5G SA: Why Monetization Is Still Stuck (Slicing, URLLC, MEC)

5G Standalone was supposed to change everything. Cloud-native core, service-based interfaces, clean separation of control and user planes. Over 90 operators have deployed it. And yet, in 2026, monetization remains largely theoretical. The Opensignal report puts it bluntly: “Architecture Deployed, Monetisation Pending.”

This is not a technology problem. It is a scaling problem. And understanding it requires examining the three pillars that were supposed to generate new 5G revenue: network slicing, URLLC, and MEC.

The 5G SA paradox: deployed but not profitable

More than 90 operators worldwide have launched 5G SA networks. That number sounds impressive, but it represents only 10% of MNOs globally. Worse: the momentum is slowing. In 2024, 12 new SA launches were recorded. In 2025, only 5.

Why the slowdown? Because early deployments have not yet demonstrated a clear return on investment. Operators that were waiting are watching those who took the leap, and what they see does not encourage acceleration.

This diagram illustrates the promise: a modular architecture where each slice can be configured, billed, and monitored independently. In theory, it is a massive B2B revenue generator. In practice, the blockers are numerous.

Network slicing: the Asia-Pacific paradox

Network slicing is the most publicized 5G SA use case. The principle is elegant: create isolated virtual networks on a single physical infrastructure, each with its own throughput, latency, and reliability guarantees.

The raw numbers are encouraging. Industry tracking identifies 118 slicing use cases across 56 CSPs. But only 65 have moved beyond the proof-of-concept stage. And the geographic concentration is striking.

Asia-Pacific captures 91% of slicing revenue

According to ABI Research, the global network slicing market is estimated at $6.1 billion in 2025, projected to reach $67.52 billion by 2030. But 91% of that revenue is generated in Asia-Pacific, with China accounting for 95% of that segment alone.

This means that network slicing, in Europe and North America, is still a niche product. The reasons are multiple:

1. BSS/OSS chain complexity. Selling a slice requires the ability to configure, provision, bill, and monitor it in real time. This demands a complete BSS overhaul that most European operators have not yet undertaken.

2. No structured demand. Enterprises do not know what a slice is. They do not know what they could do with one. The market-making work remains to be done, and operators are not naturally equipped to lead it.

3. Unclear pricing. How much does a slice cost with 50 Mbps guaranteed throughput and 10ms latency for a 4-hour sports event? Nobody has a standardized answer. Dynamic pricing remains experimental, though T-Mobile announced a dynamic slicing system for early 2026.

4. Interoperability. A slice that works on one operator’s network does not necessarily work on another’s. For multinational enterprises, this is a major friction point. GSMA’s work on cross-operator slicing is still in its early stages.

Why China dominates

China succeeded because three conditions were met simultaneously: strong political will (slicing is part of the national 5G plan), a massive subscriber base to amortize infrastructure costs, and deep vertical integration between operators and domestic equipment vendors.

Replicating this model in Europe or Africa is a structural challenge, not an ambition problem.

URLLC: the 1ms promise that stays on paper

Ultra-Reliable Low-Latency Communication (URLLC) is the second pillar of 5G SA monetization. 3GPP specifications define a clear target: 99.999% reliability and 1ms end-to-end latency.

On paper, URLLC opens high-value markets: remote surgery, autonomous driving, real-time industrial control. In reality, no commercial deployment achieves these specifications under real-world conditions.

Why 1ms is unachievable today

End-to-end latency does not depend solely on the radio network. It includes backhaul transport, core processing, and application response time. Even with an optimized air interface, field measurements show typical latencies of 5 to 15ms in the best 5G SA deployments.

Achieving 1ms would require:

• Processing at the antenna edge (hence MEC, which we will examine next)
• Real-time synchronization across all network layers
• Near-zero congestion on the relevant segment
• Devices capable of prioritizing URLLC packets

None of these conditions are met at commercial scale in 2026. URLLC remains confined to test environments, closed industrial campuses, or media demonstrations. This is not monetization: it is applied R&D.

The impact for operators

URLLC was supposed to justify premium pricing for B2B services. But as long as specifications are not reproducibly met, enterprise contracts remain indexed on classic SLAs (20 to 50ms, 99.9% availability). The URLLC premium does not exist yet.

MEC: the edge promise that lacks nodes

Multi-Access Edge Computing (MEC) is the third element of the 5G SA monetization trilogy. The idea: move data processing closer to the user, at the network edge, to reduce latency and enable real-time applications.

In theory, MEC is the natural companion to slicing and URLLC. In practice, deployment is anecdotal.

The BT case: one site in three years

The most revealing example is BT in the United Kingdom. Three years after launch, BT has only one operational AWS Wavelength site, in Manchester. There are no announced plans to open additional sites.

This is not a BT-specific problem. MEC suffers from three fundamental blockers:

1. Infrastructure cost. Deploying compute nodes at the network edge is expensive. It requires servers, cooling, physical security, and above all maintenance at sites often not designed to host IT infrastructure.

2. Edge congestion. Early field reports reveal a counterintuitive problem: overloaded edge nodes that show higher latency than centralized cloud. Edge is not magically faster: it is faster only when properly sized and not saturated.

3. Partnership fragmentation. AWS Wavelength, Azure Edge Zones, Google Distributed Cloud: each hyperscaler has its own MEC platform. For an operator, choosing a partner means locking enterprise customers into a specific ecosystem. And for developers, it means multiplying integrations.

This quadrant summarizes the situation: eMBB and FWA (Fixed Wireless Access) are the only truly monetized use cases. Asian slicing is progressing. Everything else remains in experimental or niche zones.

The B2B2X model: the missing key

The industry agrees that 60% of new 5G revenue will come through B2B2X models: the operator sells a service to an enterprise, which integrates it into its own end-customer offering. This is the logical framework for slicing (a slice for a factory, a stadium, a campus), URLLC (an SLA for an industrial process), and MEC (edge compute for a business application).

But the B2B2X model requires capabilities that operators do not naturally possess:

• Consultative selling: understanding customer business processes, not just selling connectivity
• System integration: interfacing the slice with the customer’s ERP, MES, or CMMS
• Contractual SLAs: guaranteeing service levels with financial penalties, not best-effort commitments
• Usage-based billing: charging based on actual consumption, not monthly flat rates

This shift is comparable to the one telcos never completed with cloud ten years ago. The question is whether they will succeed this time.

What the field reveals

Field diagnostic tools reveal a systematic gap between theoretical specifications and actual 5G SA network performance. Drive test latency measurements show values 5 to 15 times higher than URLLC targets. NAS/RRC signaling analysis shows that slicing is often poorly negotiated during network attachment. And indoor measurement campaigns reveal zones where the allocated slice cannot be maintained due to radio signal degradation.

These findings are not failures. They are the realities of a maturing technology. But they explain why monetization lags behind deployment.

Outlook 2026 to 2028

Three factors could accelerate progress:

1. Standardized network APIs. The CAMARA project (GSMA/Linux Foundation) aims to expose network capabilities (slicing, QoS, location) through standardized APIs. If developers can order a slice via an API as simply as a cloud server, adoption will follow.

2. Regulatory pressure. In Europe, 5G coverage obligations are beginning to transform into quality-of-service requirements. Operators that cannot monetize slicing will need to justify their investments differently.

3. Repeatable industrial use cases. Early successes in ports, mines, and airports need to be standardized and packaged so that other enterprises can adopt them without custom projects.

Conclusion

5G SA is not a failure. The architecture is solid, the specifications are ambitious, and early deployments prove the concept works. But monetization will follow only if the industry solves three concrete problems: simplify slicing provisioning, achieve reproducible URLLC performance, and densify MEC beyond proof-of-concept.

Until then, the most honest answer to “where are the 5G SA revenues?” is: in Asia-Pacific. And for the rest of the world, the window of opportunity is still open, but it will not stay open indefinitely.