Hosted PBX, start to finish.
What it is, how it compares to UCaaS and on-premises systems, what features matter at enterprise scale, what it really costs, and how to migrate without breaking what already works. Built for IT and communications leaders evaluating cloud voice.
What is hosted PBX, defined.
Hosted PBX shifts three things off your plate: the hardware (no on-prem call servers), the lifecycle (no upgrade planning, no spare parts), and the uptime responsibility (SLA-backed, geo-redundant by default). What you keep: configuration, user lifecycle, integrations, and compliance posture. For most enterprises, that trade is worth the loss of absolute hardware-level sovereignty — but the calculus changes for regulated or high-availability sites, which is where hybrid architectures come in.
Hosted vs. UCaaS vs. on-prem, side by side.
These three architectures sound interchangeable in marketing copy. They are not. They differ on what you manage, what fails when, and how they price.
Enterprise features, at scale.
Six capabilities separate enterprise-grade hosted PBX from the per-seat tools sold to small businesses. Each one is about governance at scale, not feature count.
Intelligent call routing
Skill-based and time-of-day logic, hunt groups, conditional auto attendants, policy-driven failover across sites.
Softphones & mobile
Business number follows the user across desk phone, desktop, and mobile. Codec negotiation adapts to bandwidth.
200+ integrations
CRM, ERP, ITSM, Microsoft Teams, vertical applications. Open APIs and webhooks for custom workflows.
Analytics & QoS
Call volume, concurrency, and Quality of Service metrics — latency, jitter, packet loss — down to site or user level.
Voicemail & retention
Centralized voicemail with voicemail-to-email, compliance-aligned retention controls, and audit-grade logging.
Global PSTN replacement
One platform across regions, local-number presence in each operating country, consolidates carrier relationships.
Security & compliance, non-negotiable.
Voice goes global the moment you move it to the cloud — and that changes the security model. Three risks that don't exist on-prem the same way:
Toll fraud is a security issue, not a billing issue
A single compromised SIP credential can generate six-figure liability in minutes through international call-injection. Voice credentials need zero-trust rigor: fraud detection, geographic restrictions, anomaly alerting.
Data sovereignty > data residency
Residency is where data sits. Sovereignty is whose laws apply. For healthcare, legal, government, and financial services, that distinction decides whether a cloud deployment is even viable.
E911 in distributed environments
Remote workers break static emergency locations. Compliance with Kari's Law and RAY BAUM'S Act requires dynamic "dispatchable location" data — floor and room — for every user, every move.
Cost & pricing, without surprises.
Three pricing models, three different math problems. Pick the wrong one and you'll overpay for years.
Per-user subscription
Flat or tiered monthly rate per named user. Simple to budget, no usage modeling required, scales linearly with headcount.
Concurrent call paths
Pay for simultaneous active calls, not registered users. Far more economical when a large workforce uses the phone intermittently.
Platform fee + per-user
Fixed platform fee covers shared infrastructure and admin overhead; incremental per-user rate scales down at volume. Front-loaded for small orgs, efficient at scale.
What the headline rate doesn't include
Hybrid & remote work, anywhere.
Hybrid deployments
For many organizations, hybrid isn't a migration step — it's the end state. Cloud handles the distributed knowledge workforce; on-prem keeps regulated, high-availability, or specialized-integration sites under direct control.
- Cloud for distributed users and modern collaboration
- On-premises for regulated or local-survivability sites
- SIP trunking bridges both during transition
- Single admin interface manages both environments
Remote & hybrid workforce
Hosted PBX ties business numbers to user profiles, not desk ports. The phone follows the person — across home offices, branch sites, and mobile devices — with the same policies and audit trails.
- Business number across desk, desktop, and mobile
- Provisioning in minutes via web portal
- QoS visibility per user, location, and time of day
- Call data stays in the business environment, not on personal carriers
Migration, phased.
A governed enterprise migration takes three to six months. Treating it as a procurement exercise instead of a governance project is the single biggest cause of post-cutover failures.
Discovery
Inventory current state, identify analog devices (fax, paging, elevator phones), document integrations.
Architecture
Design routing logic, integration scope, dispatchable-location strategy, compliance posture, failover paths.
Network readiness
SD-WAN, QoS marking, disable SIP ALG, validate bandwidth and codec policy at every site.
Pilot
Validate call quality and routing logic in a real environment with a representative user group before scaling.
Cutover
Phased rollout, number porting (4–8 weeks for international), coexistence with legacy, user training, decommission.
Different size, different priorities
If you're a small or mid-sized business, most of this guide is over-engineered for your reality. You don't need SCIM lifecycle automation or concurrent call path modeling — you need predictable per-seat pricing, fast setup, and minimal IT overhead. Same architecture, different evaluation criteria.
Quick answers, 30 questions deep.
The five questions IT and procurement teams ask most often. The full 30 — covering security, hybrid, emergency services, and more — live on the FAQ page.
Plan your hosted PBX strategy with a Mitel specialist.
Have a Mitel communications expert walk through your architecture, compliance requirements, and migration path. No cost, no commitment.