The Operational Backbone Behind Trusted AI in Retail

AI can make retail service feel faster, more personal and more responsive. But for customers, the real test is simple: was the answer accurate, useful and actionable?

That is why agentic retail experiences cannot succeed as a front-end layer alone. When a customer asks where an order is, whether an item is available for pickup or how to return a purchase, the experience depends on something deeper than conversational design. It depends on connected operations. Real-time inventory visibility, order status, fulfillment logic, returns workflows and customer service data all have to work together if AI is going to feel credible.

Pandora’s transformation shows what this looks like in practice. Publicis Sapient helped the brand modernize key parts of its commerce, fulfillment and service foundation, including the migration of order management to IBM Sterling Order Management on Cloud in Australia and New Zealand. That work strengthened the omnichannel capabilities behind the customer experience, reducing manual processes, improving inventory visibility and supporting services such as Click-and-Collect and Store Fulfillment. It also created better tools for post-purchase inquiries and returns.

For operations leaders, that is the more important story. Helpful AI is only possible when the underlying retail architecture is connected enough to support trustworthy answers and consistent execution across channels.

In retail, many of the most common customer questions are operational questions. Where is my order? Can I pick this up in store? Is this item actually available? What happens if I need to return it? Those moments sit at the intersection of service, commerce and fulfillment.

If the systems behind them are fragmented, even a well-designed digital experience will fall short. Customers may get generic answers instead of real updates. Associates may have to switch between tools. Service teams may rely on manual workarounds. Inventory may appear visible in one channel but not another. The result is friction, inconsistent service and lower confidence in digital interactions.

Pandora’s work addressed that issue at the foundation. The order management modernization in Australia and New Zealand focused on consolidating e-commerce fulfillment and improving the flow of operational data needed to support cross-channel service. The new environment enabled real-time inventory visibility, improved ship-to-home services, optimized returns and refunds, and supported Click-and-Collect and Store Fulfillment.

That kind of modernization matters well beyond fulfillment efficiency. It changes what the business can confidently promise customers, what service teams can resolve more quickly and what digital experiences can automate without undermining trust.

Before the modernization, Pandora faced operational friction that many retailers will recognize: multiple manual processes, limited inventory reconciliation and outdated customer service tools. These issues made it harder to support a stronger omnichannel model and created unnecessary complexity in the customer journey.

Publicis Sapient worked with Pandora and IBM to identify those pain points and redesign the order management layer to support a more connected operating model. By moving to IBM Sterling Order Management on Cloud, Pandora gained a stronger foundation for cross-channel fulfillment and post-purchase service.

The value of that shift is practical and immediate.

With stronger inventory visibility, the business can make better fulfillment decisions and support selling across channels with more confidence. With improved ship-to-home and store-based fulfillment capabilities, it can route orders in ways that are more responsive to customer expectations. With optimized returns and refunds, it can reduce friction after purchase rather than treating service as a disconnected afterthought.

And with better tools for handling delivery questions and post-purchase inquiries, customer service becomes more informed, faster and easier to scale.

Pandora’s customer service agent, Clara, is a good example of why operational integration matters. Clara helps automate routine questions such as order status and jewelry care, but its value does not come from sounding conversational alone. Its value comes from being connected to the systems that allow it to provide real answers.

Clara integrates with Service Cloud and Commerce Cloud and uses Pandora’s IBM Sterling environment to provide real-time order updates. That is what turns a common customer question into a credible digital interaction. Instead of offering a scripted response, the experience is grounded in live operational data.

This is an important lesson for retail leaders evaluating AI. Customers do not judge service automation only by tone or speed. They judge it by whether it helps them move forward. If an agent cannot access the order, the inventory position, the fulfillment status or the returns context, the experience may be polished but it will not be trusted.

In that sense, modern order management is not separate from AI strategy. It is part of the enabling architecture that makes AI useful.

The order management work did more than improve back-end operations. It helped support a broader omnichannel model in which commerce, fulfillment and service can operate with more continuity.

In Australia and New Zealand, the new platform supported Click-and-Collect, Store Fulfillment and better handling of returns and inquiries. More broadly, the work created a foundation that could be extended into additional markets over time.

That scalability matters. Retailers rarely struggle because they lack digital ambition. More often, they struggle because their operating model is too fragmented to scale new experiences consistently. Publicis Sapient’s work with Pandora addressed that challenge by connecting commerce modernization, fulfillment improvement and service enablement into a more unified retail foundation.

That broader foundation was also supported by changes in how Pandora worked. Publicis Sapient helped remove silos between marketing, commercial operations and technology while applying cross-functional ways of working and a test-and-learn approach. The result was not just a new platform stack, but a more agile way to launch and improve omnichannel experiences.

Pandora was able to launch pilot programs in under three months, including experiences such as Click-and-Collect, find in-store, store locators, virtual queuing and in-store appointment booking. That speed reflects the value of connected operations as much as connected technology.

For COOs, digital operations leaders and commerce executives, the implication is clear: if you want AI to improve service and support cross-channel selling, start by strengthening the systems that define operational truth.

That means modernizing order management so inventory, fulfillment and returns data can be used in real time. It means reducing manual processes that slow service and introduce inconsistency. It means giving customer service teams better tools for post-purchase interactions. And it means designing omnichannel capabilities such as Click-and-Collect and Store Fulfillment as part of a connected experience, not as isolated features.

Pandora’s example shows that agentic service becomes more credible when it is built on a stronger operational backbone. Publicis Sapient helped create that backbone by modernizing commerce, strengthening omnichannel fulfillment and connecting the systems that support customer-facing interactions.

The result is a retail model where better service is not only more automated, but more trustworthy. And in an era when every brand is exploring AI, that may be the difference that matters most.

Relevant Links

• Reimagining In-Store Magic: Pandora’s Agentforce Journey with Publicis Sapient
• Reimagining In-Store Magic: Pandora’s Agentforce Journey with Publicis Sapient
• Pandora’s APAC commerce modernization
• Luxury and specialty retail has always relied on a kind of intelligence that is hard to scale: the ability to read the moment.
• How Pandora Made Retail Transformation Stick
• Danmark: Fra global retailtransformation til mere menneskelige digitale kundeoplevelser (Europe)
• Cómo llevar la calidez de tienda al canal digital: lecciones para líderes de retail en México (LATAM)
• FAQ (FAQ)
• What to Know About Publicis Sapient’s Work With Pandora: 12 Key Facts for Retail Leaders (LIST)