Before an enterprise chooses a vendor, a spectrum model, or a rollout plan, it needs to answer a more fundamental question: who should actually own and operate the network. Four ownership models dominate current private 5G deployments, and each represents a genuinely different trade-off between control, capital investment, operational burden, and long-term flexibility.
|
Model |
Capital Cost | Operational Burden | Control |
Accounting Treatment |
|---|---|---|---|---|
| Enterprise-owned (SNPN) | Highest | Highest — needs internal capability | Full | Predominantly capex |
| Operator-managed / NaaS | Lowest | Lowest | Limited, governed by contract | Predominantly opex |
| Neutral host / hybrid | Moderate, shared across tenants | Shared with host/partner | Partial, split by layer | Mixed, depends on structure |
| Spectrum-leased | Moderate | Depends on paired infrastructure model | Exclusive spectrum, time-bound | Depends on paired infrastructure model |
Enterprise-Owned: Maximum Control, Maximum Responsibility
A fully enterprise-owned Standalone Non-Public Network means the organisation holds the spectrum rights, owns the radio and core infrastructure, and operates the network directly. This model delivers the strongest data sovereignty, the greatest control over technology refresh timing, and the fewest external dependencies, an enterprise decides when to upgrade, how to configure security policy, and retains full ownership of the network as a long-term strategic asset. It also carries the highest capital cost and the greatest operational burden: the enterprise needs the internal capability, or a strong integration partner, to design, deploy, and run the network on an ongoing basis. This model tends to suit large, strategic, long-duration deployments where the enterprise is planning to treat the network as core infrastructure rather than a service.
Operator-Managed: Lower Capital Outlay, Recurring Dependency
A managed service or Network-as-a-Service model shifts ownership and day-to-day operation to a vendor or mobile network operator, with the enterprise paying a recurring fee for a defined level of service. This dramatically lowers the upfront capital investment and the internal operational burden, making it a more accessible entry point for enterprises without deep in-house telecom expertise. The trade-off is a longer-term dependency: the enterprise’s network access, performance guarantees, and cost trajectory are governed by the terms of an ongoing commercial relationship rather than being fully within its own control, which makes contract terms, SLA enforcement, exit provisions, and pricing stability, the most consequential part of this model to get right.
Neutral Host and Hybrid Models: A Middle Ground
Hybrid arrangements, where the enterprise owns the radio access layer while a vendor or operator provides the core and spectrum, or where a neutral host operator serves multiple tenants within a shared physical infrastructure, sit deliberately between the two extremes above. These models are particularly common in multi-tenant environments like airports, shared industrial parks, or large campuses, where several organisations benefit from sharing physical radio infrastructure while maintaining logically separate, secure network slices for their own traffic. The key evaluation point for hybrid and neutral host models is contractual clarity at the boundary between what the enterprise controls and what the host or partner controls, particularly around fault resolution and change management when an issue spans that boundary.
Spectrum-Leased: A Newer Fourth Path
A structured spectrum-leasing arrangement, increasingly viable following documented examples like the Jawaharlal Nehru Port Authority’s 2026 procurement, offers a distinct ownership profile: the enterprise gets dedicated, exclusive spectrum access without holding a licence directly, typically paired with either enterprise-owned or hybrid infrastructure ownership above the spectrum layer. This path is worth active consideration specifically in markets where CBRS-equivalent shared spectrum isn’t available and direct licensing is impractical for a single site, giving those enterprises a genuine path to dedicated spectrum access that didn’t reliably exist a few years ago.
Capex Versus Opex: The Financial Accounting Dimension
Beyond operational control, ownership model has a direct financial accounting impact worth involving finance stakeholders in early. An enterprise-owned network is predominantly a capital expenditure, an asset the organisation depreciates over its useful life, which affects balance sheet structure and depreciation schedules. A managed service model is predominantly an operating expenditure, a recurring cost that flows through the income statement rather than the balance sheet. Some organisations have a clear institutional preference for one treatment over the other, driven by broader capital allocation policy, tax considerations, or how technology investment is reported to the board, independent of the operational merits of either model. Surfacing that institutional preference early, rather than discovering it only once a preferred technical model is already selected, avoids a late-stage reversal that can cost significant planning time.
Transition Paths Between Models Over Time
Ownership model isn’t necessarily a permanent choice made once. Enterprises commonly start with a managed service model to prove out the business case with lower upfront risk, then transition toward enterprise ownership once the deployment has demonstrated clear value and justifies the larger capital commitment — effectively using the managed service phase as an extended, lower-risk pilot for the ownership decision itself. The reverse transition, from enterprise-owned toward a managed model, is less common but does happen, typically when an organisation’s strategic priorities shift away from treating network infrastructure as a core competency it wants to build in-house. Planning for the possibility of a future transition, even when starting with a clear initial preference, is worth building into contract terms from the outset — an initial managed service agreement with weak exit and data portability provisions can make a later transition to enterprise ownership far more costly and disruptive than it needs to be.
Matching the Model to the Organisation
The right ownership model depends less on which offers the most control in the abstract, and more on an honest assessment of the enterprise’s own capability, deployment scale, financial reporting preferences, and time horizon. Organisations planning a single, large, long-duration deployment with in-house telecom capability tend toward enterprise ownership. Organisations wanting to move quickly with minimal internal operational burden tend toward managed service models, often as a deliberate first phase ahead of a possible later transition. Multi-tenant sites tend toward neutral host or hybrid arrangements. And enterprises in markets without accessible dedicated or shared spectrum increasingly have a genuine fourth path in structured leasing. None of these is universally correct, the decision is specific to each organisation’s circumstances, and it’s the first question worth resolving before moving into vendor evaluation or spectrum procurement in detail.
TeckNexus’s RFP Scorecard Generator activates different evaluation criteria depending on which ownership model you choose — https://tecknexus.com/private-network-rfp-scorecard-generator/
