McKinsey's 2026 Food and Beverage Report Assumes a Data Foundation Most CPG Leaders Haven't Built
There is a well-documented pattern in retirement finance where people who spent decades successfully accumulating wealth cannot bring themselves to spend it. The psychological machinery that made them excellent savers - caution, preservation instinct, aversion to drawdown - becomes a liability the moment the environment requires a different mode. Kiplinger's reporting on retirement spending behavior captures it plainly: the skills that built the asset become the obstacle to deploying it.
The consumer packaged goods industry is running the same pattern at an organizational scale, and the two McKinsey pieces published in the same week illustrate the problem with unusual clarity.
What McKinsey's CPG Report Actually Says
McKinsey's 2026 State of Food and Beverage report opens with a diagnosis that has been obvious for three years but is now urgent: the slow erosion of value in consumer goods is accelerating. Margin compression, category fragmentation, retailer consolidation, and private label expansion have reached a point where incremental optimization no longer produces meaningful results.
McKinsey identifies three moves for CPG leaders: reshape portfolios, sharpen value propositions, and fully harness technology and AI. The framing is familiar to anyone who has followed consulting outputs in this space. Portfolio rationalization is code for cutting legacy brands that no longer justify their structural cost. Value proposition sharpening means distinguishing clearly between products that can support premium pricing and those that cannot. Harnessing AI means everything from demand forecasting to dynamic pricing to personalized trade promotion optimization.
The urgency is real. The playbook is legitimate. The problem is what it assumes.
The Prerequisite Published Separately
The same week McKinsey told CPG leaders to harness AI, they published a separate piece that functions as the footnote nobody will read: data management best practices for Advanced Planning Systems deployments.
APS - Advanced Planning Systems - are the optimization engines that run CPG supply chains, demand forecasting, production scheduling, and distribution logistics. Blue Yonder, Kinaxis, SAP IBP. These are not edge cases. They are core operational infrastructure for virtually every major CPG company. And McKinsey's finding from watching these deployments succeed and fail is precise: data management is the "quiet enabler," and getting it right at the start unlocks impact faster than any other variable.
"Quiet enabler" is doing a lot of work in that framing. It is quiet because nobody builds a strategy deck around it. It is quiet because consultants do not get credit for recommending data governance work when clients want transformation narratives. It is quiet because it has no announcement, no ribbon cutting, and no press release.
But when quiet enablers are absent, every loudly announced initiative eventually runs into the same invisible wall.
This is the kind of structural prerequisite that STI's research tracks systematically - where AI-driven outcomes are predicted not by which models you deploy, but by the data foundation you built before the deployment decision. We documented the specific mechanism in the Walmart-Macy's agentic commerce gap: two retailers ran "agentic commerce" pilots in the same year. One generated 4.75x revenue per order. One shut down after six months with conversion rates three times below baseline. The variable was data architecture, not model capability.
The Accumulation Trap in Corporate Strategy
The behavioral science behind the retiree spending problem is well-established. Decades of research in loss aversion and mental accounting show that the psychological weight of a potential loss is roughly twice that of an equivalent gain. This asymmetry serves accumulation beautifully: being more afraid of losing $100 than you are excited about gaining $100 is a useful feature when the goal is to build a portfolio over 40 years.
The transition from accumulation to deployment requires inverting that psychological frame. The person who built wealth through preservation has to start treating drawdown as the intentional strategy, not the failure mode. Most cannot make that switch cleanly without an external forcing function - a financial advisor who reframes the math, or a health event that makes the urgency concrete.
CPG companies have built extraordinary capabilities over the past three decades: brand equity, retail relationships, global supply chain infrastructure, procurement efficiency, and category management sophistication. These capabilities were built by investing in what was known to work. The operational excellence that produces 400 basis points of margin improvement from logistics optimization. The marketing machine that can launch a product at scale in 18 markets simultaneously.
The AI transformation McKinsey is recommending requires deploying against something genuinely different. Portfolio reshaping means deliberately weakening brands that still generate revenue today. Data infrastructure investment means spending on foundations that will not show up in the quarterly P&L for 18 months. AI implementation means building capabilities that initially decrease the utilization of existing human expertise.
Every one of those moves activates the same accumulation psychology that freezes retirees. The organization built for preservation deploys the defense mechanisms that served it during accumulation: extended review cycles, pilot programs that never scale, "AI centers of excellence" that do not touch the core operational systems.
What the AI Agent Conference Story Actually Signals
Adweek is promoting a workshop at Social Media Week where brand marketers can build their first AI agent - a hands-on session where attendees learn to construct agents and leave with tools they can deploy immediately. There is genuine value in hands-on AI literacy for marketing practitioners. The problem is the sequence.
The marketers attending this workshop will return to organizations whose data environments range from clean to actively contradictory. They will deploy agents against customer data that has not been audited, against campaign performance data that uses inconsistent attribution windows, against product catalogs that have not been rationalized in three years.
The result is agents that produce outputs nobody trusts, get quietly shelved after three months, and leave the organization with a residual belief that "AI does not work here."
This is the agent equivalent of the APS failure McKinsey describes. The planning system arrives. The data environment is unprepared. The implementation produces results below expectations. The organization concludes the technology does not work, when the actual conclusion should be that the prerequisite was skipped.
The Compounding Cost of Sequence Errors
Sequence errors in technology deployment compound in both directions. Organizations that build the data foundation before deploying agents produce compounding returns - every agent interaction generates signal that improves the next recommendation. Organizations that deploy agents before building the foundation generate compounding noise - every failed interaction erodes organizational trust in AI, making the next implementation harder to fund and resource.
McKinsey's research on enterprise AI adoption shows fewer than 10% of companies that experiment with AI agents scale them to tangible value. That figure is not a technology capability problem. It is a sequence problem. The 10% that succeed built the unglamorous prerequisites first.
If you are evaluating your organization's readiness to act on McKinsey's CPG recommendations, our analysis tools can surface the infrastructure gaps that transformation reports will not highlight directly.
The Consulting Incentive That Obscures the Gap
There is an uncomfortable structural reason why McKinsey's CPG growth report leads with AI transformation rather than data foundation requirements. Consulting firms generate more revenue from high-visibility transformation engagements than from data governance work. Portfolio reshaping projects are large, complex, and visible. They produce clear artifacts - brand divestitures, acquisition recommendations, go-to-market redesigns - that justify substantial fees and executive attention.
Data management best practices work is smaller, slower, less visible, and harder to cost-justify in a quarterly planning cycle. It is also, by McKinsey's own analysis, the prerequisite that determines whether the transformation work succeeds.
This is not a criticism of the research quality - both McKinsey pieces are analytically sound. It is an observation about how incentive structure shapes what gets emphasized in board presentations versus what gets buried in the appendix. The infrastructure shift in enterprise AI is well-documented. What is less documented is the structural bias toward the visible half of that shift.
CPG leaders who read McKinsey's State of Food and Beverage as a mandate to pursue AI transformation without auditing their data environment first are not making a strategic mistake. They are making a sequence mistake, which is worse - because sequence mistakes are self-validating. The AI initiative fails, not because AI does not work, but because the prerequisite was absent. The organization draws the wrong lesson and spends the next two years recalibrating its investment thesis while competitors who built foundations first continue to compound.
The Non-Obvious Read on McKinsey's Dual Publication
The fact that McKinsey published both pieces in the same week is not a coincidence. Their Operations practice and Consumer practice are telling the same story from different entry points: the CPG industry has reached a moment where incremental improvements within existing capability architectures no longer close the performance gap. What closes it is a genuine replatforming - portfolio rationalization, value proposition clarity, and AI on top of a data foundation that can actually support AI.
The companies that will execute McKinsey's growth playbook successfully are not the ones that move fastest on the visible recommendations. They are the ones that treat the quiet enabler as the actual constraint, invest in it before it feels necessary, and let the transformation layer follow once the foundation is genuinely ready.
That sequence is counterintuitive. It requires spending on things that do not appear in the transformation narrative. It requires tolerating the perception of moving slowly at a moment when speed feels urgent. It requires, in a phrase, the same psychological reframe that retirees need to make: deploying deliberately against a foundation you built, rather than preserving the foundation out of the same instinct that built it.
The CPG companies that solve this will not announce it loudly. The quiet enabler enables quietly. If you are building the case for that investment internally, the data architecture gap between AI-ready and AI-attempting organizations is now quantifiable - and it is wider than most strategy presentations acknowledge.