The EV charging industry made a collective bet early on: open standards. OCPP for charger-to-back-end. OCPI for back-end-to-back-end roaming. ISO 15118 for vehicle-to-charger. The bet has largely paid off — these standards are now the foundation of the global EV charging ecosystem.
But the bet wasn’t inevitable. Several major players started with proprietary approaches. Some still maintain hybrid stacks. The “open vs proprietary” question shaped the industry in important ways, and the trade-offs are worth understanding.
This article walks through why open standards won, where proprietary still matters, and what the future likely looks like.
A short history
In the early 2010s, EV charging was a fragmented space. Each charger manufacturer (OEM) had its own back-end. If you bought chargers from three different vendors, you ran three different back-ends. Mixing brands was operationally painful.
Several initiatives launched simultaneously to address this:
- OCPP 1.2 (2010) from a consortium led by ELaadNL in the Netherlands. Targeted at charger-to-back-end communication.
- CCS and CHAdeMO as competing DC fast charging connector standards.
- OCPI later (2014-2017 various drafts) for back-end-to-back-end roaming.
- ISO 15118 for vehicle-to-charger.
Meanwhile, Tesla built its own vertically-integrated stack — proprietary connector, proprietary protocol, proprietary network, proprietary vehicle integration.
By 2020-2022, the open standards had reached production maturity. By 2024-2026, even Tesla had moved to open standards (NACS became SAE J3400; Supercharger compatibility with non-Tesla vehicles via OCPP/OCPI back-ends became reality).
Why open standards won
A few main reasons.
Multi-vendor flexibility for operators
A CPO wants to choose the best charger for each location. Without open standards, that means running multiple back-ends. With OCPP, all chargers talk to one back-end regardless of brand.
The economic value of this for operators was huge. Hardware vendors couldn’t reasonably refuse to support OCPP and still sell to operators.
Roaming
Drivers want to charge anywhere their card works. Without OCPI (or equivalent), every CPO needed bilateral integrations with every eMSP. With OCPI plus roaming hubs (coming soon), the integration count scales as O(N) instead of O(N²).
This made multi-network charging practical for users. Networks that resisted roaming lost market position.
Innovation distribution
When the protocols are open, multiple vendors can compete on implementation quality and additional features. The standards body sets the baseline; vendors compete above the baseline.
This produced a wider variety of products than a single-vendor proprietary world would have.
Vendor sustainability
Open standards reduce the risk to operators of being locked into a single vendor. A vendor whose products are 100% proprietary can hold operators hostage. A vendor whose products comply with OCPP is interchangeable, which forces the vendor to compete on actual product quality.
Counterintuitively, this is good for vendors too — they have a healthier market to sell into.
Regulatory tailwind
In Europe, AFIR (Alternative Fuels Infrastructure Regulation) mandates interoperability and roaming as requirements for publicly-funded charging infrastructure. Similar regulations are emerging in the US (NEVI grants), the UK, and elsewhere. Public money is a strong vote for open standards.
Where proprietary still matters
Open standards won the broad market, but proprietary approaches survive in specific contexts.
Vertical integration advantages
Tesla’s vertical integration during the 2012-2022 period delivered some real benefits — Supercharger uptime, integrated user experience, faster product iteration without standards-body coordination. Even after moving to open NACS, Tesla retains some of these advantages because their vehicle and Supercharger network co-evolve.
For operators who control the vehicle, charger, and back-end (some fleet operations, some specialized applications), vertical integration is still a viable choice.
Customer-facing software
The OCPP/OCPI layer is standardized. The customer-facing app, the operator’s dashboard, the analytics platform — those are all competitive differentiators. Every CPO and eMSP builds proprietary front-ends on the standard back-end.
This is “open infrastructure, proprietary experience” — a sustainable pattern.
Emerging features ahead of standards
When a new capability emerges (e.g., bidirectional charging, advanced V2G), standards bodies move slowly. Vendors who want to ship the capability now sometimes go proprietary for the initial version, then migrate to the standard when it catches up.
Tesla’s early bidirectional pilots, some early Plug & Charge implementations, several V2G grid-services programs — all started proprietary, are migrating to standards as the standards mature.
Niche applications
Specialized applications (military, off-grid, remote locations, marine, industrial) often have requirements that the mainstream standards don’t address well. Proprietary stacks fit better.
The cost of being on the wrong side
Several cautionary stories about the cost of betting against open standards.
Networks built on proprietary protocols pre-OCPP. Some early networks ran their own protocols and tied operators to their back-ends. As OCPP spread, these networks faced a difficult migration. Some never migrated and lost market position.
Vendors who refused to implement OCPP. Lost large customers who insisted on OCPP support. Some adapted; some shrank.
Networks that resisted OCPI roaming. Lost the customer expectation game. Users want one card or one app that works across networks. Resisting roaming pushed users toward networks that supported it.
The pattern: short-term proprietary advantage often loses to long-term standards-based viability.
The cost of standards body coordination
Open standards have real downsides too.
Slower to evolve. Standards bodies move at the pace of consensus. Adding a feature can take 1-3 years from proposal to release. Vendors who want to move faster sometimes find this frustrating.
Lowest-common-denominator risk. Standards have to be implementable by many vendors. This sometimes means the standard is less ambitious than any one vendor would prefer.
Implementation variation. Even with a standard, vendors implement differently. OCPP 1.6 from vendor A and vendor B may have meaningful behavioral differences. Interop testing matters.
Political tension. Standards bodies have member politics. Different members have different commercial interests. The standards reflect compromise.
These are real costs. They’ve been worth paying because the alternative (fragmentation, lock-in, no roaming) is worse.
The current state, by layer
A snapshot of how open standards stand in each layer of the EV charging stack. If you want the layer-by-layer protocol breakdown in more depth, see OCPP vs OCPI vs ISO 15118.
flowchart TD
V[Vehicle] -->|ISO 15118| C[Charger]
C -->|OCPP| B[CPO Back-end]
B -->|OCPI| E[eMSP Back-end]
E --> A[Driver App]
classDef open fill:#dcfce7,stroke:#16a34a
classDef prop fill:#fee2e2,stroke:#dc2626
class C,B,E open
class A propEach arrow is a standardized interface; the shaded nodes lean open, while the driver-facing app stays proprietary.
Charger-to-back-end (OCPP): decisive win for open standards. OCPP is everywhere. Proprietary protocols at this layer are essentially absent in new deployments.
Back-end-to-back-end (OCPI): strong win for open standards. OCPI is the dominant roaming protocol. Some legacy protocols (OICP, eMI3) coexist but are declining.
Vehicle-to-charger (ISO 15118): dominant standard but adoption uneven. ISO 15118-2 is the basis for production Plug & Charge. ISO 15118-20 is the future. No serious proprietary alternative is gaining ground.
Connector standards (CCS, NACS): mostly converged. CCS dominates Europe and most of the world. NACS (formerly Tesla proprietary) now SAE J3400 dominates North America. CHAdeMO is fading. No proprietary connector at scale.
Customer-facing software: entirely proprietary, which is fine. App design, billing UI, route planning — these compete on quality, not standards compliance.
Grid services protocols: mixed and emerging. OpenADR is the main open standard for demand response signaling. Many proprietary APIs coexist. Maturity is still happening.
What the future looks like
A few predictions.
Open standards will continue to win at the infrastructure layer. The economics overwhelmingly favor them. New layers (V2G grid services, advanced smart charging, etc.) will mostly become open standards eventually, even if proprietary implementations come first.
Proprietary value will concentrate in user experience and analytics. The differentiation moves up the stack. Operators compete on app quality, analytics depth, customer service, network coverage — not on protocol uniqueness.
Hybrid approaches will persist. Major operators will continue to have proprietary advantages alongside standards compliance. Tesla retains some vertical-integration benefits even after opening NACS.
Standards bodies will continue to be slower than the market wants. Vendor proprietary implementations will continue to fill the gap for emerging features.
What this means for buyers and operators
A few practical implications.
If you’re buying chargers: require OCPP 2.0.1 minimum. Test interop with your CSMS before committing.
If you’re choosing a CSMS: support for multiple OCPP versions, OCPI roaming, ISO 15118 PnC integration. Don’t lock in to a CSMS that only supports one OCPP version.
If you’re building a network: plan for OCPI roaming from day one. Standalone networks fail.
If you’re a vendor: open-standards compliance is table stakes. Differentiate above that, not below.
If you’re an investor: be skeptical of “proprietary moat” claims in EV charging infrastructure. Sustainable moats are in execution, scale, and customer relationships, not protocols.
The honest summary
Open standards won most of the EV charging world because the economics overwhelmingly favored them. Multi-vendor compatibility, roaming, regulatory alignment, and operator flexibility all pushed against proprietary approaches. The remaining proprietary niches are real but small. Anyone building in this space should treat open standards as the foundation and compete on product quality, customer experience, and execution — not on protocol uniqueness.