Gurhan Kok founded invent.ai in 2013 to create advanced inventory planning solutions for retail.
getty
Every retail season starts with a plan. Months before customers begin shopping, teams are forecasting demand, building assortments, planning inventory and setting financial targets. They’re deciding what products to buy, how much inventory to carry, where it should go and how it should be priced.
By the time back-to-school shopping begins, thousands of choices have already shaped what customers will see on shelves and online.
Back-to-school is one of the moments when those plans are tested. Retailers are trying to anticipate what customers will want, where they will want it and how demand might evolve, often before they have a complete view of what’s happening in the market.
Every year, as we saw this year, the season reveals the same challenge: Retailers are making decisions in a market that moves faster than traditional planning processes can support.
Customer preferences shift, trends move quickly across channels and regional demand varies. Supply chain disruptions create new constraints. A product that looks balanced in a plan months in advance may not be available in the right place when customers are ready to buy.
The Gap Between Prediction And Action
Forecasts can help retailers understand what is changing, but knowing what’s happening is only the first step. The real challenge is determining how to respond.
Many retailers have invested in better data, advanced analytics and forecasting models that help teams create more informed plans. But I’ve found that prediction only addresses part of the challenge.
In a season like the back-to-school season, a change in demand can trigger a series of actions across merchandising, inventory, allocation, replenishment, pricing and financial planning. Should inventory move to another location? Should replenishment accelerate? Should an assortment change? Should pricing or promotions be adjusted?
These choices are connected. A shift in one area can influence outcomes elsewhere, making it difficult for teams to understand the trade-offs and determine the right path forward quickly.
The retailers that want to succeed must do more than plan at the beginning of the season. They need to adjust when reality changes.
Customer Expectations Are Raising The Bar
The pressure to improve how retailers respond isn’t only coming from inside retail organizations. It’s also coming directly from customers.
During back-to-school shopping, availability matters. Families are often shopping within a specific timeframe, whether they are preparing for a new school year, replacing essentials or looking for specific products. When they can’t find what they need, they rarely wait. They look elsewhere.
Recent research found that nearly one in three U.K. shoppers experience stock gaps when shopping for fashion in stores, and many also encounter products being unavailable online. Shoppers may turn to competitors, marketplaces or other channels when availability doesn’t meet expectations.
Customers don’t see everything happening behind the scenes. They only experience the outcome: The product was either available when they wanted it, or it wasn’t. Availability is no longer just an operational measure. It’s part of the customer experience.
Retail Needs To Move Beyond Prediction
Back-to-school season highlights why retailers need to think differently about AI. Better predictions are available, but they only answer part of the question. Retail teams also need to determine what action to take when demand changes.
Consider a seemingly simple challenge: Different school districts can have different rules about what students can bring to school. Those requirements can affect which products are relevant to customers in a particular market.
An agentic AI workflow could turn that external information into a planning action. An AI agent could collect school district requirements and store-level information, classify markets based on restrictions and connect those requirements to relevant product attributes. It could then scan prior-year sales by attribute and store to understand how those rules affected demand.
Before the next season, the agent could automatically scan for updated district rules, identify what has changed and evaluate the implications for the upcoming assortment and inventory plan, recommending changes to allocation strategy.
The workflow can continue across the retail cycle: pre-season planning, pre-season execution, in-season execution and postseason analysis. Information gathered after the season can become an input into the next planning cycle, creating a continuous learning loop instead of disconnected planning exercises.
This is where agentic AI differs from a standalone forecasting tool. The agent isn’t simply producing another insight for a planner to interpret. It can access the information needed to understand the problem, connect that information to relevant retail decisions and coordinate actions across the planning process.
The architecture matters. When AI has access across modules and shared information sets, a signal discovered by one skill can inform decisions elsewhere. A school district restriction, for example, can become an input into assortment planning and allocation rather than remaining isolated as external research.
Retailers can begin building connected decision-making systems where skills, information and actions work together across functions and organizations.
Technology Should Support Human Expertise
The future of retail AI isn’t about replacing the people who understand the business best. Merchants, planners and retail leaders bring experience, customer knowledge and business judgment to every decision that business leaders need to recognize and champion. Technology can help those teams work through complexity, evaluate options and understand trade-offs while keeping human judgment at the center.
Retail organizations shouldn’t have to choose between human expertise and AI capabilities. They must combine both, using technology to support better decisions while keeping people at the center of the process.
The Future Of Retail Is Built On Adaptability
Back-to-school season is a reminder that retail will always involve uncertainty. No forecast can predict every shift in customer behavior, emerging trend or challenge that appears during a season. The goal isn’t to create a plan that never changes. It is to build the ability to adapt when it does.
As retail becomes more dynamic, the retailers that want to succeed need to be able to recognize change, evaluate their options and respond while there is still time to act. The next generation of retail AI won’t be defined only by how accurately it predicts what happens next. It will be defined by how effectively it helps retailers decide what to do when things change.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
