Composable Commerce Migration: How to Do It Right? Tips from Virto Commerce’s CTO

Desperate times call for decisive measures. For enterprises, that often means making the leap to composable commerce migration. To explore this in depth, we sat down with Oleg Zhuk, Chief Technology Officer at Virto Commerce, whose 15+ years of expertise shaped the insights behind this article. 

If you’re a digital transformation manager, enterprise architect, ecommerce director, or leading a replatforming project, you’ll find here the key signals that it’s time to move to a composable solution, along with eight essential principles for a successful migration (plus a real-life example from our team).

This article is part of a deep dive into commerce composability (check out the first part). We’ll explore the differences between composable, headless, and legacy systems, and see what makes a platform composable.

Let’s be honest, the times of building an enterprise-grade ecommerce presence from zero have passed. You will most likely not find an enterprise that sells or procures without an online presence. Today, the main dilemma in the commerce world is not whether to establish online channels, but rather whether the current ecommerce platform or system aligns with the company’s vision in the long run. In other words, to replatform, or not to replatform, that is the question (excuse our Shakespeare reference).

When the current ecommerce platform can no longer meet business needs or scale effectively, companies start looking toward more advanced ecommerce solutions. And today, there’s no architecture more advanced, flexible, scalable, and agile than composable.

Before we continue exploring the topic of composable commerce migration, it’s worth mentioning that the words migration and replatforming bear slightly different meanings, and we talk more in detail about them in our article on replatforming. Unlike an ecommerce migration, which primarily transfers data and existing functionality, replatforming means a shift in the technology stack, ensuring businesses can adapt to market demands and customer expectations. However, for the purpose of this article, we’ll use the terms migration and replatforming synonymously when describing the process of moving from legacy systems to a composable commerce architecture.

Composable commerce architecture is a modern approach to building digital commerce solutions where each part of the system is modular (i.e., composable). It’s a relatively new ecommerce term that was first introduced in the 2020s and is considered the most advanced.

At Virto Commerce, we go further with the concept of composability and differentiate platforms based on their platform and deployment composability, with the latter often being overlooked by other solutions on the market. You can learn more about composable commerce, platform and deployment composability, in our dedicated article.

To put it briefly, there is a key signal that a company needs to consider moving to a composable solution from its legacy system, and there are two accompanying technical reasons for composable commerce migration.

Often, enterprises set ambitious growth targets—sometimes aiming to multiply their business several times over within a single year. Trying to match ambition with reality, companies frequently discover that their current architecture, whether legacy or homegrown, simply cannot scale to meet new demands. Even adding headcount and resources doesn’t fix it: the existing system, in its current configuration, fails to satisfy business requirements. That failure is the clearest signal it’s time to replatform.

The core signal for replatforming to composable commerce is straightforward: you want to grow, and you can’t do it with the current platform. Onboarding and training take too long, the current platform is overloaded and slow, processes are outdated, and creating a single new order can take hours. Those high operational costs are often the primary driver for change, and this holds true for both B2B and B2C businesses.

Another frequent technical trigger to move to a composable architecture is the loss of institutional knowledge. Many legacy systems were developed long ago, and the people who built or deeply understand them are no longer part of the company. Technology has advanced, but the external experts who can safely customize, adapt, or maintain legacy systems are rare and expensive when they do exist.

There’s also another technical and cultural reason: engineering teams (fairly) lobbying for modern technology stacks. When asked whether the current legacy platform can support the new roadmap, tech teams will often point out that delivering those plans would require rewriting a large portion of the codebase and dedicating significant development resources. Over the long term, switching to a modern stack often becomes the most cost-effective option. The era of bespoke, homegrown systems is largely behind us; today, it’s usually simpler and more economical to adopt a proven platform that already solves core problems and allows for smooth scaling.

Modern composable solutions allow enterprises to pick the needed modules and out-of-the-box functionalities and concentrate on building only the custom business features required for unique business scenarios. Companies don’t have to worry about the trivial things—the platform vendor will have already designed how the system should scale, be operated, and be maintained. That reduces cost and potential risk. For example, a composable solution like Virto’s Commerce Innovation Platform can offer not only a scalable PaaS foundation but also ready-made tools and a complete development lifecycle (hosting, DevOps, and more), leaving you to customize only what truly differentiates your business.

When replatforming to a composable solution, there are two primary approaches: the so-called big bang approach and the step-by-step (or iterative) approach.

That’s the theory. But in practice, a full shutdown of the old system and simultaneous adoption of the new one almost never happens. A big bang migration is extremely rare because it often leads to business disruption, loss of revenue, and the need to rebuild your entire ecommerce presence from scratch. In today’s world, very few enterprises exist without some form of online presence, which makes this “on/off switch” type of migration highly unrealistic.

At Virto Commerce, we firmly believe that migration should be done iteratively. Modern businesses are rarely built around a single product; instead, they operate within a broader ecosystem—multiple teams, multiple products, different geographies, varied business goals and KPIs, and more. Due to this complexity, the move to a composable architecture should be seamless and executed in small, manageable steps.

1️⃣ The first risk: migrating products only to end up with the exact same result.

It’s possible to spend enormous amounts of time—a year and a half, even two years—and still realize that the product you built isn’t the one you actually needed. Why? Because you didn’t release it in parts, test it in production, and iterate along the way. Instead of following an agile, step-by-step approach, the team chased an “ideal” end state. But when the final system was delivered, it didn’t align with the real requirements. The key is to bring every piece of functionality to production readiness and test it, rather than waiting until the end for a so-called perfect system.

2️⃣ The second risk: designing an architecture that doesn’t align with modern architectural principles.  

This means the new system may lack integrability, fail to provide control over data dependencies and data flows, or worse—create cyclical dependencies that are difficult to manage. Without a forward-looking architectural foundation, scalability and flexibility quickly become major roadblocks.

3️⃣ The third risk: a gap in knowledge and competencies.

Many companies underestimate the expertise required to successfully migrate to a composable solution. They dive into the process with the team they have, assuming existing skills will be enough. But without the right competencies in place, the rollout drags on, mistakes compound, and key architectural decisions get delayed or made incorrectly.

For example, teams sometimes attempt to push forward without bringing in external experts, hoping to “learn on the go.” Instead of accelerating the project, this often results in fragile solutions and an incomplete framework that needs to be reworked later. Before starting the process of migration, it’s important to analyze what kind of team and expertise you actually need. Identify which competencies you already have and which you’re missing.

Now, let’s move on to the key principles of a successful migration from a legacy system to a composable one, addressing the risks above, as outlined by Oleg Zhuk, CTO at Virto Commerce, who brings over 15 years of expertise in this field.

The first step is to take the time to fully grasp how your existing legacy system works. This means documenting as much as possible through workshops, recordings, and shadowing employees at their workstations. The goal is to capture the current processes, architecture, systems, and areas of responsibility. This stage should be time-efficient, thorough, but not dragged out endlessly.

As mentioned earlier, migration must be gradual, seamless, and iterative. A composable transition works best when you move in small, manageable steps, continuously improving and learning as you go.

As Oleg Zhuk, CTO of Virto Commerce, puts it: 

A common mistake is to attempt to “make legacy systems composable” by rebuilding the old system piece by piece in the new environment. Instead, focus on how the legacy system and the new composable platform will coexist for a time. The goal is to ensure business continuity (profits and processes must not suffer) while employees learn to work across both systems. That means prioritizing functionality that is missing from the old platform and is the very reason for replatforming in the first place. It could be a better interface, streamlined data flows, or a reimagined order workflow. Priorities will depend on your business and its unique needs. 

Since both systems (legacy and composable) will run in parallel, you need early adopters inside the company who are willing to experiment and test the new platform. If something doesn’t work in the new system, the fallback option is still the old one, but without those early adopters providing feedback, you won’t know what needs fixing before full replatforming. Both employees and end-users should be part of testing, and they can be incentivized through bonuses, coupons, or other perks. For example, expose only 15% of users to the new system first—an iterative rollout approach. That said, every company and system is unique, so there’s no single blueprint. What we outline here are patterns that work for most enterprises.

Ideally, your first MVP should go live within 3–4 months with working scenarios that can be fully tested end-to-end. After that, additional functionality can be released in sprints. Eventually, all users are migrated to the new system, but it’s critical to keep the old one available as a backup during the transition.

Your new system is meant to last for many years, so from day one, design integrations and architectures properly. For example, in our Atomic Architecture™ approach, all data flows and integrations are managed through a dedicated middleware layer. Processes are documented, and point-to-point connections are controlled and minimized wherever possible. The system can be de-composed into Packaged Business Capabilities (PBCs) and scaled independently, ensuring long-term agility and resilience.

Another misconception is that tackling fewer elements will simplify the process. In reality, if you choose to migrate a component, focus on fully transferring one specific functionality (usually covering 1–2 roles) before moving on to the next. The migration plan should be iterative—avoid the temptation to move everything at once. Think in terms of integrability and step-by-step changes. Yes, you should always keep the big-picture transformation in mind, but just like solving a Rubik’s Cube, you’ll continuously make adjustments and improvements to both the legacy system (while it’s still live) and the new composable system. Migration is not a fixed, one-time process.

The market evolves every six months, and there is both a technology gap and a knowledge gap. As your product evolves, so must your team. But relying solely on your internal team (especially if they’ve worked with the legacy system for years) and expecting them to manage both migration and self-learning simultaneously is ineffective. That’s why it makes sense to bring in external system integration experts. They can guide the process, build a roadmap, and accelerate delivery, while your internal team learns alongside them and gradually takes over the responsibilities. Human capital—your team and their skills—is just as important as the technology itself.

Avoid wasting time and resources on maintaining the platform infrastructure when ready-made solutions already exist. For example, Virto Cloud provides out-of-the-box services for hosting and operations, eliminating unnecessary overhead. Many modern systems come with pre-built capabilities, and it’s smart to take advantage of them rather than building and maintaining everything yourself.

Let’s walk through a real example of composable commerce migration following all the principles we’ve just discussed: our work with the U.S.-based company that designs, produces, and distributes custom apparel. Before migration, they relied on a homegrown system, but over time, it had become outdated and unable to meet their evolving business needs.

The very first thing Virto Commerce did was organize a dedicated migration workshop for the client’s team.

• Representatives from different departments came together to explain how their current system worked.
• Team members with various roles shared perspectives: the technical team walked us through the architecture of their system, while the business team described their vision of what the new platform should deliver.

This gave us a 360-degree understanding of their current processes and pain points. 

Next, we conducted a thorough analysis of their legacy system:

• We mapped out the structure of the system, identified its entities, and tracked the relationships between them. 
• We documented every user flow, pinpointed issues, and gathered the client’s desired outcomes.

This stage was about making the invisible visible and fully reconstructing their system so we could understand exactly what needed to evolve. 

Based on the client’s requirements, Virto Commerce then developed and presented a new vision for their platform:

• A redesigned architecture for their products. 
• A clear role for Virto Commerce within the new ecosystem.
• A reimagined data architecture.
• Proposals for the future user interface and customer journey.
• A detailed breakdown of what needed to be customized and what had to be built from scratch to support launch. 

Crucially, this was done with the understanding that the legacy system would continue running in parallel with the new one, ensuring that business processes wouldn’t be disrupted during replatforming. 

After just three months, we delivered the first minimum viable product (MVP).

• The MVP included a complete, working end-to-end scenario for the company’s sales representatives.
• Sales reps could now enter the new system and process orders according to their requirements in minutes rather than hours. 

This early win provided our client with immediate, tangible value and set the stage for continued iterative releases. 

Virto Commerce has been on the market as a trusted ecommerce platform provider since 2011, assisting enterprises such as HEINEKEN, Bosch, De Klok Dranken, and the largest GPO in the U.S., OMNIA Partners, in their replatforming journeys.

Virto’s Commerce Innovation Platform is API-first, cloud-native, and PBC-driven, with over 80 out-of-the-box Packaged Business Capabilities or modules—from catalog, pricing, and search to checkout. Every piece of the platform is fully customizable and can be mixed and matched in a way most suitable for your business.

Virto’s platform is composable and de-composable at both platform and deployment levels, which allows companies to have full ownership of their ecommerce solution: deployment, hosting, custom logic, and integration flows tailored to their business. Think 3D printing with the freedom to define your architectural design.

If you’re looking for a trusted B2B ecommerce platform provider for your enterprise, Virto Commerce experts can help you define a gradual and painless replatforming scenario.