TransUnion today published survey research finding that 89% of senior US marketing leaders expect to increase spending on AI-enabled marketing over the next 12 to 24 months, while only 36% rate their organisation's data foundations and internal processes as highly ready to support it. The credit bureau and identity company calls the gap the AI Confidence-Readiness Paradox.

The research, released from Chicago on August 5, 2026, was commissioned by TransUnion (NYSE: TRU) from UTA Advisory, the brand advisory division of United Talent Agency. It covers 100 senior marketing and technology leaders at director level or above, all working at US enterprise brands with annual marketing budgets of at least 50 million dollars. Respondents were required to be using AI in current or planned core marketing workflows, and the sample excluded executives whose AI use was confined to basic productivity or creative tasks.

That screening matters when reading the numbers. These are not marketers deciding whether to adopt AI. They are marketers who have already adopted it, at companies with the budgets to do so properly, and a majority still cannot connect the technology to a business outcome they can defend.

Adoption is broad, scaling is not

Among the executives surveyed, nearly half indicated that AI now spans multiple core marketing functions. Only 13% said their organisations had fully scaled AI across the enterprise.

The distribution of use cases explains why. Productivity applications reached 92% adoption. Basic content and creative development followed at 89%, and creative optimisation at 83%. Beyond that executional layer the numbers fall away sharply: more than half of teams applied AI to data quality, identity and enrichment, 45% to marketing measurement, and 33% to customer or market insights.

Reported outcomes track the same shape. Three quarters of respondents, 75%, said AI reduced manual effort or time spent on marketing tasks. Beyond that, the returns thin out. Enabling activities that were not previously possible, such as new segments or new tests, was cited by 45%. Reducing media waste or improving the efficiency of media spend reached 44%. Avoiding or reducing headcount additions came in at 37%. Improving conversion or response rates from existing marketing activities was reported by 35%, and reaching a larger share of the target audience at similar spend levels by 30%.

The pattern is a technology delivering time savings reliably and revenue effects intermittently. According to TransUnion, that concentration in tactical use cases leaves the higher-value decisioning and performance applications largely untouched.

Confidence runs ahead of the foundations

Sixty-four percent of respondents said they were confident of meeting their AI-enabled marketing goals. Fifty-three percent said they had seen meaningful ROI from AI-driven marketing initiatives. Eighty-nine percent expected budgets to rise regardless.

Asked to assess readiness to reliably scale AI across marketing functions, meaning consistent and repeatable use rather than one-off deployment, respondents rated three dimensions. On people, covering the skills, talent and roles required to design, operate and act on AI-enabled marketing, 42% answered very high or high, 41% moderate and 17% low or not at all. On process, covering defined ways of working, governance and decision-making frameworks, 36% answered very high or high, 45% moderate and 19% low or not at all. On data, meaning identity, signals and outcomes that are high quality, accessible and well integrated across systems, the split was identical to process: 36%, 45% and 19%.

With a sample of 100, each percentage point corresponds to roughly one respondent, which places these figures closer to a directional reading of the enterprise segment than a precisely bounded population estimate.

"Marketers are increasingly confident in AI's ability to drive business results, but many are still working to build the foundations needed to scale it effectively," said Matt Spiegel, EVP, TruAudience Growth Strategy at TransUnion. "However, AI isn't a shortcut around data challenges. It's a force multiplier. Organizations that build strong foundations of trusted data, identity and measurement will be best positioned to close the gap between AI ambition and AI outcomes."

The data underneath

Fragmentation surfaces as the most frequently named obstacle. Sixty-seven percent reported siloed or fragmented data systems. Forty-two percent cited incomplete or missing data. Thirty-one percent struggled with data latency or timeliness.

A note on how those figures are framed. The whitepaper introduces them within a section describing respondents who rated their AI readiness as moderate to low, then presents the three percentages under a broader heading covering data used in AI-enabled marketing generally. The accompanying press release lists the same three items as top concerns without the readiness qualifier. The base for the 67%, 42% and 31% figures is therefore ambiguous between the full sample of 100 and the moderate-to-low readiness subset.

What is unambiguous is the direction. AI systems inherit the quality of their inputs, and according to TransUnion, outputs built on fragmented or unreliable data are likely to be equally flawed, with the effect concentrated in measurement and optimisation rather than in content generation.

That finding sits alongside a body of research PPC Land has tracked since late 2025. Publicis Sapient's Guide to Next report, published in November 2025, argued that AI projects fail because of data discipline rather than model quality, identifying a confidence-capability gap across enterprises that had left most organisations in pilot mode. In July 2026, DigiCert research covering 1,001 respondents found nearly half of firms unable to fully trace AI decisions back to source data.

Measurement stays anchored to cost

Sixty-five percent of marketing leaders said they measure the impact of AI-enabled marketing primarily through estimated cost or time savings, such as hours saved or reduced manual effort. Marketing mix modelling was used by 42%. Formal lift tests or experiments, including A/B tests and randomised holdouts, reached 35%. Multi-touch attribution or other advanced attribution models stood at 31%, the same share as those relying only on platform or vendor reporting such as dashboards from media or technology platforms.

At the tail of the distribution, 22% said they rely primarily on stakeholder perception or anecdotal feedback, and 14% do not currently measure the impact of AI-enabled marketing at all.

Two structural limits compound the problem. Seventy percent of marketing leaders said cross-channel blind spots prevent them from accurately proving full-funnel impact. Sixty-nine percent said data blind spots inside walled gardens limit their ability to evaluate AI effectiveness. Fewer than half, 48%, said they have enough visibility into platform-level AI to make optimisation decisions with confidence.

The consequence TransUnion describes is a closed loop: teams deploy AI to improve execution, measure it through cost savings, then make further investment decisions without a complete view of performance, which keeps AI tied to incremental gains.

Walled garden reporting limits are not a new complaint. A TransUnion and EMARKETER survey of 196 US marketing professionals published on October 21, 2025 found 54.1% of marketers reporting no year-over-year improvement in measurement confidence and 14.3% reporting a decline, with 40.8% citing walled garden reporting limitations. The IABState of Data 2026 report, announced February 7, 2026, found that up to 75% of marketers considered attribution, incrementality tests and marketing mix models to underperform on rigour, timeliness, trust and efficiency, and estimated roughly 32 billion dollars in unrealised value across media investment and productivity.

Governance lags deployment

Fifty-five percent of respondents reported formal AI approval processes in place. Forty-six percent had documented AI usage guidelines. Thirty-three percent had established response plans for failures, errors or data risks.

Among leaders who rated their organisations' overall AI readiness as low to moderate, 52% cited limited coordination across marketing, data and technology teams. Forty-five percent lacked clear processes for evaluating AI use cases, and the same share said AI usage remained inconsistent or ad hoc.

The skills picture inside that subset is similar. Sixty-nine percent reported limited AI knowledge across marketing teams, 57% cited a lack of in-house expertise and 46% pointed to insufficient training.

Ownership is being negotiated between internal teams and outside vendors. Asked where they expect to rely on external partners over the next 12 to 24 months, 56% named technical implementation and integration. Strategy and use-case design followed at 41%, measurement and attribution at 39%, data and identity foundations at 35%, change management and training at 30%, and governance and risk at 23%. Budget limitations, at 30%, and data privacy and security concerns, at 23%, were the barriers most often named to working with external partners at all.

According to the whitepaper, that mix points toward a hybrid arrangement in which internal teams set priorities, manage risk and hold oversight while external partners handle execution and specialist work.

Leaders and laggards

The report closes with a two-column division. AI leaders embed the technology across planning, activation, measurement and optimisation, run unified data and governance frameworks, connect AI initiatives to business outcomes through clear measurement, and evolve operating models to support continuous adoption. AI laggards keep AI concentrated in isolated use cases such as content generation and campaign execution, contend with fragmented data and inconsistent governance, struggle to prove impact beyond efficiency, and remain stuck in experimentation and pilots.

"What this research makes clear is that AI success is no longer defined by access to the technology itself," said Michael Burke, principal at UTA Advisory. "The real differentiator is whether organizations can connect their data, measure outcomes and operationalize AI at scale."

The whitepaper cites two external reference points for the same argument. Huble's AI Data Readiness Report from April 2025 found that under 10% of executives considered their organisations fully AI-ready. Gartner, in February 2025, predicted that through 2026, 60% of AI projects would be abandoned because organisations lacked AI-ready data.

Why this matters for marketers

The commercial position behind the research is explicit. TransUnion sells identity resolution and cross-channel measurement, and the report's diagnosis, that walled garden and cross-channel blind spots are the binding constraint on proving AI's value, describes the market its TruAudience unit serves. The whitepaper ends with an invitation to build the data, measurement and operating model needed to scale AI with TransUnion. That framing does not invalidate the survey figures, but it does shape which problems the study foregrounds.

TransUnion's own positioning has moved in that direction over the past year. On May 20, 2026, the company and Googleannounced an integration bringing YouTube into TransUnion's multi-touch attribution solution, making TransUnion the sole MTA provider for marketers on the platform. More recently the company joined Universal Ads as one of its first Audience partners, alongside Klaviyo and LiveRamp. Each of those moves addresses a piece of the blind-spot problem the new survey quantifies.

For media buyers, the operational reading is narrower than the headline. Budgets are rising on the strength of a 64% confidence figure and a 53% ROI figure, which means a substantial share of incremental AI spend in enterprise marketing is being committed by teams that have not yet demonstrated a return on the spend already made. With 65% of organisations evaluating AI through cost and time savings, the metric most likely to be presented to finance in the next budget cycle is one that says nothing about revenue.

The measurement methods that would say something about revenue remain minority practice. Marketing mix modelling at 42%, experiments at 35% and multi-touch attribution at 31% describe an industry where fewer than half of enterprise marketing organisations run any causal method against their AI investments. A separate reading of the same tension appeared in Incubeta research covered in May 2026, which found 70.4% of marketing leaders confident their budgets were deployed effectively while 41.6% conceded waste.

There is also a consistency problem across the industry's own research base. MiQ's global survey of 3,169 marketers, released in November 2025, found 72% planning to apply AI in more ways while only 45% felt confident applying it. WARC and TikTok research published on July 14, 2026 recorded 88% reporting higher creative volume from generative AI against 45% reporting higher quality. TransUnion's 64% confidence figure sits above MiQ's 45% and below the vendor-commissioned effectiveness claims circulating elsewhere. Sample sizes, screening criteria and commissioning relationships differ in every case, which is precisely why the composition of TransUnion's panel, 100 executives at brands spending 50 million dollars or more a year, defines the boundary of what these particular numbers can support.

Timeline

Summary

Who: TransUnion (NYSE: TRU), the identity and information company, commissioned the research from UTA Advisory, the brand advisory division of United Talent Agency. The named voices are Matt Spiegel, EVP of TruAudience Growth Strategy at TransUnion, and Michael Burke, principal at UTA Advisory. Respondents were 100 senior marketing and technology leaders at director level or above.

What: A survey and whitepaper describing what TransUnion calls the AI Confidence-Readiness Paradox. Key figures include 89% expecting AI-enabled marketing investment to rise over the next 12 to 24 months, 64% confident of hitting AI goals, 53% reporting meaningful ROI, 42% rating people readiness as high, 36% rating data and process readiness as high, 13% having fully scaled AI, 65% measuring AI through time and cost savings, 70% citing cross-channel blind spots and 69% citing walled garden blind spots.

When: Released on August 5, 2026 from Chicago, with the whitepaper carrying a 2026 copyright and document reference US-4500593-260717-5. The investment and partner-reliance questions cover a forward window of 12 to 24 months.

Where: The sample is confined to US enterprise brands with annual marketing budgets of 50 million dollars or more, spanning retail, consumer goods, technology, telecom, financial services, automotive, travel and hospitality, media and entertainment, healthcare and pharma, and fast food, with B2B and professional services capped at 5% of the pool.

Why: The research argues that the constraint on AI performance in marketing has shifted from access to technology toward organisational readiness, specifically connected data, causal measurement and clear ownership. For advertisers, the practical significance is that rising AI budgets are being approved against efficiency metrics rather than revenue evidence, while the measurement methods capable of producing that evidence remain in use at fewer than half of the organisations surveyed.