Build On Buy: The Third Retail Technology Model Emerging In The AI Era

SVP SW Portfolio Management at Toshiba Global Commerce Solutions, Yevgeni drives innovation and delivers cutting-edge retail technologies.

getty​AI is making software easier to create. For retailers, this changes a decades-old technology decision and raises a more important question: What solutions are actually worth building?

Over the years, some of the biggest technology decisions in retail started with the same question: Should we build it, or should we buy it?

Building offers control and the ability to differentiate. On the other hand, buying provides access to proven technology, scale and capabilities retailers did not have to create and maintain themselves.

The rapid rise of AI is changing that equation, which ultimately reduces both time and resources needed.

Software that once required months of development can increasingly be created or modified much faster. User experiences, workflows, reports, integrations and applications are becoming easier to build.

This shift changes the economics of innovation. Rather than choosing between complete ownership and complete dependence on a vendor, retailers can now be far more deliberate about where they invest internal resources and where they leverage established technology. That balance is giving rise to a model: build on buy.

Build on buy means first buying the mission-critical foundation that must perform, scale and continuously evolve, while preserving the freedom to build the experiences, workflows, intelligence and differentiation above it.

AI is changing the economics of software, but it is not changing every layer equally.

Consider a retailer creating a new associate experience. Building the interface is becoming easier. Creating the workflow is also becoming easier. Connecting an API, generating a report or developing an AI assistant can take a fraction of the effort it once did.

But underneath that experience are many challenging issues. Is the inventory accurate? Is the price correct? Is the associate authorized to take an action? What happens if a payment fails halfway through a transaction? What happens if connectivity disappears? Can an action be reversed? Can the retailer understand exactly what happened afterward?

Within that process, scalability also needs to be considered. Can the capability perform consistently across thousands of locations, different devices, store configurations and network conditions?

With the help of AI, this technology helps engineers address these problems but does not remove them entirely.

The cost of creating software is coming down much faster than the cost of operating consequential systems.

That changes where the real value sits.

As visible functionality becomes easier to reproduce, durable value moves deeper into the technology stack.

It moves toward transaction integrity, authoritative business context, security, policy, resilience, integration, observability and recovery.

Eventually, a digital experience meets physical reality. The product must exist. The price must be right. The payment must work. Inventory must reconcile. Devices must operate. Associates may need to intervene. Stores still need to function when a network, cloud service or another dependency has a problem.

These are difficult capabilities because in addition to writing the software, they must work correctly, repeatedly and at scale for many years.

This is why I believe world-class technology platforms are becoming more important as AI makes development easier.

The foundation must perform on an ordinary Tuesday and during the busiest shopping day of the year. It must scale as transaction volumes, locations, devices and experiences grow. It must remain secure and resilient. It must recover from failures and allow continuous change without creating continuous instability. A retailer choosing a foundational platform is making a decision that may remain with the business for years. During that time, infrastructure will change. Security threats will change. Regulations and payment methods will change. AI architectures will change. Entirely new ways of interacting with retailers will emerge.

A world-class platform provider should absorb much of that complexity and continuously invest in performance, scalability, security, compatibility and resilience.

The retailer is buying today’s software along with accumulated engineering experience and continuous evolution.

One interesting outcome is that the easier it becomes to build what sits on top of the platform, the more important the platform underneath becomes.

This leads to a different question for technology leaders.

Instead of asking, “Can we build this?” ask, “Is this where we should invest our capability to build?”

Retailers may want to build their own customer experiences, associate applications, unique workflows, AI agents and proprietary intelligence. These are areas where their brand, operating model, data and ideas can create real differentiation to gain a more competitive advantage instead of recreating template foundational technology underneath those experiences.

This is an important distinction from traditional customization. A modern platform should assume the retailer will build and customize. It should expose capabilities through open interfaces, provide common business semantics and allow retailer applications, partners and AI agents to interact safely with the core.

The goal is to create your own differentiation on top of an industrialized foundation rather than modifying someone else’s application.

That is the real promise of build on buy.

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?