How the Market Is Actually Layered
Buy a business phone service in Australia and you are almost certainly buying from a company that owns none of it.
The typical structure has four layers. A wholesale platform vendor writes the software and sells white-labelled access to it. A network operator carries the calls and holds the interconnects. A data centre or cloud provider runs the compute, sometimes offshore. And on top sits a reseller — the company whose logo is on your invoice, whose website you read, and whose support number you will ring.
This is not a scandal. Reselling is a legitimate business model and plenty of resellers are good at the parts they control: onboarding, account management, being pleasant to deal with. Specialisation produces efficiency, and a small reseller with excellent local service can be a better experience than a large operator with a bad one.
The problem is not competence. It is authority. A reseller cannot fix what a reseller cannot see — and nothing in a sales process reveals where that line sits.
The distinction this whole article rests on
The Escalation Ceiling
Every provider has a point beyond which they stop being able to act and start being able only to relay. Call it the escalation ceiling. Below it, they investigate, adjust and fix. Above it, they lodge a ticket with someone else and wait — and while they are waiting, they know precisely as much as you do.
| Where the ceiling sits | What they can do below it | What happens above it |
|---|---|---|
| Reseller of a wholesale platform | Account changes, call flows, user setup, billing, explaining features | Raise a ticket with the platform vendor. Then wait. No visibility, no priority, no ability to escalate beyond a queue position |
| Reseller with its own support engineering | The above, plus real diagnosis of configuration and endpoint faults | Better diagnosis, same wall. They can tell you accurately what is wrong and still cannot fix it |
| Operator of its own network and platform | Everything above, plus routing, capacity, interconnects, platform behaviour, incident response | ✓ Only genuinely external causes — an access network fault, an upstream carrier, a customer-site issue |
Notice that the third row still has a right-hand column. That is the honest part, and we come back to it. But the difference between “we are waiting on our supplier” and “our engineers are on it” is the difference between an incident with an owner and an incident without one.
The Same Fault, Two Providers
Concrete beats abstract. Here is one realistic fault, handled two ways.
The fault: from about 2pm, roughly one in ten inbound calls drops after eight seconds. Outbound is fine. It is intermittent, so it is hard to demonstrate, and by the time anybody has the provider on the line it has stopped for a bit.
| Hour | With a reseller | With the operator |
|---|---|---|
| 0:00 | You ring. They are helpful. They ask you to reboot the router and collect examples. | You ring. They ask for two example call times and pull the call detail records while you are on the line. |
| 0:30 | You have three examples. They lodge a ticket with the platform vendor. | They can see the calls failing at a specific point in the path and confirm it is not your equipment. |
| 2:00 | Ticket acknowledged. No detail. You are asked whether anything changed on your network. | Cause identified. You are told what it is in plain language and given an expected resolution time. |
| Next day | Vendor asks for a packet capture. Nobody at your provider can take one. | Fixed. A short written explanation of cause and remedy follows. |
| Day 4 | Resolved, cause described vaguely as “an upstream issue”. You never find out what happened. | — |
Where the real cost sits
Not in the fault. In the four days of somebody in your business acting as an unpaid project manager between companies that have no contract with each other — chasing, re-explaining, collecting evidence for a party they cannot speak to. That time never appears on an invoice, which is exactly why it never appears in a comparison.
What Owning the Network Changes About Diagnosis
Four specific capabilities, and each maps to a class of fault that is otherwise very hard to resolve.
Both halves in one view
Call records and network path data side by side. Most intermittent voice faults are only visible in the correlation between the two, which is why split ownership makes them so hard to catch.
Routing is a decision, not a request
If a route is degraded, the party who owns it can change it. A reseller can only ask, and wait for someone else’s change window.
Priority means something
Your fault competes with other customers of the same company, not with the entire wholesale customer base of a platform vendor who has never heard of you.
You get a real cause
Not “an upstream issue”. A description of what happened and what was changed — which is what you need to decide whether it can happen again.
That last one matters more than it looks. A vague resolution is not a resolution; it is a fault you cannot plan around. The businesses that get burned twice are usually the ones who were never told what happened the first time.
The Honest Limits
An argument you cannot state the limits of is a sales pitch. So, plainly:
| Nobody controls | What that means in practice | What a good provider still does |
|---|---|---|
| The NBN access network | The physical connection into your building belongs to NBN Co. A fault in it is a fault nobody else can fix. | Owns the ticket end to end, chases it, keeps you informed, and does not make you the messenger. |
| Mobile carriers | If a caller’s mobile network is having a bad day, that is their network. | Can prove it is not your system, which ends the argument rather than prolonging it. |
| Your internal network | Your router, your switch, your Wi-Fi, your firewall’s SIP ALG setting. | Diagnoses it accurately and tells you exactly what to change, rather than declaring it out of scope. |
| Handset manufacture | VOCPhone supplies certified handsets rather than manufacturing them. That is a supply relationship like anybody’s. | Certifies and provisions them, so the provisioning and platform behaviour are still one company’s responsibility. |
The point is not omnipotence
No provider owns everything, and one that claims to is telling you something about their sales culture. The point is how much of the path is inside one company, and who owns the ticket for the part that is not. Those are two different questions and both matter. A provider that owns the network and the platform is answering yes to the first and can still be judged on the second.
What 15 Years and 99.99% Actually Describe
Two numbers get quoted constantly and neither is self-explanatory, so here is what they mean operationally.
15+
Years operating in the Australian market
99.99%
Network uptime
24/7
Australian human support, not a chatbot and a queue
1
Company that owns the network, the platform and the support desk
Fifteen years is not a loyalty badge. In telecommunications it means having been through the events: the NBN migration, the 3G shutdown, the copper retirement, several major national outages, and the arrival of an entirely new regulatory regime around emergency calling and scam prevention. Institutional memory of those is why the twelfth incident is handled better than the first.
99.99% is roughly 52 minutes of unavailability across a year. The useful thing about the figure is not the number itself but that it belongs to a network somebody actually operates. A reseller quoting an uptime figure is quoting somebody else’s, usually from a wholesale agreement they cannot show you.
The question that tests it
Ask any provider: “When you say 99.9-something percent uptime, whose network is that measured on, and can I see your outage register?” Since 30 June 2026, Australian providers are required to maintain public outage registers recording start time, locations, service types, cause and restoration. The claim is now checkable, which makes it worth checking.
The Platform Still Has to Be Good
All the accountability in the world does not rescue mediocre software. A well-supported bad phone system is still a bad phone system. So here is the other half, framed as tests to apply to any provider including this one.
| Test | Why it decides things | VOCPhone |
|---|---|---|
| Apps on every platform, free | Most staff will live in the app rather than on a handset. If it is an afterthought, so are they. | ✓ Windows, Mac, iOS, Android — included |
| AI included, not upsold | Whether AI sits in a premium tier moves total cost more than any other single answer. | ✓ AI phone agents with Australian accents, transcription and summaries |
| Real integrations and open APIs | A phone system that cannot reach your CRM gets replaced within three years. | ✓ Salesforce, HubSpot, Zoho, Xero, Monday and 1000+ apps |
| Number portability | If leaving is hard, staying is not a choice. | ✓ Bring your existing numbers |
| Transparent pricing | Per-seat pricing you can read on a page beats a quote you have to negotiate for. | ✓ Published plans and pricing, with a price guarantee |
| Data in Australia | Recordings and AI transcripts are personal information. Jurisdiction is a compliance question. | ✓ Australian hosted |
Six Questions That Reveal the Answer
You cannot tell a reseller from an operator by looking at a website. You can tell in about four minutes on a phone call, by asking these.
- “Do you own the network my calls travel over, or do you buy it wholesale?” A direct answer either way is fine. Hesitation, or a redirect to how good their support is, is the answer.
- “If I report a fault you cannot fix, who do you escalate to, and what visibility do I get?” This locates the escalation ceiling precisely. Ask it and then be quiet.
- “Can your support team see call detail records and network path data at the same time?” This is the capability that resolves intermittent faults. Either they have it or they do not.
- “When you quote uptime, whose network is that measured on, and can I see the outage register?” Registers are now mandatory. A provider who cannot point you at theirs is telling you something.
- “Which country is my call data stored and processed in?” A good answer names a country. “In the cloud” is not an answer.
- “What is the exit process and how long does porting take?” Asked before signing, the answer is honest. Asked after, it is a negotiation.
When a Reseller Is the Right Choice
Sometimes it genuinely is, and pretending otherwise would be dishonest.
Your IT provider bundles it
If your managed service provider already runs everything else and does it well, adding phones to that single relationship may beat splitting it — because for you, they are the single point of accountability.
You need something very specific
A niche vertical platform with features nobody else builds. Take the feature and accept the support structure that comes with it.
Phones genuinely do not matter much
Three users, low volume, no customer-facing calls. Buy on price. The accountability argument is real but it is not free, and it is not worth much to you.
You are multinational
An Australian operator is the right answer for Australian operations. Twelve countries is a different problem with different trade-offs.
For everyone else — an Australian business between five and a couple of hundred people, where the phone ringing is connected to revenue — the accountability argument is the one that pays, and it pays on the day you were not planning for.
How to Decide
Three questions, in order, and they take about twenty minutes.
1. What does an hour of downtime cost?
Not precisely — roughly. Calls per hour, conversion rate, average value. If the number is meaningful, accountability is worth paying for. If it is negligible, buy on price and stop reading.
2. Who currently does the chasing?
When something breaks now, who in your business becomes the go-between? What is their time worth, and how often does it happen? That is the cost you are actually deciding about.
3. Ask the six questions
Of every provider on your shortlist, including your current one. The answers will sort the list faster than any feature comparison, and they cost nothing but a phone call.
If you want the network ownership argument in more depth, we own the network covers it, and redundancy and single-network risk covers what to do about the parts nobody controls.
Frequently Asked Questions
What is the difference between a phone system reseller and a network operator?
Authority, mostly, and it is invisible until something breaks. The Australian market is layered: a wholesale platform vendor writes the software and sells white-labelled access, a network operator carries the calls and holds the interconnects, a data centre runs the compute, and a reseller puts a logo on the result and answers the support line. A reseller can do plenty well - onboarding, account management, call flows, billing, being genuinely pleasant to deal with - and a good small reseller can beat a bad large operator on service. What a reseller cannot do is fix anything it cannot see. When a fault sits in the platform or the network, the reseller lodges a ticket with somebody else and then waits alongside you, knowing exactly as much as you do. An operator that owns its own network and platform can investigate routing, capacity, interconnects and platform behaviour directly. The practical test is not competence, it is where the point sits beyond which your provider can only relay rather than act.
What is an escalation ceiling and why does it matter?
It is the point beyond which a provider stops being able to act and can only pass your problem to someone else. Below the ceiling they investigate, adjust and fix. Above it they lodge a ticket and wait, with no visibility, no priority and no ability to escalate beyond a queue position. For a reseller of a wholesale platform, that ceiling sits fairly low - roughly at the boundary of your account configuration. A reseller with its own support engineering has a higher ceiling and can often tell you accurately what is wrong, but still cannot fix it. An operator of its own network and platform has a ceiling that only genuinely external causes reach: an access network fault, an upstream carrier, or something inside your own building. The reason it matters is that nothing in a sales process reveals where the line sits, and you find out during an incident. The question that locates it precisely is: if I report a fault you cannot fix, who do you escalate to and what visibility do I get?
How does owning the network change fault diagnosis?
In four specific ways, each mapping to a class of fault that is otherwise very difficult to resolve. First, both halves appear in one view - call detail records alongside network path data - and most intermittent voice faults are only visible in the correlation between the two, which is exactly why split ownership makes them so hard to catch. Second, routing becomes a decision rather than a request: if a path is degraded, the party who owns it can change it, while a reseller can only ask and wait for somebody else's change window. Third, priority means something real, because your fault competes with other customers of the same company rather than with the entire wholesale customer base of a platform vendor that has never heard of you. Fourth, you get a genuine cause rather than the phrase 'an upstream issue' - which matters because a vague resolution is not a resolution, it is a fault you cannot plan around, and the businesses that get burned twice are usually the ones never told what happened the first time.
Does any provider control the whole path from caller to callee?
No, and any provider suggesting otherwise is telling you something about their sales culture rather than their engineering. Four things sit outside everyone's control. The NBN access network - the physical connection into your building - belongs to NBN Co, and a fault in it is a fault nobody else can fix. Mobile carriers own their own networks, so if a caller's mobile network is having a bad day, that is their network. Your internal environment is yours: the router, the switch, the Wi-Fi and the firewall's SIP ALG setting. And handsets are manufactured by hardware companies; VOCPhone supplies certified handsets rather than building them. What differs between providers is not omnipotence but two separate things: how much of the path sits inside one company, and who owns the ticket for the part that does not. A provider that owns its network and platform answers yes to the first, and should still be judged on whether it chases the rest rather than making you the messenger.
What does 99.99% uptime actually mean for a business phone system?
Roughly 52 minutes of unavailability across a full year, which is a useful benchmark but not the most informative part of the claim. The more important question is whose network the figure describes. A provider that owns and operates its network is quoting its own measurement of its own infrastructure. A reseller quoting an uptime figure is quoting somebody else's number, usually drawn from a wholesale agreement they cannot show you, for a network they cannot see. Since 30 June 2026 the claim has become checkable: Australian providers must maintain public outage registers recording when an outage started, which locations and service types were affected, the cause and the restoration time. So the question worth asking is not just what your uptime is, but whose network that is measured on and where your outage register is published. A provider who cannot immediately point you at theirs has answered a different question than the one you asked.
What should I ask a phone provider before signing?
Six questions, and they take about four minutes on a call. Do you own the network my calls travel over, or do you buy it wholesale - a direct answer either way is fine, while hesitation or a pivot to how good the support team is constitutes the answer. If I report a fault you cannot fix, who do you escalate to and what visibility do I get - ask it, then stay quiet and let them answer. Can your support team see call detail records and network path data at the same time, which is the specific capability that resolves intermittent faults. When you quote uptime, whose network is that measured on and can I see your outage register, since registers are now mandatory. Which country is my call data stored and processed in - a good answer names a country, and 'in the cloud' is not an answer. And what is the exit process and how long does porting take, which is honest when asked before signing and a negotiation when asked after.
Is buying from a reseller ever the right decision?
Yes, in four situations. If your managed service provider already runs your IT well and bundles phones into that relationship, adding voice to a single supplier you trust may beat splitting it, because for practical purposes they are your single point of accountability. If you need something genuinely niche - a vertical platform with features nobody else builds - take the feature and accept the support structure attached to it. If phones genuinely do not matter much to your business, meaning three users, low volume and no customer-facing calls, buy on price, because the accountability argument is real but it is not free and it is not worth much to you. And if you are multinational, an Australian operator is the right answer for Australian operations but twelve countries is a different problem with different trade-offs. For the businesses in between - Australian, five to a couple of hundred people, where the phone ringing is connected to revenue - accountability is the argument that pays, and it pays on the day you were not planning for.
