The CPG AI Readiness Trap
You don’t need perfect data to start creating better decisions
There is a piece of advice that has quietly slowed AI adoption across CPG organizations for years:
“Get your data right.”
It is well-intentioned advice. It has been repeated by consultants, reinforced by technology vendors, and embedded into countless transformation roadmaps. Over time, it has created a widely accepted belief that meaningful AI capability only becomes possible after an organization has completed a large-scale data harmonization effort.
For many organizations, that belief has become an unintended barrier to progress.
The result is that teams spend months—or years—preparing for AI while delaying the very business outcomes they hope AI will help achieve. Forecast accuracy remains inconsistent. Trade spend remains difficult to optimize. Pricing decisions remain reactive. Commercial teams continue to spend valuable time gathering information rather than acting on it.
The organizations making the most progress have adopted a different mindset. Data maturity and AI value creation are not sequential activities. They happen together.
The Data Readiness Assumption
Much of the market still treats data readiness as the finish line.
The prevailing assumption is that once retailer POS, syndicated, shipment, TPM, financial, and supply chain data are connected, governed, and harmonized, insights will naturally emerge. Entire categories of vendors have been built around this premise, positioning data integration as the primary obstacle standing between organizations and better decisions.
A strong data foundation is incredibly valuable. Organizations absolutely need trusted, connected, decision-grade data.
The problem is that data readiness and decision readiness are not the same thing.
Connecting data does not automatically create pricing intelligence. It does not reveal promotional elasticity. It does not optimize assortment. It does not improve forecast accuracy. It does not determine the best commercial action.
Many organizations have invested heavily in modern data platforms only to discover they are still left answering the same questions:
-Which promotions should we run?
-Which prices should we change?
-Which SKUs should we rationalize?
-Where should we invest trade dollars?
-How should we respond to changing retailer behavior?
The data exists. The decisions remain.
That is because data platforms solve a data problem. Commercial AI for CPG solves a decision problem.
The ability to connect retailer POS, syndicated, shipment, inventory, and financial data is fundamentally different from understanding how pricing, promotions, assortment, forecasting, retailer behavior, and commercial execution interact to drive business outcomes.
Organizations need both. The mistake is believing one automatically creates the other.
What Foundation Models Changed
The emergence of foundation models changed expectations in a profound way.
When OpenAI introduced ChatGPT and Anthropic introduced Claude, they demonstrated that useful intelligence no longer needs to be built entirely from scratch. These systems arrived with embedded knowledge, reasoning capabilities, and pattern recognition before ever seeing an organization’s proprietary information. Internal data makes those systems more relevant. It does not make them intelligent.
That distinction matters because many organizations continue to evaluate commercial AI through the lens of traditional analytics projects, where value could only emerge after significant data preparation was complete. Increasingly, that is no longer true.
The best AI systems arrive with meaningful capability already embedded within them. Proprietary data improves context, relevance, and precision, but it is no longer the sole source of value.
Commercial Intelligence Should Not Start at Zero
This same principle increasingly applies to commercial AI.
One of the assumptions many organizations still make is that commercial AI platforms begin with no understanding of pricing, promotions, forecasting, retailer behavior, assortment strategy, or revenue growth management. The expectation is that the system must first learn everything from company-specific data before it can become useful.
In practice, the strongest commercial AI platforms increasingly arrive with embedded commercial intelligence already built into them.
At Insite AI, years of work have gone into codifying commercial knowledge around pricing, promotions, assortment, forecasting, retailer behavior, category dynamics, demand transfer, and revenue growth management. The result is a commercial intelligence layer that arrives with meaningful capability from day one and becomes increasingly precise as proprietary customer data is incorporated.
A useful analogy is an experienced RGM leader, category expert, or external consultant. They do not walk into an organization knowing every customer, every SKU, or every retailer relationship. Yet they are still capable of providing meaningful guidance immediately because they bring frameworks, pattern recognition, and commercial expertise developed through years of experience. As they gain context, their recommendations improve. Your data does not create their expertise, it sharpens it.
Purpose-built commercial AI operates in much the same way. Embedded commercial intelligence provides the starting capability. Proprietary data improves precision, relevance, and specificity over time. Both matter. Neither replaces the other.
The Organizations Moving Fastest
Across the customers and prospects we speak with, the goals are remarkably consistent. Organizations want more accurate forecasts, stronger pricing decisions, more effective promotions, better retailer planning, and faster responses to market volatility.
They want to improve trade profitability. They want greater confidence in pricing decisions. They want to reduce commercial friction between teams and shorten the distance between signal and action. Those outcomes are not achieved through data harmonization alone. They require the combination of trusted data, commercial intelligence, and decision frameworks capable of turning information into action.
The organizations making the most progress are not waiting for perfect data before they begin pursuing those outcomes. They are creating value while continuing to strengthen their data foundation.
Rather than treating data transformation and AI deployment as sequential activities, they are pursuing them in parallel. They are improving forecast accuracy while improving data quality. They are optimizing trade spend while expanding data connectivity. They are accelerating pricing decisions while strengthening governance and harmonization.
This approach creates something many organizations underestimate: momentum.
Business stakeholders begin seeing measurable outcomes. Commercial teams gain confidence. Cross-functional alignment becomes easier. Investment becomes easier to justify because value is visible.
The conversation shifts from preparing for transformation to realizing it.
The Real Question
For years, organizations have asked:
“Is our data ready for AI?”
Increasingly, the better question is:
“What can AI do for me right now?”
At Insite AI, we help organizations answer that question by combining embedded commercial intelligence with each company’s unique data, processes, and business context. The objective is not to bypass data maturity. It is to accelerate time-to-value while continuously strengthening the foundation that supports better commercial decisions.
The organizations pulling ahead are not waiting for perfect conditions. They are improving data maturity and creating business value at the same time.
The future of CPG is not better reporting. It is better decisions, made faster.
