AI adoption is rising faster than marketing returns

Why AI adoption is rising faster than marketing performance

Only 6% of marketing organizations report significant performance improvements from AI, even as the technology becomes a standard part of marketing operations. Companies are expanding their use of AI rapidly, yet substantial financial gains remain concentrated among a relatively small group of higher-performing organizations.

New research from Bain & Company points to organizational decisions as a key difference. Higher-performing marketing teams are more likely to coordinate AI investment centrally, redesign workflows, restructure roles, and use customer intelligence to guide decisions. These practices are associated with stronger financial outcomes, even as other companies catch up in basic AI adoption.

For chief marketing officers (CMOs), the findings raise a practical question: how should organizations evaluate AI investments when access to the technology is widespread? Employee adoption rates and automated task counts help measure deployment. Understanding business value requires examining how AI changes marketing operations.

AI adoption is rising faster than financial returns

Bain & Company’s research, AI in Marketing: How Leaders Achieve Double the Revenue Impact, compares marketing organizations with different levels of business performance. The findings draw on the 2026 Leaders and Laggards Survey of 1,397 respondents and Bain & Company’s 2025 Marketing Survey of 1,232 respondents.

The comparison identifies 121 marketing leaders, defined as companies achieving 11% annual revenue growth and seven-point annual market share growth. Bain examines how these organizations differ from lower-performing companies, which it classifies as laggards.

Both groups increased their use of AI substantially between 2025 and 2026. But adoption grew faster among laggards, narrowing the gap between the groups.

The adoption gap fell from 27 percentage points to 17 percentage points in one year. Laggards have also nearly caught up with leaders in routine applications such as content creation and media buying. As these uses become common, adoption rates reveal less about why some organizations achieve better financial outcomes.

The financial differences remain substantial. Bain finds that leaders are twice as likely as laggards to attribute double-digit marketing-driven revenue growth or cost savings to AI initiatives.

This comparison comes from two survey questions asking respondents what proportion of marketing-driven revenue growth and cost savings over the preceding year they attributed to AI-enabled initiatives. It measures how often organizations report substantial AI-attributed benefits. It does not establish that leaders generate twice as much revenue from AI.

Bain identifies three organizational practices associated with the performance gap: centralized AI strategy, extensive workflow redesign, and greater emphasis on customer-focused applications. Together, these practices show how organizations direct investment, change how people work, and use information to improve marketing decisions.

The findings establish associations rather than independent causal effects. Bain & Company also provides consulting services related to AI and organizational transformation, giving it a commercial interest in the management changes its research highlights. The survey comparisons are therefore most useful for understanding how higher-performing organizations operate and where their practices differ.

Centralized strategy directs AI investment

The first difference concerns who decides how AI should be used. Higher-performing marketing organizations are substantially more likely to follow a strategy established by senior leadership.

Bain’s 2026 survey finds that leaders are 1.8 times more likely to have a centralized AI roadmap with clearly defined priorities. Laggards, meanwhile, are 10 times more likely to have no defined marketing AI strategy.

These arrangements shape which projects receive investment and how their outcomes connect to business goals. When marketing teams adopt AI independently, their choices tend to reflect immediate departmental needs. Content teams may focus on creative production, while media teams concentrate on campaign execution and optimization.

Both activities can improve efficiency, but their wider value depends on whether teams share information and coordinate decisions. Faster content production offers limited benefits to the overall operation when audience selection, campaign approvals, and spending decisions still move through disconnected processes.

A centralized strategy allows senior executives to establish common priorities across those functions. Leadership can identify the business outcomes AI should support, decide where shared investment makes sense, and coordinate the teams responsible for implementation.

Consider a customer intelligence initiative that analyzes audience behavior and campaign performance. Its findings become more useful when teams responsible for audience selection, creative development, campaign execution, and spending decisions can act on the same information. Central direction helps establish the responsibilities and processes that make this coordination possible.

Executive sponsorship addresses barriers between teams

Coordinated investment also helps organizations overcome obstacles that individual marketing teams cannot resolve independently.

Bain identifies budget limitations, uncertain return-on-investment targets, legal and risk concerns, and organizational silos as barriers to AI transformation. Addressing them often requires decisions involving finance, technology, legal teams, and senior management.

Executive sponsorship can establish shared investment criteria, coordinate legal and operational requirements, and provide authority to change responsibilities across departments. A marketing manager may be able to introduce an AI application within one team, but redesigning a process that crosses several functions generally requires broader agreement.

That authority matters when AI changes how information moves through the organization. Customer intelligence, for example, may need to flow from analytics into creative planning, campaign execution, and budget management. Each handoff involves people, systems, and decisions that must work together.

Leaders commit larger shares of their budgets

Strategic coordination is also accompanied by greater spending.

More than 40% of leaders allocate at least 11% of their marketing budgets to AI-related activities, compared with approximately one-quarter of laggards. Bain also finds that leaders are twice as likely to devote more than 25% of their annual marketing budgets to AI.

These allocations cover AI use cases, software licenses, data, and research. Although baseline spending is similar between the groups, leaders appear more frequently among organizations making larger financial commitments.

Larger budgets can fund initiatives that span several marketing functions. Integrating AI into operational systems, developing customer intelligence capabilities, and redesigning workflows may require coordinated spending on technology, data, training, and implementation.

Still, investment and organizational strength can reinforce each other. Companies with greater financial resources may find it easier to establish centralized programs, recruit specialized employees, and fund extensive changes. Their spending patterns could therefore reflect existing advantages alongside their AI priorities.

For executives, the question is whether funding matches the changes an organization intends to make. A centralized roadmap needs resources for implementation, while substantial spending needs shared priorities to produce coordinated results.

The next question is what happens when those investments reach the teams doing the work.

Workflow redesign changes how marketing operates

Bain finds that marketing leaders are 3.7 times more likely to fundamentally redesign workflows around AI. They are also twice as likely to restructure teams and job descriptions to reflect AI capabilities.

These differences matter because adding AI to an existing task and redesigning the surrounding process create different operational possibilities.

A marketing organization might introduce generative AI, which creates content from instructions, into its established production process. Employees can use it to draft copy, develop images, or prepare campaign materials while keeping the existing sequence of briefing, review, approval, and distribution.

This can reduce the time spent on individual activities. Yet the overall process may still be constrained by the same approval stages, information gaps, and coordination requirements.

Workflow redesign addresses those constraints together. For example, AI-generated customer insights could inform the initial creative brief, support the development of several campaign options, and help teams evaluate those options before production. Campaign performance could then guide subsequent creative and audience decisions.

In this arrangement, AI changes when information becomes available, who uses it, and how quickly teams respond. Connecting the stages of work allows faster production to support faster decisions.

Faster production creates new operational choices

The potential increase in production capacity is substantial.

One unnamed CMO interviewed by Bain described teams of five to 10 people producing output comparable to traditional teams of 40 or 50 employees. The comparison reflects one executive’s experience with AI-enabled production.

Another CMO described national promotional campaigns moving from planning to launch in a fraction of the time previously required. Shorter production cycles give teams more opportunities to develop campaign versions, respond to changing conditions, and refine creative work.

But additional capacity creates value only when the surrounding processes can keep pace. If campaign approvals, customer analysis, resource allocation, and performance evaluation continue at their previous speed, faster asset creation may have little effect on overall execution time.

The opportunity extends beyond producing more material. Teams can rethink how they plan work, test alternatives, review results, and decide what deserves further investment.

Marketing roles are shifting toward judgment and coordination

Changes in production also affect the skills organizations need.

The CMO who described faster national promotions explained the shift:

“We don’t need as many doers. The process of doing will get faster every day, so you need people that are extremely good at knowing what to create the first time.”

The executive’s point is that deciding what to produce becomes more important as production gets easier. When AI reduces the effort needed to generate content or execute established tasks, marketers can focus more on customer needs, commercial priorities, and the decisions that guide production.

These responsibilities require judgment across several marketing disciplines. Employees need to understand which messages are appropriate, which customer problems deserve attention, and how different activities contribute to business objectives.

One retail CMO anticipated greater demand for employees who can coordinate AI agents across specialized marketing functions. AI agents are systems that can carry out assigned tasks and coordinate multiple steps toward an objective.

A marketer working with several agents might establish campaign objectives, assign activities, review outputs, resolve conflicts, and decide when human intervention is necessary. The role involves coordinating work across the campaign process while retaining responsibility for important decisions.

Another unnamed CMO emphasized recruiting marketers who are:

“generalist enough”

The phrase reflects a preference for employees who understand several marketing functions and can coordinate work across them.

During the expansion of digital marketing, many organizations developed specialized roles around individual channels, technologies, and execution tasks. AI is prompting some executives to reconsider how those responsibilities should be distributed.

Specialist knowledge remains valuable when employees must evaluate complex customer, channel, or technical decisions. The emerging emphasis is on combining that expertise with broader commercial judgment and the ability to coordinate multiple activities.

These accounts reveal changing expectations for marketing roles. They also help explain why Bain’s higher-performing organizations are more likely to reconsider team structures alongside their workflows.

Integrated tools make redesigned workflows practical

Redesigned workflows also depend on how AI connects to the systems employees already use.

Bain finds that leading organizations are 2.4 times more likely to integrate AI directly into their marketing tools. Meanwhile, 6% of leaders report relying on general-purpose AI applications, compared with 16% of laggards.

General-purpose products such as ChatGPT, Copilot, and Gemini can support writing, analysis, research, and planning. They commonly rely on foundation models, large AI models capable of supporting many different applications.

These tools can be useful across marketing functions. The operational question is how their outputs reach the systems where employees make and execute decisions.

Consider a marketer using a general-purpose assistant to analyze campaign results. The employee may need to collect performance data, provide context, interpret the response, and transfer recommendations into another application.

When AI is integrated into campaign management tools, relevant information can appear directly within the workflow where decisions happen. This reduces manual transfers, connects analysis with execution, and makes it easier for teams to act on updated information.

Integration can also connect customer intelligence, campaign activity, and performance evaluation. These connections matter because a decision made in one function often affects work elsewhere.

Leaders and laggards generally have access to similar underlying foundation models. Bain’s findings associate stronger performance with differences in how organizations incorporate those capabilities into everyday marketing operations.

Once teams can produce work faster and share information across connected systems, they have more opportunities to test customer decisions. This creates another major difference between leaders and laggards.

Customer experimentation creates more opportunities to learn

Bain finds that leaders are 5.2 times more likely to prioritize customer-centric AI applications than laggards. They are also 1.5 times more likely to identify personalization and improved customer experience as intended outcomes of AI deployment.

The personalization comparison comes from a survey question asking organizations what they hope to achieve through AI. It therefore measures strategic priorities rather than observed improvements in customer experience.

Those priorities shape where organizations apply the technology. Customer-focused AI applications can help teams analyze consumer behavior, identify emerging preferences, coordinate activity across channels, and develop more relevant interactions.

For example, customer intelligence can guide audience selection, message development, and decisions about how experiences differ between channels. AI-enabled omnichannel coordination, which connects activity across customer contact points, can help teams apply those decisions consistently.

The value grows when organizations repeatedly evaluate their assumptions and use the results to improve subsequent decisions.

Leaders run more marketing experiments

Bain’s largest reported behavioral difference concerns experimentation.

Marketing leaders are 8.5 times more likely than laggards to conduct at least 100 marketing experiments per month.

This comparison measures the relative likelihood of reaching the survey’s highest experimentation category. It does not mean that leaders conduct 8.5 times as many experiments on average.

The survey asked organizations how many new marketing experiments they design and launch monthly for important brands in priority markets.

Respondents selected from the following categories:

Experimentation allows marketers to test assumptions about customers. A campaign team might believe that a particular message will perform better with a specific audience. Testing provides evidence to guide the next decision.

More frequent experiments create additional opportunities to examine messages, audience selections, creative approaches, and customer experiences. AI can support this work by reducing the effort needed to develop alternatives, analyze information, and interpret outcomes.

But conducting more tests creates value only when organizations use what they learn. Results that stay within one department, or require lengthy approvals before influencing execution, have limited operational effect.

Bain’s findings suggest that leaders are more likely to combine experimentation with redesigned workflows and coordinated decision-making. This allows them to move from producing alternatives to deciding which ones deserve further investment.

Nestlé uses AI to accelerate product concept development

Nestlé provides an example of how customer intelligence can shape decisions earlier in development.

The company built a proprietary AI platform to improve consumer understanding and accelerate product concept testing. Its implementation brings market research, idea generation, and evaluation into a shared process.

Employees were trained to use the internal platform to analyze market trends and signals from social media. Those inputs helped teams identify consumer interests and develop potential product concepts.

AI then supported the generation and refinement of those concepts. The process also included simulations of consumer responses, allowing employees to evaluate proposed ideas before moving them further through development.

This sequence connects activities that might otherwise involve separate research, creative, and evaluation stages. Market intelligence informs concept generation, concepts become candidates for testing, and simulated feedback helps teams refine them.

Nestlé reported that the initiative reduced its development cycle from six months to six weeks. The shorter cycle allows the company to move more quickly from identifying a consumer opportunity to developing and evaluating a corresponding product concept.

That acceleration can also increase the number of ideas an organization can explore within a given period.

Nestlé additionally reported nearly three times the return on investment, with financial gains reflecting both cost reductions and top-line growth.

These are company-reported outcomes. Nestlé has a commercial interest in demonstrating the effectiveness of its investment, and the reported results do not isolate how much improvement came from AI independently of the surrounding operational changes.

Still, the implementation illustrates a substantial change in how development decisions are made. Employees learned a new platform, incorporated its outputs into several stages of work, and used AI-supported customer intelligence to refine concepts before committing further resources.

The shorter development cycle reflects the combined effect of technology, employee skills, and redesigned processes. It also shows how experimentation can shape decisions beyond campaign production.

Customer learning becomes commercially useful when spending changes

Experiments generate information about customer behavior. Their financial impact depends partly on whether organizations use that information to change strategy and resource allocation.

Bain finds that nearly 70% of marketing leaders regularly or extensively adjust strategy and spending using AI-generated insights. Among laggards, the corresponding figure is 31%.

This difference suggests that leaders are more likely to incorporate AI into recurring management decisions.

A marketing team might discover that a particular audience responds better to a different message or channel. The next decision is whether that finding should change campaign planning, resource allocation, or the wider marketing strategy.

When spending stays unchanged despite useful new information, the organization captures fewer potential benefits from its experiments. Frequent testing becomes more valuable when managers can adjust execution based on the results.

Bain reports that leaders are twice as likely to use AI for demand forecasting and revenue modeling. They are also three times more likely to continuously redistribute spending across marketing channels using AI-generated insights.

Demand forecasting estimates future customer demand, while revenue modeling examines the potential commercial effects of different decisions. AI-supported analysis can help organizations update these estimates as new information becomes available.

For example, a company may combine revised demand expectations with campaign performance data to reconsider where additional marketing spending is most likely to support its objectives.

This changes the rhythm of marketing management. Traditional planning often establishes spending priorities before execution, followed by periodic reviews. AI-informed operations can support more frequent reassessment throughout a campaign or planning period.

Managers still need to consider business objectives, risk, customer needs, and constraints beyond the available data. AI gives them additional information and more opportunities to evaluate their choices.

Experimentation and resource adjustment therefore become part of the same process. Tests generate evidence about customer behavior. Forecasting and modeling help managers interpret the business implications. Budget decisions determine how the organization responds.

This process becomes especially important when customer-facing changes are expensive or difficult to reverse.

Walmart tests physical store changes with digital twins

Walmart shows how AI-supported experimentation can extend into physical operations.

The retailer has created more than 1,700 digital twins using Nvidia Omniverse. A digital twin is a digital representation of a physical environment used to simulate conditions and evaluate proposed changes.

Walmart’s implementations include photorealistic three-dimensional representations of stores. These environments allow the company to examine customer traffic, test store layouts, and evaluate merchandising changes before implementing them physically.

The process includes testing planograms, which specify how products should be arranged and displayed in a store.

AI agents help anticipate customer behavior and identify potential congestion points within the simulated environments. Walmart can use those findings to adjust layouts or merchandising plans before employees make changes in physical stores.

The retailer reports that its digital twin technology has made store resets 10% faster, generating time and cost savings across its retail network.

Walmart has a commercial interest in demonstrating the value of these investments, so the reported improvement represents the company’s account of its implementation.

The operational application is straightforward. Physical store changes require coordination, labor, and resources. Simulating proposed adjustments allows Walmart to evaluate potential outcomes before committing those resources.

The approach connects customer behavior analysis with merchandising execution. Teams can identify problems such as congestion and reconsider layouts before implementation begins.

Nestlé and Walmart illustrate different uses of AI-supported experimentation. Nestlé applies customer intelligence and simulated responses to product concepts, while Walmart uses digital environments to evaluate physical store changes.

Both examples show how organizations can evaluate decisions earlier in a process. For marketing leaders, the opportunity is to identify where better information can reduce the cost, time, or uncertainty of making customer-facing changes.

Measuring AI maturity through operational decisions

Bain’s findings suggest that conventional adoption metrics capture only part of AI maturity.

The number of employees using AI, the number of tools deployed, and the volume of content generated indicate how widely an organization has introduced the technology. Business value also depends on whether those capabilities influence priorities, responsibilities, customer decisions, and resource allocation.

A more useful assessment examines what the organization can accomplish through its AI investments.

Six questions follow from the differences Bain identifies between leaders and laggards:

1. Are AI priorities connected to enterprise objectives? Senior leadership should establish the business outcomes AI is expected to support and coordinate investment across functions. Executive sponsorship matters especially when initiatives require shared resources or changes beyond marketing.
2. Have workflows and responsibilities changed? Executives should examine how teams plan, execute, review, and improve their work. This includes deciding which responsibilities require human judgment and where AI can support execution.
3. Is AI integrated into operational tools? Teams should be able to use AI-generated information within their normal workflows. Integration can reduce manual transfers between systems and help employees coordinate decisions.
4. Can teams conduct more customer experiments and use the results? Greater testing capacity matters when findings influence customer experiences, campaign development, and subsequent decisions. Organizations should examine how quickly experimental results reach the people responsible for acting on them.
5. Do AI insights regularly influence strategy and spending? Forecasting, customer intelligence, and performance analysis should help managers reconsider resource allocation as conditions change.
6. Are financial outcomes measured separately from activity? Executives need to distinguish technology usage, operational improvements, and business results such as revenue growth and sustainable cost savings.

These questions are practical implications drawn from Bain’s research rather than a formally validated assessment framework. They help make investment decisions more specific.

For example, an organization considering additional AI spending can examine whether its main constraint is technology access, data availability, fragmented workflows, or the authority to change existing processes.

The answer should determine where the next investment goes.

If teams already use capable AI tools but struggle to incorporate their outputs into campaign decisions, improving system integration or redesigning responsibilities may create greater value than purchasing additional licenses.

If customer experiments generate useful insights but spending decisions remain disconnected from those findings, management processes may need attention. The priority becomes establishing how evidence reaches decision-makers and when it should trigger a change in execution.

Conversely, organizations lacking the infrastructure or resources to support integrated AI operations may need further investment before extensive workflow redesign becomes practical.

These choices also depend on organizational constraints. Companies facing restricted budgets, uncertain return-on-investment expectations, legal and risk requirements, or fragmented decision-making authority may struggle to undertake large transformations.

Bain’s leader category is defined partly by existing business outcomes: 11% annual revenue growth and seven-point annual market share growth. Companies already achieving stronger results may have management capabilities, financial resources, and organizational flexibility that make extensive AI transformation easier.

The relationship may also work in the other direction. Changes to strategy, workflows, and experimentation could contribute to stronger performance.

Bain’s survey establishes associations between organizational practices and business outcomes. It does not isolate the individual causal effects of those practices or establish that adopting them will produce equivalent returns.

The same distinction applies to Nestlé and Walmart. Their implementations demonstrate substantial uses of AI-supported experimentation and operational adjustment, but their reported improvements do not establish how much of each outcome can be attributed independently to AI or organizational redesign.

Nor does Bain’s research establish that laggards are predominantly smaller or less-resourced companies. The barriers it identifies nevertheless help explain why organizations differ in their ability to implement similar changes.

For executives, AI maturity must therefore be evaluated alongside an organization’s capacity to change. A company with limited resources may need to improve a focused process before expanding into a larger transformation.

A better-funded organization may already have the technology and data it needs, making coordination and decision-making its more pressing constraints.

This distinction also changes how spending should be evaluated. Higher-performing companies tend to commit larger budgets to AI, but expenditure is an input. Its commercial value depends on the capabilities those investments create and how employees use them.

For CMOs deciding where to invest next, the most useful starting point is a specific operational constraint. Identifying where customer information becomes difficult to use, where decisions are delayed, or where experiments fail to influence execution provides a basis for choosing between new technology, system integration, training, and workflow redesign.

The investment can then be evaluated against the operational change it was intended to produce and the financial results that follow.

Final thoughts

AI adoption is becoming less useful as a measure of competitive advantage in marketing. As more organizations gain access to similar tools and capabilities, the differences in business performance increasingly depend on how those technologies are integrated into operations and decision-making.

Bain’s research shows that higher-performing marketing organizations are more likely to centralize AI strategy, redesign workflows, and use customer intelligence to guide experimentation and spending. These practices are associated with stronger financial outcomes, although the findings do not establish that AI transformation alone produces those results.

For CMOs and business leaders, the priority is to determine where AI can improve decisions rather than simply accelerate existing tasks. Faster content production offers limited value if campaign planning, customer analysis, and budget allocation remain disconnected.

That requires investment decisions to account for more than software capabilities. Organizational structure, employee responsibilities, data integration, and management processes all influence whether AI-generated insights translate into measurable business improvements.

As AI adoption continues to spread, executives will need to evaluate success through changes in operational performance and financial outcomes. The distinction between deploying AI and using it effectively will become increasingly important as the technology becomes a standard part of marketing operations.