The Fault Nobody Owns
Start with the scenario that costs Australian businesses the most, because every other argument for bundling is downstream of it.
Calls have been dropping out for a fortnight. Not always — mid-morning and mid-afternoon mostly, and never when anyone is testing. You ring the phone provider. They check the platform, look at your call records, and tell you truthfully that the platform is healthy and the packet loss is happening before the traffic reaches them. It is a network problem.
So you ring the internet provider. They test the line, confirm it is performing within specification, note that a speed test looks fine, and tell you truthfully that the connection is not the issue. It is a phone system problem.
Both of them are right, and that is the problem
Voice faults are very often interaction faults. The connection is fine on average and briefly terrible at the moment that matters. The platform is fine and receiving damaged audio. Neither supplier can see the other's half, neither is lying, and neither has any commercial reason to keep digging. The business ends up running the investigation — with no access to either network, no packet captures, and a receptionist apologising to customers in the meantime.
When one provider supplies the access service and the voice platform, this conversation cannot happen. There is no boundary to argue about, no second company to ring, and no scenario where two correct answers add up to no answer. The person you reach can see the call in the platform and the connection carrying it, in the same view, at the same time.
That is the benefit. Everything else in this article is real, and none of it is as valuable as that.
Prioritisation: Only Works If Someone Owns the Path
"Voice prioritisation" is a phrase used loosely enough to be almost meaningless, so it is worth being precise about which parts of the journey can actually be controlled.
| Segment | Who controls it | Can voice be prioritised? |
|---|---|---|
| Your LAN and router | You, or whoever configured the router | ✓ Yes — and this is where most avoidable damage occurs |
| The access tail to the provider | Your internet provider | Partly — depends on the plan class and how the provider handles it |
| The provider's own network | Your internet provider | ✓ Yes, if the voice platform is inside the same network |
| The public internet between providers | Nobody in particular | ✗ No — this is the segment bundling removes |
Read that last row carefully, because it is the actual mechanism. Bundling does not give anybody magical control over the public internet. What it does is shorten the journey so that less of it is uncontrolled. If your access provider is also your voice provider, your call traffic can stay inside one network from your router to the platform. If the two are separate companies, your audio has to cross a boundary neither of them owns, at a peering point chosen for commercial reasons, on a path that can change without notice.
The single most common cause of bad calls
It is worth naming, because it is fixable in ten minutes and almost nobody does it: someone in your office uploading something large while a call is in progress. A cloud backup kicking off at 2pm, a designer pushing a video, an accountant syncing a huge file. On an asymmetric connection the upstream saturates, your own router queues the voice packets behind the bulk traffic, and every call in the building degrades simultaneously.
QoS on your router fixes this by marking voice and letting it jump the queue. A provider that supplies both layers will configure it as part of the install rather than leaving it as an exercise for you. For the diagnostic detail, see how to troubleshoot VoIP and SIP problems.
The Upload Number Everyone Ignores
Australian businesses shop for internet by download speed because that is how it is advertised. For a business that talks to customers, the download figure is close to irrelevant and the upload figure decides how your calls sound.
Everything your business sends travels upstream: your voice on every call, your camera in every video meeting, your screen when you share it, and every file you upload. Consumer plans are built for the opposite pattern — receiving streamed video — so they are deliberately lopsided.
~100 kbps
Per concurrent call, each direction, once packet overheads are counted
10:1
Download-to-upload ratio on a consumer 500/50 plan
2.5:1
Ratio on business 250/100 — the same connection, built for talking
Voice itself is remarkably light. Ten simultaneous calls need only about a megabit each way. The problem is never the calls — it is what the calls have to share the upstream with. Fifty megabits of upstream sounds generous until one machine decides to back up 40 GB to the cloud, at which point every call in the building is competing for scraps. A hundred, two hundred or four hundred megabits of upstream absorbs that without anyone noticing.
One question to ask before you buy any connection for a business that uses phones
"What is the upload speed, and what is the typical evening upload speed?" The published typical figures matter because they are measured under load. On Uniden Voice business plans they are stated openly: 237/96 on the 250/100 tier, 467/191 on 500/200 and 957/388 on gigabit. A provider that will not publish typical speeds is telling you something.
There is a second reason business plans matter here, and it catches people out: home plans use CGNAT, a shared public IP address. Business plans include a static IP, which is what you need for site-to-site connectivity, remote access, hosting anything locally, and for allow-listing your traffic. Our guide to how much bandwidth a phone system needs goes through the sizing arithmetic properly.
What Business nbn Actually Costs
Published Uniden Voice pricing, so you can compare it against your current arrangement rather than against a range in a brochure. Every plan includes unlimited data, one phone line, free setup, free number porting, month-to-month terms with no lock-in, Australian-based support and a 99.99% carrier-grade network uptime target. Plan changes are free once a month.
| Plan | Speed | Typical evening | Monthly | Suited to |
|---|---|---|---|---|
| nbn Home | 500/50 100/20 on FTTN, FTTC, FTTB |
92/17 on the 100/20 tier | $79.90 (reg. $99.90) | Home offices and sole traders. CGNAT shared IP. |
| nbn Business | 250/100 | 237/96 | $89.90 (reg. $104.90) | The default for most small businesses. Static IP, business SLA, priority support. |
| nbn Business | 500/200 | 467/191 | $109.90 (reg. $124.90) | Teams doing video, cloud backup and heavy file movement alongside calls. |
| nbn Business Gigabit | 1000/400 | 957/388 | $139.90 (reg. $164.90) | Larger sites, contact centres, and anywhere local server hosting matters. |
Add the unlimited calls pack at $15/month and, for most small businesses, the entire telephony line item on the P&L is the phone system seats plus fifteen dollars. Full current details are on the nbn plans page, and the per-user platform pricing is on the plans and pricing page. Prices shown here are the published rates at the time of writing and are illustrative for comparison — confirm current pricing before you budget.
The Bundled vs Split Comparison
A worked example for a ten-person business, using round illustrative figures so you can substitute your own. The purpose is not to prove a specific saving — it is to show where the differences sit, because most of them are not in the headline rate.
| Line item | Split across two providers | Bundled with one provider |
|---|---|---|
| Business internet | Business plan from ISP | Business plan, phone line included |
| Voice line rental / service fee | Charged separately by the phone provider | ✓ Included |
| Call charges | Per-call or a separate pack | Unlimited calls pack, $15/month |
| Router configuration and QoS | Yours to arrange, or an install fee | ✓ Configured for voice at install |
| Bills to reconcile monthly | Two, on different cycles | ✓ One |
| Contracts and renewal dates | Two, rarely aligned | ✓ One, month-to-month |
| Support numbers when calls sound bad | Two — and a likely dispute between them | ✓ One, with visibility of both layers |
| Time lost to a disputed intermittent fault | Weeks, unbudgeted, recurring | ✓ One owner, one diagnosis |
The first three rows are the ones a CFO will notice. The last five are where the money actually goes, and they never appear in a price comparison because they are paid in the owner's time rather than on an invoice. We take the consolidation argument further in what you save by putting communications with one provider.
Five Benefits That Never Appear on a Comparison Table
One install, one appointment
The connection, the router, the QoS configuration, the handsets and the apps get set up in one visit by people who know what the other half needs. No sequencing two suppliers who have never spoken to each other.
One renewal conversation
Two contracts on different cycles means you are always mid-term on one of them, which is exactly how businesses end up unable to change anything. One month-to-month relationship means every decision is available every month.
Growth without renegotiation
Adding staff means adding seats, and if that pushes the connection, the tier moves — free once a month. Growth stops being a procurement exercise.
Diagnosis instead of debate
Call quality data and connection data in one place means a fault is a fact rather than a theory. This is the difference between "we think it's your internet" and "here is the four-second period where packets were dropped".
One accountable Australian organisation
The connection and the platform both answer to the same Australian company — which also means your voice data stays on Australian infrastructure. See where your calls actually live.
Porting handled once
Free number porting as part of the same order, sequenced by the people doing the install — rather than coordinated by you across two companies with different processes and different lead times.
The Honest Answer on Outages
The obvious objection to bundling: doesn't putting calls and internet on one connection mean losing both at once? It deserves a straight answer in two parts.
Part one: your handsets stop, your number does not
This is the part that genuinely changes with a cloud phone system, and it is easy to underrate. Because the system lives in the network rather than in your comms cupboard, your phone number is not attached to your building. When the connection drops, calls can be re-routed instantly to:
- the mobile app over 4G or 5G, so staff keep taking calls on the business number from their phones;
- mobile numbers, a second site, or a home worker;
- an AI voice agent that answers in an Australian accent, takes the details, answers common questions and books appointments while you are offline.
Compare that with an on-premise PBX on the same failed connection, where the answer is silence and a voicemail nobody can retrieve. The failure mode of a cloud system is inconvenience; the failure mode of a box in your building is invisibility.
Part two: the caveat bundling genuinely carries
One provider means one access service
If your operation cannot tolerate any interruption, mobile failover is a mitigation, not redundancy. Real redundancy means a second access path — ideally a different technology on a different physical route. That is an architecture decision, and it is a good one for contact centres, clinics, and anywhere a lost call is a lost patient or a lost order. Bundle the primary service for accountability, then add the second path deliberately. We set out the whole design in why one network is a single point of failure.
It is also worth knowing what your provider owes you when something does break. Australian carriers now face enforceable obligations around telling customers about significant outages and keeping public outage records — see public outage registers and the new transparency rules.
When Bundling Is the Wrong Decision
We would rather you buy the right thing than the bundled thing. Two cases where splitting is correct:
You already have a good business-grade service under contract
If you have fibre ethernet with a real availability SLA, or an nbn service you are genuinely happy with and meaningful termination costs, breaking it early to consolidate rarely pays. Do it in two steps: bundle the voice now, bundle the access at renewal. You get most of the accountability benefit immediately at no exit cost.
You need deliberate path diversity
If downtime is genuinely unacceptable, you want two services on different technologies and different physical paths, which means more than one relationship somewhere by definition. Bundling optimises for accountability; diversity optimises for survival. Larger sites should want both — bundle as primary, diverse second path behind it.
Notice what is not on that list: "because the bundled price isn't the cheapest line on a comparison site". Almost every business that chases the cheapest connection for a phone-dependent operation ends up paying for it in the fault nobody owns.
Why 2026 Is the Year It Comes Up
This question is being asked more this year than last, and not by coincidence. Three things are forcing it onto the agenda at once.
| What is happening | Why it puts bundling on the table |
|---|---|
| Copper disconnection and forced fibre upgrades | If your service is being migrated anyway, you are already changing the connection — the cheapest moment to reconsider both layers is while one of them is being replaced. See forced fibre upgrades and copper disconnection. |
| Retail price rises across the major providers | When bills move, businesses actually read them. Two invoices from two suppliers rising on two different cycles is the classic trigger for consolidation — see the July 2026 nbn price rises. |
| Higher-speed tiers becoming standard | Full-fibre footprint and faster tiers mean the upload constraint that made cloud voice marginal in some premises has gone. Read what full-fibre upgrades mean for your business phone. |
If your premises is in a new fibre estate rather than the nbn footprint, the same logic applies with a better connection underneath it — see getting connected with Uniden Voice on LightFast Networks.
The Switchover Sequence
Done in this order there is no window where a customer rings and gets nothing. The order is the whole trick.
| Step | What happens | Risk if you skip it |
|---|---|---|
| 1. Install the new connection alongside the old | The new nbn service goes in and is tested while the existing service still carries everything. | Discovering an install problem at the moment you need the line to work. |
| 2. Build and test the phone system on it | Extensions, call flows, IVR, ring groups, apps and handsets configured and tested — with live calls still on the old service. | Learning your call flow is wrong while customers are inside it. |
| 3. Train the team before cutover, not after | Twenty minutes on the app and the handset while there is no pressure. The apps run on Android, iOS and PC, so no hardware purchase is required. | Day-one confusion being blamed on the platform. |
| 4. Port the numbers | The one irreversible step. Free porting, scheduled cutover window, existing numbers retained — the number on your van and your invoices keeps working. | Nothing, if steps 1–3 are done. This is why they come first. |
| 5. Cancel the old service — after the port completes | Only once numbers are live on the new service and tested from an outside line. | The single most common way businesses create an outage. |
If you are moving from a physical PBX rather than another cloud provider, hosted PBX versus on-premise covers what changes and what you can keep, and how Uniden Voice works on the nbn covers the mechanics of running business voice over an nbn service.
Frequently Asked Questions
What to Read Next
Bundling is a decision about accountability. These cover the layers either side of it.