AI in Marketing Agencies Statistics 2027: Adoption, Use Cases & Impact
By Tushar Dey
17 min read
These AI in marketing agencies statistics cover what agency owners and the people who hire them actually ask us about: how many agencies use AI and how deeply, what it gets used for, whether the time it saves shows up in profit, and what it is doing to pricing, headcount and the SEO work clients buy.
We used 20 reports from 17 organizations: the 4A’s trade body, analysts (Forrester, Gartner, McKinsey), large marketer surveys (Salesforce, HubSpot, Marketing AI Institute), Pew, and a few vendor surveys we label as such. Every figure links to whoever ran the research, with year and sample size where published. If we couldn’t trace a number to its source, we left it out.
Last updated: November 2026.
Key AI in marketing agencies statistics (quick list)
- Nine in 10 US marketing agencies use generative AI, and half use agentic AI for marketing execution (Forrester and the 4A’s, 2026).
- Agency genAI use rose from 61% in 2024 to 75% in 2025, though mostly on under half their work (4A’s State of GenAI Adoption 2025).
- 61% of agencies treat AI as a cost of doing business, and fewer than 10% sell genAI as its own line of business (Forrester, 2026).
- 81% of agencies say staff productivity is their main goal for generative AI (Forrester, 2026).
- 22% of CMOs say GenAI has reduced their reliance on agencies for creativity and strategy (Gartner, n=402).
- About one in three agencies has already been asked by a client for an “AI discount” (Productive survey of 180+ agencies, 2025).
- 53% of marketers expect AI to eliminate more marketing jobs than it creates, up from 40% in 2023 (Marketing AI Institute, n=1,858).
- Google users click a regular search result on 8% of visits when an AI summary appears, versus 15% when it doesn’t (Pew Research Center, 2025).
- 88% of marketers have started optimizing for AI-generated answers (Salesforce State of Marketing, n=4,450).
- Legal concerns are the top barrier to agency genAI use, cited by 77% of agencies (4A’s, 2025).
AI adoption in marketing agencies: how many use AI?
Nearly all of them, in some form. How many depends on what “use” means, and surveys define it differently.
The most credible agency number comes from the 4A’s, the US agency trade body, which runs an annual study with Forrester. Its 2025 edition found 75% of agencies using generative AI, up from 61% in 2024. The June 2026 edition puts it at nine in 10, with half also using agentic AI (systems that run multi-step tasks) for execution.
Vendor surveys report even higher numbers. Basis Technologies’ 2025 Advertising Agency Report found 98% of agencies have AI in their workflows and nearly 40% use genAI daily, against under 10% in 2023. NinjaCat’s 2025 survey of 547 US agency decision-makers found 91%.
| Survey | Who was asked | Sample | Uses AI |
|---|---|---|---|
| Basis Technologies, 2025 | Senior agency staff, North America/global | 170+ | 98% |
| NinjaCat, 2025 | US agency decision-makers | 547 | 91% |
| Forrester & 4A’s, 2026 | US marketing agencies | Not published | ~90% (genAI) |
| 4A’s & Forrester, 2025 | US agencies | Not published | 75% (genAI) |
| Salesforce, 10th ed. (fielded late 2025) | Marketing decision makers, global | 4,450 | 75% |
| HubSpot, 2025 | Marketers, global | 1,508 | 66% (74% in US) |
Agencies vs in-house marketing teams
On the numbers we have, agencies look ahead of the brands they work for, though the surveys ask different people different questions, so read the gap as rough. HubSpot’s 2025 AI Trends for Marketers report (1,508 marketers, early 2025) found 66% of marketers use AI in their role. Salesforce’s tenth State of Marketing report (4,450 marketers, late 2025) puts it at 75%.
Depth is another matter. The Marketing AI Institute’s 2025 report (1,882 respondents, 28% from marketing services firms) found only 14% of teams “scaling” AI and 46% piloting. Across all business functions, McKinsey’s 2026 State of AI survey (n=1,719) found 44% say AI is scaling across their enterprise, up from 38%.
Source: Basis (n=170+), NinjaCat (n=547), Forrester & 4A’s, Salesforce (n=4,450), HubSpot (n=1,508)
What this means for you: “We use AI” is no longer a selling point; every agency on a shortlist will say it. Ask how much of the work it touches and who checks it. Nearly 80% of agencies used genAI on less than half their work in 2025, so most shops are still at the bolt-on stage.
The daily-use number is a better gauge than the headline. Basis found under 10% of agency staff used AI every day in 2023, and a third didn’t use it at all. Two years later nearly 40% used genAI daily. That is the shift that changes how a team is staffed and how long a brief takes to turn around.
What marketing agencies use AI for: generative AI in advertising agencies
The use-case side of the AI in marketing agencies statistics is less exciting than the pitch decks suggest: most agency AI use is internal. Forrester’s 2026 study found the top job for generative AI is making staff more productive (81%), followed by summarizing documents and communications (74%) and research and competitive intelligence (70%). Basis found the same pattern a year earlier: ideation and brainstorming (86%) and research (72%) were the top reasons agencies use AI.
Source: Forrester & 4A’s, 2026 and Forrester blog, June 2026
Forrester’s headline is blunt: agencies use AI “to cut costs at the expense of creativity.” Moving from genAI to agents, the share prioritizing creative ideation drops from 65% to 35%. Agencies are pointing agents at operations and keeping idea work with people.
NinjaCat’s agency sample leans more analytical: 74% use AI for data and insights and 72% for campaign strategy and optimization, and 59% say they are actively using AI agents. The common thread across all three surveys is that the first wins are in research, summarizing and reporting. Those are the hours clients rarely see on an invoice, which is part of why the savings are hard to price.
AI writing still gets edited before it ships
Few teams publish AI output untouched. In HubSpot’s survey, 55% of marketers use generative AI for text content, but only 4% use it to write entire pieces. Just 7% don’t edit AI output; 56% significantly revise it and 38% make minor tweaks.
For agencies, that editing step is what the client is paying for. Price and scope it as such, and don’t let it disappear from timelines because “the AI wrote it.” Our list of AI writing tools for agencies rates assistants on brand-voice control and review workflows, which matter more than raw speed.
Time saved and productivity: what the data supports
Time savings is why most people adopt AI, and it’s the benefit the data supports best. The Marketing AI Institute found 82% of marketers say cutting time on repetitive, data-driven tasks is the main outcome they want, the highest share it has recorded.
- 89% of agencies cite better employee productivity as a genAI benefit, and 69% cite efficiencies and cost savings, up from just 13% in 2024 (4A’s, 2025).
- Among CMOs, the top GenAI benefits are time efficiency (49%), cost efficiency (40%) and capacity to produce more (27%) (Gartner, 2025).
- 64% of marketers track productivity gains and 55% track time saved as their main AI success metrics (HubSpot, 2025).
- About 80% of respondents across industries say AI has improved their personal productivity, but only 37% report any positive EBIT contribution (McKinsey, 2026).
That McKinsey gap is the one to remember. People feel faster; far fewer companies can show it in profit. For an agency, saved hours only become profit if you bill the same for less effort or fill the freed time with paid work. Otherwise the client gets the benefit and you get the tool bill.
Be wary of “agencies save X hours a week” headlines. None of the primary sources above measures agency hours saved in a sound way, so we haven’t quoted one. The 4A’s also found about a quarter of agencies have ROI concerns about genAI and about a third struggle with strategy and budget for it, which fits: plenty of activity, thin proof of payback.
If you want your own number, measure it on one repeatable deliverable (a monthly report, a blog brief, a paid search audit). Log hours for a month without AI and a month with it, including review and fix time. That gives you a figure you can defend in a pricing conversation, which no industry survey will.
How AI is changing agency pricing and billable hours
If a task that took six hours now takes two, hourly billing punishes the agency for getting faster. Clients have noticed. Productive’s survey of 180+ agencies found about a third had been asked for an “AI discount,” and nearly half expected it soon. You might expect a rush to value-based pricing. The data says otherwise.
| Share of agencies that… | Figure | Source |
|---|---|---|
| Treat AI as a cost of doing business | 61% | Forrester & 4A’s, 2026 |
| Sell genAI as its own line of business | Fewer than 10% | Forrester, 2026 |
| Plan to monetize agentic AI within 24 months | 31% | Forrester & 4A’s, 2026 |
| Moved away from billable hours to value or outcome pricing | 28% | NinjaCat, 2025 (n=547) |
| Sell AI solutions or automations as a service | 45% | NinjaCat, 2025 (n=547) |
| Use value-based pricing | 31% (2024) to 18% (2025) | Promethean Research, 2026 |
| Asked by a client for an “AI discount” | About 1 in 3 | Productive, 2025 (n=180+) |
The surprise is Promethean Research’s 2026 State of Digital Services report: value-based pricing fell from 31% to 18% of agencies in a year. The roughly 22% that raised rates in 2025 grew faster than the 73% that held steady (the article doesn’t publish its sample size or the growth gap). And in Productive’s March 2026 pulse report (174 agencies), 52% of agencies reporting several positive AI outcomes kept or raised prices while improving margins, up from 47% six months earlier.
Two more details from that pulse report matter if you’re setting prices for 2027. Among agencies still rolling AI out, 61% say they are still working out how to price it, up from 47%. And clients have not been asking for AI discounts any more often than six months before. The pressure is real but it isn’t accelerating, which gives you time to repackage before your next renewal round.
What clients want from agency AI
Clients want AI to make agencies cheaper first and better second. When Forrester asked corporate marketers what they want from agency AI, cost efficiency won by a wide margin. Forrester also notes that AI has stayed a cost of business for most agencies for the third year running.
Source: Forrester, June 2026
Gartner’s 2025 CMO Spend Survey (402 marketing leaders) points the same way. Budgets were flat at 7.7% of revenue, 59% said that wasn’t enough, and 39% planned to cut labor costs and agency allocations. 22% said GenAI had already reduced their reliance on agencies for creativity and strategy.
Clients don’t always understand what they’re asking for: 73% of agency respondents in Digiday’s 2025 Media Agency Report, cited in its April 2026 research, said clients don’t understand agentic AI. In that April 2026 survey of 62 agency professionals, 38% named the effects of AI as their biggest challenge for 2026 (29% external effects such as new tools, 9% internal ones such as workforce development). With a sample that small, treat it as a mood reading.
Consumers are a separate gap. IAB’s January 2026 study (104 ad executives, 505 Gen Z and Millennial consumers) found 83% of executives say their company uses AI in creative, up from 60% in 2024. They think 82% of young consumers like AI-made ads. Only 45% do. That 37-point gap was 32 points in 2024, so it is getting wider as adoption grows.
The practical upshot: agencies should expect procurement to ask where the AI savings went, and have an answer in scope, turnaround or output volume. Clients should ask how the agency discloses AI-made creative. IAB found 89% of advertisers using genAI disclose at least sometimes, but fewer than half always do.
AI and agency jobs: hiring, layoffs and forecasts
Marketers are gloomier about AI and jobs than the headcount data justifies, so far.
What people expect
The Marketing AI Institute found 53% of marketers expect AI to eliminate more marketing jobs than it creates in the next three years, up from 40% in 2023. Only 24% expect the reverse. Among consultants and agency professionals it’s 59%, the gloomiest group in the sample. That makes sense: agency staff do the production work that is easiest to automate.
What is actually happening
- More than 40% of agency respondents said their agency had layoffs in the past 12 months, up from 36% the year before (Basis, 2025). The survey doesn’t say AI caused them; budget cuts and lost accounts are in there too.
- Only 3% of agencies made significant staff reductions because of AI (Productive, March 2026, n=174). Hiring is slowing, though.
- Very few agencies predict headcount growth over the next two years, and most forecast a reduction of less than 10% (4A’s, 2025).
- 39% of CMOs want to reduce labor spending (Gartner, 2025).
The main long-range forecast is three years old. Forrester’s 2023 agency workforce forecast projected US agencies will lose 32,000 jobs to automation by 2030, or 7.5% of the workforce, with genAI behind nearly a third. Clerical (28% of genAI-driven losses), sales (22%) and market research roles (18%) were most exposed; the more original the work, the safer.
For agency owners: so far this means slower hiring, not mass layoffs, with junior production and reporting roles hit first. Train people on review, strategy and client work, which clients still pay a premium for.
AI search, GEO and the new services agencies sell
The biggest change AI has made to agency services is in where clients’ customers find information. AI Overviews and assistants like ChatGPT answer many questions without a click, which is pushing agencies to sell generative engine optimization (GEO), also called answer engine optimization.
| AI search signal | Figure | Source and sample |
|---|---|---|
| Monthly users of Google AI Overviews | 2 billion+ | Alphabet, Q2 2025 |
| Google searches that showed an AI summary | 18% | Pew, March 2025, 68,879 searches |
| Click on a regular result, with vs without AI summary | 8% vs 15% | Pew, 900 US adults |
| CTR change for position one when an AI Overview appears | -58% | Ahrefs, 300,000 keywords, Dec 2025 |
| Value of an AI search visitor vs an organic visitor | 4.4x | Semrush, 2025 |
| Marketers optimizing for AI-generated answers | 88% | Salesforce, n=4,450 |
Pew’s study of real browsing data is the cleanest evidence, because it tracked what 900 US adults did rather than what they said. Only about 1% of visits to a page with an AI summary ended in a click on a link inside the summary, and 26% ended the browsing session there, against 16% without one. Ahrefs’ February 2026 update put the click loss for position one at 58%, up from 34.5% in its April 2025 study (the new comparison spans two years, so part of the jump is method).
The upside: AI referral traffic converts well. Semrush’s AI search study estimated an AI search visitor is worth 4.4 times a traditional organic visitor, based on conversion rate, across 500+ marketing and SEO topics. Treat its projection that AI search visitors overtake traditional search visitors for those topics by early 2028 as a vendor extrapolation, and note that Semrush doesn’t publish the visit volumes behind the 4.4x.
Salesforce found high performers are 2.2 times more likely than underperformers to have optimized for AI search, and 85% of marketers say AI is reshaping their SEO strategy. In the proposals we see, GEO has gone from an add-on pitch in 2024 to a standard line item. We track the firms doing it well in our rankings of GEO agencies and AEO agencies, and our SEO agency statistics cover the effect on SEO pricing and results.
If you’re buying: ask any SEO agency how it measures visibility in AI answers as well as rankings. A team that can’t show citation tracking across ChatGPT, Perplexity and AI Overviews is still selling 2022 SEO.
Barriers: legal risk, accuracy and training
Legal risk is the biggest brake on agency AI use. The 4A’s found legal concerns were the top barrier for 77% of agencies in 2025, the same as in 2024. Lack of understanding of genAI tools more than doubled, to 42%. In Forrester’s 2026 study, the top barriers were accuracy and bias (63%), legal concerns (62%) and privacy and security (55%). For AI agents, the blockers were lack of expertise (54%) and data infrastructure gaps (51%). NinjaCat’s agency respondents put technical integration (57%) and resistance to change (55%) at the top, and 32% had tried building their own AI tools and hit integration problems.
Training is the gap nobody is filling. The Marketing AI Institute found 62% of marketers cite lack of training as a top barrier, and 68% effectively get none from their employer. 75% have no AI roadmap, and 63% have no generative AI policy or don’t know of one.
Fix this and it becomes a selling point. A written AI policy, a disclosure approach and a documented review step answer most questions from client legal and procurement teams. When we review shops for our list of AI marketing agencies, a written policy is one of the first things we ask for.
How to read AI in marketing agencies statistics
Before you quote any of these AI in marketing agencies statistics in a pitch deck or board report, check these:
- “Uses AI” varies by survey. Some count anyone who tried ChatGPT once; others ask about scaled deployment. The 98% and 14% figures above answer different questions.
- US skew and small samples. The 4A’s, Forrester and NinjaCat cover US agencies only. Digiday’s survey had 62 respondents and IAB’s had 104 executives: fine for direction, weak for precise percentages.
- Forecasts aren’t data. Forrester’s 2030 jobs forecast and Semrush’s 2028 search projection are models. Quote them as predictions, and treat anything older than 2024 as history.
- Gated reports. The Forrester/4A’s 2026 study and the 4A’s 2025 study don’t publish sample sizes on their public pages, and Gartner’s and McKinsey’s pages block or time out for many readers. We link the publisher’s landing page; quote the publisher, not the site you found it on.
What these AI statistics for agencies mean for you
Read together, the AI in marketing agencies statistics above say three things: use is close to universal, depth is still shallow, and the money mostly hasn’t moved yet. Here is what we would do with that.
If you run an agency
- Move production work to fixed-scope pricing before clients ask for an AI discount. A third already have.
- Decide whether AI is a cost or a product. 61% treat it as a cost; 31% plan to monetize agentic AI within two years, and NinjaCat found 45% already sell AI automations as a service. Our ranking of AI automation agencies shows what productized AI services look like.
- Write an AI policy and a disclosure statement. Legal concerns block 77% of agencies, and clients’ legal teams have the same worries.
- Automate reporting and admin first, where the gains are clearest. Start with our list of workflow automation tools.
If you hire agencies
- Ask how much of your work AI touches and who reviews it. “We use AI” tells you nothing.
- Ask for savings as more output or faster turnaround. A fee cut on its own tends to cut the senior time on your account.
- Check AI search visibility in any SEO proposal. With click-through rates falling where AI Overviews appear, rankings alone are an incomplete measure.
You can filter our agency directory by service to find shops offering AI, GEO and automation work.
FAQ
What percentage of marketing agencies use AI?
Nine in 10 US marketing agencies use generative AI (Forrester and the 4A’s, 2026), up from 75% in 2025 and 61% in 2024. Basis puts it at 98%. Far fewer use it across most of their work.
How are marketing agencies using AI?
Mostly for internal productivity: 81% name staff productivity as their main genAI goal, 74% use it to summarize documents and 70% for research (Forrester, 2026). Ideation (86%) topped Basis’s 2025 list.
Will AI replace marketing agencies?
Not wholesale. 22% of CMOs say GenAI has reduced their reliance on agencies (Gartner, 2025), and Forrester forecasts 7.5% of US agency jobs will be automated by 2030. Only 3% of agencies have made significant AI-driven staff cuts (Productive, 2026).
Is AI changing how agencies charge clients?
Slowly. A third of agencies have been asked for an “AI discount” (Productive, 2025) and 28% have moved away from billable hours (NinjaCat, 2025). But 61% treat AI as a cost of business, and value-based pricing fell from 31% to 18% (Promethean Research).
How much does AI search reduce website traffic?
Pew found users click a regular result on 8% of visits when an AI summary appears, versus 15% without. Ahrefs estimates AI Overviews cut click-through for the top result by 58%, which is why many SEO agencies now sell GEO too.
Where do AI in marketing agencies statistics come from?
Mostly from trade bodies (the 4A’s), analysts (Forrester, Gartner) and software vendors surveying their own users (Basis, NinjaCat, Productive). The 4A’s and Forrester data is the most agency-specific; treat vendor numbers as upper bounds.
Sources
- Forrester & 4A’s, State Of AI Inside US Marketing Agencies, 2026
- Forrester blog, June 2026
- 4A’s, State of GenAI Adoption 2025
- Forrester, Agency AI Workforce Forecast 2030 (2023)
- Gartner 2025 CMO Spend Survey
- Marketing AI Institute, 2025 State of Marketing AI
- Salesforce State of Marketing, 10th edition
- HubSpot 2025 AI Trends for Marketers
- McKinsey, The State of AI 2026
- Basis 2025 Advertising Agency Report
- NinjaCat 2025 agency research
- Productive, Agencies in the AI Era (2025)
- Productive pulse report (2026)
- Promethean Research, 2026 State of Digital Services
- Pew Research Center, AI summaries and clicks (2025)
- Ahrefs, AI Overviews click study (2026)
- Alphabet Q2 2025 earnings remarks
- Semrush AI search study (2025)
- IAB, The AI Ad Gap Widens (2026)
- Digiday+ Research (2026)
