Few CPG and Retail Firms Are Converting AI Experiments Into Measurable Returns, New Report Finds

The gap between AI experimentation and AI value creation in consumer goods and retail is wider than most boardrooms would like to admit. A new leadership report, launched by The Consumer Goods Forum in collaboration with Boston Consulting Group at the 2026 Global Summit in Vienna, puts hard numbers behind that gap. The title says it plainly: AI in CPG and retail: how winners are pulling ahead.

The data tells a clear story. Seventy-six percent of CPG companies remain in what the report calls the "explorer phase." Only 18% have moved beyond pilots to actively scale AI's impact across their operations. Retailers look somewhat further along, with 45% scaling impact, but even there, 40% have barely begun the journey.

That split matters because it suggests technology access isn't the differentiator. Most firms in both sectors have some exposure to AI tools. What separates leaders from laggards is execution discipline, not novelty.

The ROI Measurement Problem

Here's where the report gets uncomfortable for executives who've already greenlit AI budgets. Best-in-class performers are generating returns of two to five times their investment. That's a meaningful number, the kind that justifies continued capital allocation and signals genuine competitive advantage.

But more than half of organisations surveyed don't measure ROI on their AI investments at all.

That's not a rounding error. It's a structural blind spot. Without measurement, companies cannot distinguish between tools that are earning their keep and those quietly draining budget with no clear payback. It also means boards and investors are, in many cases, approving continued AI spend on faith rather than evidence. For a sector that prides itself on supply chain precision and margin discipline, this is a notable inconsistency.

The financial opportunity on the table is substantial. The report projects the addressable value will expand by a factor of 1.7 as AI capabilities mature from today's assisted tools into more autonomous models capable of independent decision-making across commercial functions. That's not a marginal uplift. It implies the gap between winners and laggards will widen further, not narrow, as the technology matures.

Measurement failures compound that risk. Companies that can't quantify today's returns will struggle to justify the bigger bets required to capture tomorrow's expanded opportunity. Capital tends to flow toward demonstrated value, and in AI's case, demonstrated value requires someone to actually track it.

The report doesn't just diagnose the problem. It offers a strategic framework organised around six considerations the report says belong squarely with the C-suite, not delegated to IT or innovation teams. Three themes stand out from that framework, based on the report's own framing.

First, investment has to align tightly with commercial priorities. AI deployed for its own sake, divorced from revenue growth, margin improvement, or supply chain efficiency, is the kind of spending that shows up in budgets but never in results. Second, operating models need rethinking. Bolting AI tools onto legacy processes rarely produces the step-change gains that justify the investment. Third, leaders need clear guardrails around cost and risk. Scaling AI without governance invites the kind of overruns and compliance exposure that erode any ROI gains on the back end.

The report's central prescription is a shift from isolated point solutions toward what it calls orchestration, embedding AI deeply into commercial operations rather than treating it as a bolt-on capability. That distinction between tool and system is, in practical terms, the difference between the 18% of CPG companies scaling impact and the 76% still exploring.

Supply chain functions are an obvious proving ground for this shift. CPG and retail businesses have spent decades optimising logistics, inventory, and demand forecasting through incremental technology upgrades. AI applied with real orchestration, rather than as a standalone analytics layer, has the potential to compress those optimisation cycles meaningfully. But the report's own data suggests most organisations aren't there yet, and won't get there without addressing the measurement gap first.

Retailers' stronger scaling numbers relative to CPG companies likely reflect proximity to consumer data and faster feedback loops. Retailers see demand signals, pricing responses, and inventory turns in near real time, which creates more immediate incentive and more immediate evidence to justify scaling AI investment. CPG companies, operating further up the value chain with longer lead times between decision and outcome, may simply have less visibility into whether their AI bets are working, which could partly explain why experimentation persists longer before scaling begins.

The Bottom Line

The report's framing, "how winners are pulling ahead," is a warning as much as a description. A 2 to 5x ROI for best-in-class performers, combined with a projected 1.7x expansion in addressable value as AI capability matures, creates a widening performance gap between companies that can prove their AI investments work and those that can't.

For executives in consumer goods and retail, the near-term priority isn't necessarily more AI spending. It's building the measurement infrastructure to know whether existing spending is paying off. Over half the industry currently lacks that visibility. Until that changes, continued investment in AI capability, however promising in theory, remains a bet made largely in the dark. The companies willing to confront that gap now, rather than after the next budget cycle, are the ones most likely to be counted among the winners the report describes.