Oracle and Deloitte point to a new era for retail: AI, connected processes, and faster results

2026 study highlights the evolution of artificial intelligence in Oracle Retail, the integration of business journeys, and an important shift in priority: less isolated technology, more efficiency and measurable results.

Retail is entering a new phase of technological transformation. After years of investments in digitization, automation, and system modernization, a new question begins to take center stage: how to turn these capabilities into concrete results for the business?

This is one of the main discussions presented in the 2026 Retail Industry Global Outlook, published by Deloitte, in a chapter dedicated to Oracle's vision for the future of the sector.

The material gathers perspectives from Jeff Warren, Oracle's Customer Success Executive for Consumer Industries; Paul Woodward, Global Head of Product, Strategy, and Innovation at Oracle Retail; and Rudy Kulas, leader of the Oracle Retail practice at Deloitte Consulting.

The analysis points to an important shift: retailers are not just looking for more modern systems. They need more connected, intelligent operations that can respond quickly to market changes.

In this scenario, artificial intelligence ceases to be a complementary technology and begins to occupy an increasingly strategic role in executing business processes.

Retailers no longer want isolated systems. They want journeys that work.

Historically, large retail operations have been built on different specialized applications.

Merchandising, planning, purchasing, inventory, pricing, stores, e-commerce, logistics, and finance systems have evolved to meet specific needs.

While these solutions are essential, complexity arises when a single business journey depends on the interaction between various applications, data, and teams.

The challenge is no longer just to ensure that each system works correctly.

It is to ensure that the operation works as a whole.

According to the vision presented by Oracle executives, retailers are increasingly interested in the outcomes of journeys, regardless of how many applications are needed to execute them.

This represents a relevant change in the way investments in technology are evaluated.

A replenishment operation, for example, does not start and end in a single system. It may involve demand forecasting, planning, order generation, suppliers, receiving, distribution, and availability in stores.

The value is not just in automating each step, but in connecting all of them to improve decision-making and reduce manual intervention.

Artificial intelligence: from recommendation to execution

The application of artificial intelligence in retail is not new.

For years, predictive algorithms and machine learning models have been used in demand forecasting, customer segmentation, price optimization, and discount management.

More recently, generative AI has expanded the possibilities for interacting with systems, interpreting information, and producing content.

Now, a third dimension is beginning to gain traction: Agentic AI.

While predictive AI identifies patterns and generative AI produces responses and content, agentic AI can coordinate steps, interact with tools, and execute actions within defined processes.

In the context of retail, this evolution opens up possibilities such as:

• Monitoring stock exceptions and initiating investigation flows.
• Supporting the identification and resolution of supply issues.
• Coordinating activities between planning, purchasing, and suppliers.
• Analyzing scenarios and recommending operational adjustments.
• Automating repetitive tasks that depend on information distributed across systems.

These possibilities depend, of course, on the capabilities available, the integrations, and the control mechanisms adopted in each operation.

The most relevant transformation is not just in the ability to produce intelligent responses.

It is in the possibility of incorporating intelligence into the operational journeys themselves.

Oracle plans to expand AI in its retail portfolio

One of the highlights of the material is the strategy presented by Oracle to incorporate artificial intelligence into its portfolio.

According to Paul Woodward, the company has set a goal to make approximately 50 AI-powered features available in its retail products by 2026, many of which will be available at the time of publication.

The strategic direction combines different approaches to artificial intelligence.

Predictive models remain important for refining forecasts and planning decisions.

Generative resources can make interaction with applications more intuitive and reduce complexity for users.

Meanwhile, AI agents represent an evolution in the automation of complete processes.

For the retailer, the main change lies in the expectation that these capabilities will be incorporated into the solutions used daily, rather than relying solely on parallel innovation initiatives.

This also elevates the importance of preparing the processes and data that will feed these resources.

Return on investment needs to come faster

Another relevant point presented in the study is the increasing pressure for results in artificial intelligence projects.

Jeff Warren recounts the example of a CIO who set different expectations for traditional investments and AI initiatives.

While a merchandising or planning project might have a return horizon of approximately two years, artificial intelligence initiatives should demonstrate value in just three months.

The example illustrates a shift in mindset, although it does not represent a universal rule for all projects.

The adoption of AI is increasingly being evaluated by its ability to generate concrete results in shorter periods.

This means that indicators such as reduction of manual activities, productivity, forecast accuracy, product availability, and operational efficiency become even more important in evaluating initiatives.

Technology needs to be measured not just by what it can do but by the results it actually delivers.

The biggest mistake may be starting with technology

The popularization of artificial intelligence also creates a risk: starting projects based on the possibilities of technology without clearly establishing the problem that needs to be solved.

The study advocates a business-driven approach.

Instead of starting by asking where to apply AI, the recommended path is to identify a journey with real improvement opportunities and determine how different technological capabilities can contribute to the outcome.

In retail, this means starting from concrete questions:

• How to reduce stockouts without excessively increasing inventory?
• How to improve the productivity of merchandising teams?
• How to decrease the time needed to react to demand changes?
• How to simplify relationships with suppliers?
• How to improve the quality of forecasts?
• How to reduce the number of manual activities between systems?

From these challenges, it becomes possible to assess where predictive models, generative resources, or AI agents truly add value.

This approach also helps avoid investments in initiatives that generate interesting demonstrations but struggle to produce consistent gains in production.

The new value architecture in retail

The transformation proposed by executives can be interpreted as an evolution in the way to structure operations.

DiagramaClique para ampliar

The model reinforces an important idea: artificial intelligence does not replace the need for well-structured processes, reliable data, and consistent integrations.

On the contrary, these capabilities become even more relevant.

The greater the autonomy expected from an AI solution, the greater the trust needed in the information and rules that underpin its decisions and actions.

Suppliers also enter the transformation

Collaboration with suppliers is another highlight of the document.

In large networks, merchandising teams need to manage thousands of SKUs, multiple suppliers, and enormous volumes of information.

Order management, receiving, documentation, payments, and deliveries often involve manual activities and processes distributed across different applications.

The report mentions an Oracle initiative aimed at improving collaboration between retailers and suppliers, allowing them to participate more directly in managing information related to documents, receipts, payments, and shipments.

According to executives, early results indicated a reduction in the operational burden on merchandising teams.

The strategic implication is relevant.

Automating does not just mean executing tasks faster. It means freeing professionals for higher-value activities, such as negotiation, analysis, and decision-making.

Supply chains need to react, not just plan

Changes in trade policies, tariffs, transportation costs, and supply conditions continue to pressure the operations of large retailers.

The material from Deloitte and Oracle highlights the difficulty companies face in reacting quickly to these changes using traditional processes.

Even when information is available, analyzing alternatives and executing changes can require significant manual efforts.

A smarter supply chain needs to be able to identify impacts, evaluate scenarios, and support mitigation decisions more quickly.

This involves connecting information from suppliers, purchasing, inventory, planning, and distribution.

Artificial intelligence can enhance this capability, but it depends on an operational structure that allows transforming analyses into actions.

Multicloud and integration gain importance

The study also addresses the role of multicloud environments.

In Rudy Kulas's assessment, large companies need the flexibility to operate applications and data across different technological environments, avoiding excessive dependence on a single platform.

For retail, this topic gains relevance because operations often involve corporate systems, digital platforms, data services, and specialized solutions from different suppliers.

The goal should not just be to connect infrastructures.

It is to allow data and processes to circulate securely and efficiently between applications, regardless of where they are hosted.

This flexibility can be especially important as AI resources need to access information distributed throughout the company.

The future of retail will be defined by execution capability

The perspectives presented by Deloitte and Oracle point to a scenario where artificial intelligence will be increasingly perceived as less of an isolated tool.

Its presence should grow within the very processes of merchandising, planning, replenishment, and supplier relationships.

However, technological adoption alone does not guarantee competitive advantage.

The differential will lie in the ability to identify relevant problems, connect data, integrate systems, and transform new capabilities into measurable operational improvements.

For large retailers, this represents an important evolution in the digital transformation agenda.

It is not enough to modernize applications individually.

It is necessary to build journeys that work end-to-end, with less friction, better decisions, and continuous adaptability.

The next phase of retail transformation will not be defined solely by who uses more artificial intelligence.

It will be defined by who can transform intelligence into execution — and execution into results.