A Frankenstack is a retail tech stack built from disconnected point solutions. Most retailers have built their tech landscape in layers: systems added over years, vendors that don't talk to each other, integrations nobody fully understands. That's a Frankenstack. EVA replaces the complexity with one connected commerce platform.
The definition
A Frankenstack is a retail technology environment built up over years of tactical decisions: legacy systems, point-to-point integrations, and bolt-on solutions that were never designed to work together.
The result is a fragile, expensive architecture where every change risks breaking something else, data lives in silos, and real-time visibility across channels is structurally out of reach.
It's the single biggest reason retailers can't deliver unified commerce. Inventory accuracy, order orchestration, and consistent customer experiences all depend on a coherent data model. A Frankenstack doesn't have one.
A Frankenstack isn't a foundation to build on. It's a constraint to replace.
What we did
Retail technology doesn’t become complicated overnight. It happens one system, one integration and one workaround at a time. And eventually, the stack becomes impossible to untangle.
We called that problem the Frankenstack and registered the term as a European Union Trademark so retailers could put a name to the complexity they have been living with.
A shared name creates a shared conversation. And a shared conversation makes change possible. We are here to make that change happen by replacing the Frankenstack, not adding to it.
Symptoms
Changing one system risks breaking others. Without a shared data layer, every dependency is a liability and every release a risk.
Inventory, orders, and customer data live in separate systems and sync on a schedule. By the time data lands, it's already out of date.
Click & collect, ship-from-store and endless aisle require manual operational workarounds. The architecture doesn't support them natively.
IT spends the majority of its budget keeping integrations alive, leaving little capacity to build the capabilities that drive growth.
Every new market, channel, or feature requires a bespoke integration project. The stack doesn't scale, it multiplies.
Every major system feels impossible to replace. The complexity that created the Frankenstack keeps it in place.
25+
Fragmented systems. The number a typical mid-market retailer manages simultaneously, each with its own data model, API contract, and failure mode.
~40%
Of IT budget. The share consumed by integration maintenance rather than innovation or capability-building in a typical Frankenstack environment.
2 to 3×
Longer time-to-market. For new capabilities compared to retailers running on a unified commerce platform, due to integration overhead at every step.
In 2015, we founded New Black with a simple belief: retail should not be held back by its technology. The Frankenstack was making businesses more complex, fragmented and difficult to evolve. Our mission was clear: unify retail on one platform and give retailers the freedom to focus on their customers. That’s where EVA was born.
With EVA, we replaced a fragmented retail setup with one unified commerce foundation, transforming POS, loyalty and fulfilment across 1,000+ stores in months, not years.
KIKO Milano
1,000+ stores in 25 countries
The path forward
In a Frankenstack, customer records, inventory positions, and order data exist in separate systems. EVA replaces this patchwork with a single data layer where every channel reads from and writes to the same information simultaneously.
Batch syncs and nightly jobs leave your inventory data hours behind reality. EVA's Inventory 360 gives you live stock positions across warehouses, stores, and distribution, enabling ship-from-store, endless aisle, and accurate click & collect without manual reconciliation.
A Frankenstack makes order orchestration brittle: every flow is custom-engineered, every edge case a potential failure. EVA's OMS routes orders across ecommerce, stores, and fulfilment networks from one rule-based layer, with no bespoke integration work required.
In a Frankenstack, each new market adds more systems and local workarounds. EVA includes built-in fiscalisation and digital tax reporting across 45+ countries, so compliance is embedded in every transaction, not bolted on afterwards.
From phased migration to full switch
Order Management
Unified order routing
POS & Mobile POS
Store operations
Inventory 360
Real-time visibility
Clienteling
Customer data
Loyalty Engine
Retention programs
Promotion Engine
Pricing & offers
Headless Commerce
API-first frontendFAQ
A Frankenstack is a retail tech stack built from disconnected point solutions added over time, without a unified architecture. The result: tools that don't share data, create operational silos, and break when any single integration fails. In practice, this means a POS, ecommerce platform, OMS, and loyalty tool chosen independently and held together by custom integrations nobody fully owns. The Frankenstack is a registered EU trademark (EUIPO #019227705) coined by New Black.
The clearest signs: inventory that differs between your store and webshop, an IT team spending more time maintaining integrations than building capabilities, and customer data living across systems that can never be fully reconciled. If launching click and collect required a custom project, or if your teams work around the systems rather than with them, you have a Frankenstack.
New Black coined the term in Amsterdam in 2015, after seeing the same fragmented architecture pattern across every major retail implementation they worked on. The name references Frankenstein's monster: something assembled from parts that were never designed to work together. New Black holds the registered EU trademark (EUIPO #019227705, Class 42).
Yes. Frankenstack is a registered EU trademark owned by New Black B.V., filing number EUIPO #019227705, Class 42 (Software services). The trademark is active. New Black registered it to establish ownership of the concept and ensure consistent use across the industry.
A Frankenstack grows without a plan: systems added for individual problems, connected through point-to-point integrations, never reconciled into a coherent whole. A composable commerce platform is built on a unified data model where capabilities share data natively. The difference is not the number of systems but the presence or absence of that shared data layer. Frankenstacks accumulate complexity; composable platforms are designed to contain it.
The only durable fix is replacing it with a platform that has a unified data model at its core. Adding middleware or an API gateway reduces friction but doesn't eliminate the root cause: independently managed systems with separate data models. Those layers add cost, latency, and new failure points. Retailers on EVA have consolidated POS, OMS, inventory, and ecommerce onto one platform, either all at once or starting with the most painful capability.
No. Middleware patches friction; it doesn't fix the structure. When systems have separate data models, an integration layer still translates and synchronises between them, introducing latency, failure points, and maintenance overhead. Every vendor API update or new system added means the integration layer needs updating too. The only fix is consolidating onto a platform where no translation is needed in the first place.
Direct costs include integration maintenance, overlapping vendor licensing, and IT time keeping connections alive. Indirect costs are often larger: slower time to market because every new feature needs an integration project, missed revenue from capabilities like endless aisle that aren't viable on a fragmented stack, and teams spending their time working around systems. Technical debt in a Frankenstack isn't just a technology cost. It's a human one.
Retailers on EVA have gone live at enterprise scale in under 12 months, including deployments across more than 1,200 stores in multiple markets. EVA's unified architecture removes most of the custom integration work that extends traditional timelines, since POS, OMS, and inventory already share one data model. Both phased and all-in approaches work. The right starting point is whichever part of the Frankenstack is causing the most immediate friction.
Optimisation works when the architecture is sound. A Frankenstack's problems are structural: the systems were never designed to share data, and no amount of tuning changes that. You can improve individual components, but you cannot give a fragmented stack a unified inventory position, a shared customer record, or real-time cross-channel visibility. Those require a coherent data model that a Frankenstack doesn't have. At some point, the only decision that resolves the problem is a structural one.
We’ll uncover where fragmentation costs you most and show how a unified foundation can replace it.
