SBC-as-a-Service: Why Enterprises Are Ditching On-Prem Session Border Controllers
On-premises SBCs are a relic of the TDM era. HookZ Voice Edge delivers carrier-grade session management with zero hardware, elastic scaling, and built-in compliance recording.
By 2028, the majority of new enterprise session border controller capacity will be consumed as a service rather than deployed as customer-owned appliances.
Key Findings
- Appliance SBCs are sized for peak plus headroom and therefore run at low average utilisation across the estate—typically well under 30%.
- The hidden cost of appliance SBCs is lifecycle: firmware maintenance, certificate management, hardware refresh and dual-site redundancy, which together frequently exceed the original capital cost over five years.
- Regulatory recording and lawful intercept obligations are the most common reason enterprises believe they must keep SBCs on-premises; in most jurisdictions this is no longer technically or legally required.
- Multi-cloud UC adoption has made distributed session control a requirement, which appliance topologies serve poorly.
Recommendations
- Model SBC cost per concurrent session including lifecycle and redundancy, not per appliance; the comparison changes materially.
- Validate recording, retention and residency requirements against the actual regulation rather than inherited assumption before ruling out a service model.
- Adopt a session-based commercial model for variable workloads and a channel-based model for stable trunk capacity; most enterprises need both.
- Retain on-premises session control only where deterministic sub-10ms media handling or physical isolation is a hard requirement.
The Appliance Model Under Strain
The on-premises SBC made sense when voice traffic terminated on physical trunks in a small number of buildings. Enterprise voice today originates from cloud UC platforms, contact-centre-as-a-service, mobile clients and application-embedded calling, and terminates across multiple carriers in multiple jurisdictions. Anchoring that topology to a pair of appliances in a data centre adds latency, cost and a single point of architectural rigidity.
Elastic demand makes the mismatch worse. An enterprise that occasionally runs a campaign requiring five times its baseline concurrency must either buy for that peak permanently or refuse the campaign.
Benchmark: Appliance vs. Service
| Dimension | On-premises appliance | Virtual SBC (self-managed) | SBC-as-a-Service |
|---|---|---|---|
| Capacity model | Peak + headroom, fixed | Fixed VM sizing | Elastic per session |
| Typical utilisation | 20-30% | 30-45% | 70-90% |
| Time to add capacity | Weeks to months | Days | Minutes |
| Redundancy cost | Duplicate hardware | Duplicate VMs | Included in fabric |
| Lifecycle burden | Enterprise | Enterprise | Provider |
| Multi-region reach | Per-site deployment | Per-region deployment | Native distributed mesh |
| 5-yr cost index | 100 | 72 | 44-58 |
When On-Premises Still Wins
- Hard media-latency requirements in trading, broadcast or industrial control environments.
- Air-gapped or classified networks where external session control is prohibited by policy.
- Existing sunk investment with more than two years of remaining useful life and stable, non-growing concurrency.
Bottom Line
The strategic question is no longer whether session border control can be consumed as a service—it demonstrably can, at carrier grade—but whether the enterprise's remaining on-premises footprint is justified by a genuine constraint or by inherited assumption. Most enterprises find, on audit, that it is the latter.
This analysis is published by HookZ.ai Research for enterprise planning purposes. Benchmark ranges are directional and derived from modelled reference estates; actual results vary by estate composition, region and operating model.
Want this benchmarked against your estate?
Our solutions architects can run a tailored TCO analysis, migration roadmap or architecture review.