A brand team might ship a generative content pilot. A shopper marketing team might be building AI-assisted audience segments for a retail media campaign. Somewhere else, a digitally forward team is testing a new commerce or forecasting use case. Each looks like progress, and on its own terms, it is. But too often, a year later, what you have is a handful of good experiments that, unfortunately, never became organizational capabilities.

Adobe worked with Incisiv to survey 5,633 executives across 12 industries and 11 markets, including 521 from consumer goods. The key takeaway for leading marketers in consumer goods: the constraint is rarely the model. It's what happens to a good use case the moment it needs something outside the team that built it.

You don't own the finish line.

61% of consumer goods respondents sit in what the research calls the Frozen Middle: AI is active in individual functions, but the coordination needed to repeat that success across teams hasn't caught up. Every other industry in this study has some version of that gap. Consumer goods has a harder one, because so much of the value sits past the brand's own walls.

Take one retail-media use case. It touches brand and portfolio leaders, ecommerce, shopper and retail media teams, sales, and data and technology internally, then reaches out to media publishers, retail partners, distributors, and marketplaces externally. None of them run the same systems, report to the same P&L, or work on the same clock. A workflow that runs cleanly inside your building stalls the moment it needs a retailer's data-sharing sign-off on a timeline you don't control. That's not a rollout problem. It's a design problem, and most consumer packaged goods organizations haven't designed for it yet

No one is holding the pen.

Only 16% of consumer goods respondents have named a formal AI owner with a real title and budget authority. Everywhere else, it's a committee, a project-by-project assignment, or nobody in particular. Committees are fine for input. They are terrible at the one thing a cross-partner use case actually needs — somebody who can make a binding call when marketing, legal, privacy, security, and a retailer's data team all want something slightly different.

The approval funnel shows the cost. Out of every 100 proposed AI use cases in consumer goods, 15 make it to production. The biggest absolute loss happens immediately at the IT evaluation stage. The sharpest proportional attrition comes later, as ideas move through integration, budget, and production readiness. The research doesn't explain why each use case was removed, but the pattern exposes the cost of discovering feasibility and dependency questions too late. Neither pattern is an argument for a lighter review. Both are arguments for identifying the data, the permissions, and the owner before a proposal spends three months proving itself unworkable.

The handoff habit.

76% of consumer goods respondents still manage business and IT through siloed, transactional, or consultative relationships. Business owns the outcome; IT receives the request or advises on it. What they generally don't do is own the result together.

That model was built for longer planning cycles. AI doesn't wait politely. Questions about data access, integration, governance, and measurement show up simultaneously and need answers together. Advice helps, but shared accountability works differently.

Measuring it after the fact measures noise.

Just 12% of consumer goods respondents measure AI's return comprehensively. Most track what's easy: hours saved, content volume, or audiences built. Useful, but none of these metrics tells you whether a use case moved household penetration, buy rate, or gross profit. And once a workflow is live, pricing moves, promotions run, distribution shifts, media flights end, and separating AI's contribution from all of that becomes close to impossible. The organizations that can make a credible investment decision set the baseline before launch, not after.

What the top 6% actually do.

The study identified 342 Leaders across all 12 industries representing the top 6% of active respondents by operating-model maturity and deployment stage. This isn't a consumer goods cohort, and the research doesn't claim that these behaviors caused scale. But the contrast is instructive. Five behaviors show up much more often among Leaders:

• Real ownership. Leaders are 5.8 times more likely to have a formal AI owner with a title and budget authority than consumer goods organizations, where only 16% have named one.
• Day-one measurement. 44% of Leaders measure ROI comprehensively, against 12% in consumer goods, because they define success before the workflow launches, not after.
• Genuine partnership. 81% of Leaders run business and IT at the collaborative or integrated level, against 24% in consumer packaged goods, with shared OKRs replacing the handoff.
• A learning culture. Fewer than 1% of Leaders show destructive failure patterns, blame and repetition, compared with 28% in consumer goods.
• Empowered champions. 42% of Leaders name finding and backing internal champions as the single most impactful culture move they made, ahead of every other lever measured.

None of these is a technology purchase. They're organizational decisions, and they reinforce each other: an owner with real authority is what makes a shared OKR possible, a shared OKR is what makes day-one measurement a joint responsibility instead of an afterthought, and a champion who shows a peer how a workflow actually behaves moves adoption faster than any all-hands ever will.

Build the organization before the use cases multiply.

Consumer goods is still early enough to get this right.

AI experimentation will expand. More brand teams will use generative AI. More commerce, shopper, media, insights, and operational teams will find applications worth pursuing. The question is whether every new use case will have to solve for ownership, data access, measurement, governance, and partner coordination from scratch.

The organizations that address those questions now have an opportunity to make the next wave of AI adoption very different from the first.

The model will change. The use cases will change. What won’t change are the questions that need to be answered. Who owns the outcome? How will we measure it? Which teams and partners need to be involved from the start? And what needs to be true before this moves beyond an experiment?

That is where AI readiness in consumer goods really starts.

Read the full report to see where consumer goods organizations stand and what the Leaders did differently.

https://business.adobe.com/fragments/resources/cards/thank-you-collections/consumer-goods