versus paid CAC.
Data Gaps and Consumer Doubts Slow Agents Down
Data is the first block. Only 26% of marketers are fully happy with how their data is unified. Also, just 58% have full access to service data, 56% to sales data and 51% to commerce data. An agent cannot act well on data it cannot see.
Data and agent use go together. Teams with well-unified data are 60% more likely to use AI agents. Besides, 81% of marketers would trust AI to reply to customers, yet scattered data holds them back. Also, 84% admit they still send generic campaigns.
People are the second block. In a survey of 402 CMOs, 80% said staff fear and anxiety get in the way of AI tests. So training and clear rules matter as much as tools. Our template for an AI acceptable use policy covers the basics.
Consumers Want AI Content Labeled
Buyers are wary too. Half of US consumers in a survey of 1,539 said they would prefer brands that do not use GenAI in their messages and ads. Also, 68% often wonder whether the content they see is real.
Labels ease some of that worry. In a smaller panel of 335 US consumers, 78% rated clear AI labels as very important or the top factor for trust. So any agent that writes to customers should follow a labeling rule from day one.
Forecasts for Agentic Marketing Through 2028
Marketing leaders expect AI to do far more of the work soon. They put AI-driven automation at 16% of marketing work this year and 36% by 2028. By our math, that is 2.25 times as much automated work in about two years.
Gartner says 60% of brands will use agentic AI for one-to-one customer contact by 2028. It also says these agents will act like digital concierges for customers. However, it expects GenAI shopping tools to bring in less than 10% of e-commerce revenue.
Then there is the ad market. Two-thirds of US ad buyers already plan to focus on agentic buying this year. So the first big wave of marketing agents will likely land in paid media before it reaches brand or content work.
A Readiness Checklist We Built From These Numbers
We turned the main blockers in the data into a short checklist. Each line maps to a number above. So if your team fails a line, you know which gap is holding agents back.
- Unify customer data first, since only 26% of marketers are fully happy with theirs.
- Pick one paid-media task for a pilot, because ad buyers report the widest agent use.
- Set a budget line for AI, using 15.3% of marketing spend as the current average.
- Keep a person on deal sign-off, as only 45% of buyers see agents running direct deals.
- Label AI content, since 78% of consumers in one panel rate labels as very important.
- Train staff early, because 80% of CMOs name fear as a barrier.
For context on how AI is changing other trades, see our construction marketing statistics. The same pattern shows up there, with wide AI use and slower moves toward full automation.
What Marketing Leaders Want to Know About AI Agents
What percentage of marketers use AI agents?
About 13% of marketing teams use self-directed agents, and 19.2% of marketers use agents to run campaigns end to end. By contrast, 75% to 86.4% use AI in some form, depending on the survey.
How much of the marketing budget goes to AI?
CMOs put 15.3% of marketing budgets into AI this year. AI-ready firms put in 21.3%. For a $100 million firm, our math gives about $1.19 million for the average and $1.90 million for AI-ready firms.
What is agentic AI in marketing?
It is AI that takes a goal and acts on it across tools, with little human input. For example, an agent can shift ad budget, test creatives or reply to a customer. A chatbot that only drafts text is not an agent.
How much time do AI agents save marketers?
Marketers expect to save up to eight hours a week with AI agents. For AI use in general, about one-third report 10 to 14 hours saved per team each week.
Will AI agents replace marketers?
The data points to more automation alongside people. Leaders expect 36% of marketing work to be automated by 2028. Still, 73.4% of marketers see AI working alongside them on most of their duties.
Marketing Surveys and Filings Used for This Data
- Salesforce: 75% of Marketers Have Adopted AI, Yet Still Use It to Send Generic Campaigns (19 February 2026)
- Salesforce: 10 Ways Agentic AI Is Supporting Marketing Teams, State of Marketing 10th Edition data (20 May 2026)
- Salesforce: Tenth Edition State of Marketing report page (27 January 2026)
- HubSpot: State of Marketing and trends report, 2026 edition (updated 2026)
- HubSpot: AI predictions for marketing, citing the 2026 State of Marketing report (2026)
- IAB: 2026 Outlook Study, A Snapshot of US Ad Spend, Opportunities and Strategies for Growth (January 2026)
- IAB: US Digital Video Ad Spend to Surpass $80B in 2026 (5 May 2026)
- Gartner: 2026 CMO Spend Survey on AI budgets and readiness (11 May 2026)
- Gartner: Marketing Leaders Expect AI Automation of Marketing Work to Double by 2028 (11 May 2026)
- Gartner: 60% of Brands Will Use Agentic AI for One-to-One Interactions by 2028 (15 January 2026)
- Gartner: The Future of Marketing, Trends and Predictions for 2026 (2026)
- Gartner via Business Wire: 50% of Consumers Prefer Brands That Avoid GenAI in Consumer-Facing Content (16 March 2026)
- Salesforce: Record Second Quarter Fiscal 2027 Results (26 August 2026)
