Network slicing is the most demonstrated and least monetised capability in 5G. Almost every operator in the region has shown a slice on a stage. Very few have issued an invoice for one. It is worth being precise about why, because the blockers are specific and most of them are not in the radio.
What a slice actually is
A network slice is an end-to-end logical network — radio resources, transport, core user plane and policy — carved out of shared infrastructure with its own performance characteristics and its own policy treatment. It requires a 5G standalone core with the service-based architecture, the network slice selection function, and a policy layer that can enforce the slice’s parameters. On a non-standalone network anchored to a 4G core, what is usually called slicing is quality-of-service prioritisation with better branding.
The technical distinction matters commercially. QoS prioritisation is best-effort with preference. A slice can be sold with a number in the contract.
The four things that have to be true to sell one
1. An SA core. Non-negotiable. Slicing is a Release 15/16 5GC capability.
2. Charging that understands slices. This is the most common blocker in practice. If the charging function cannot rate a slice, the commercial team cannot price it, the billing system cannot invoice it and finance cannot recognise the revenue. A slice you cannot bill for is a demonstration.
3. Policy and orchestration. Creating, modifying and tearing down a slice has to be an operational procedure, not a project. If provisioning a customer slice takes an engineering change request, the product has no margin.
4. Transport that honours the contract. A slice guaranteed in the core and contended in the backhaul is not a slice. Transport slicing or hard capacity reservation has to be designed alongside it.
Where the buyers actually are in this region
The realistic first customers in the GCC are not consumer segments. They are:
- Government and public safety — isolated capacity with priority and pre-emption, and a data-handling posture that can be stated in writing.
- Broadcast and events — guaranteed uplink for a fixed window at a fixed location. This is the easiest slice to sell because the value is obvious and the duration is bounded.
- Ports, logistics and industrial sites — where the alternative is building a private network. A slice is the cheaper option for the customer and the higher-margin option for the operator, provided the latency guarantee is real.
- Enterprise connectivity with a service level — fixed-wireless access sold against an SLA rather than a best-effort tariff.
A sequence that works
Start with one slice type, one customer, one commercial construct. Broadcast or events is the usual right answer: short duration, contained geography, a customer who already understands paying for guaranteed capacity, and a failure mode that is embarrassing rather than contractual.
Prove the whole chain end to end on that one case — provisioning, enforcement, monitoring, reporting and, critically, invoicing. Then productise. Operators that start by building a general slicing capability and then look for customers tend to end up with an impressive platform and no revenue line.
The honest position
Slicing is real, it works, and it is worth building. It is not a reason to buy an SA core on its own — the core case rests on VoNR, URLLC and the enterprise proposition together. But for an operator that already has SA, slicing is the capability most likely to produce a genuinely new revenue line rather than a cheaper way to sell the old one.