Coverage and Reachability Are Different Words
Almost every conversation about regional connectivity conflates two things, and separating them changes what you spend money on.
Coverage
Whether a mobile network puts a usable signal on a particular patch of ground. Determined by tower placement, spectrum and terrain. Decided by carriers and regulators. You have essentially no control over it.
Reachability
Whether a person who rings your business number ends up speaking to someone. Determined by how your number is hosted, how calls route, and what happens when nobody picks up. You control all of it.
Every headline about roaming, satellites and coverage obligations is about the first column. Every dollar of revenue you lose to an unanswered phone is in the second. It is entirely possible — and common — for a business with genuinely terrible coverage to answer every call it receives, and for a business in full signal to lose a third of them.
The question is not “how many bars do we have?” It is “when someone rang us at 4:40pm on Thursday, what happened?” Only one of those has an answer you can change this month.
So the rest of this is organised around failures rather than technologies. Find yours, apply the fix, and read the policy section at the end only if you want to know whether waiting would have been smarter. It would not have been.
The Four Failures, and the Fix for Each
“No coverage” describes four different situations that share nothing except a symptom.
| Failure | Symptom | Fix | Cost to fix |
|---|---|---|---|
| 1. No signal at the premises | The site has poor or no mobile signal but does have internet of some kind | Run voice over the internet connection. The mobile network stops being relevant | Low — usually a configuration change |
| 2. The indoor blackspot | Signal in the yard, nothing in the shed, workshop or cool room where people work | Wi-Fi coverage plus a phone app that calls over it | Low — one or two access points |
| 3. Genuinely out of range | Staff in vehicles or on properties with no network at all | Nothing carries voice here today. Capture what they miss; use messaging for safety | Low, but it is mitigation rather than a fix |
| 4. The link itself drops | Weather, power, a cut cable, a provider outage | Layered failover, configured in advance and tested | Moderate — a second path costs money |
Which one do you have?
Most regional businesses assume they have failure 3, the hard one, when they actually have failures 1 and 2, the cheap ones. Ten minutes with a phone, walking your own site, settles it. Businesses spend years working around a shed with no signal without ever discovering that an access point on the wall would have ended it.
Failure 1: No Signal at the Premises
The most common and the most misunderstood. A great many regional operators believe that poor mobile signal means they cannot have a proper phone system. The opposite is closer to the truth.
A cloud phone system runs over an internet connection. It does not use the mobile network at all. If the site has fixed broadband, fixed wireless or satellite internet, desk handsets and computer applications work exactly as they do in a city office, and the mobile signal outside is irrelevant to them.
What voice actually needs is modest and specific:
~100 kbps
Per call, each direction
<30 ms
Jitter — the one that matters
<1%
Packet loss
<150 ms
One-way latency
Ten calls at once is roughly one megabit each way. The constraint is almost never bandwidth. It is consistency — a fast connection with erratic timing sounds worse on a call than a slow, steady one, which is why speed test results are close to useless as a predictor of call quality.
The satellite distinction that catches everyone
“Satellite internet” is two entirely different products for voice. Low earth orbit services operate close enough that latency lands in the same range as terrestrial links, and voice works acceptably — the practical limitation is brief interruptions during satellite handover. Geostationary services sit around 36,000 km up, and the round trip imposes a delay that no equipment, codec or configuration can remove: the result is the talk-over effect familiar from old satellite phone calls. If a site is on satellite internet, establish which type before assuming voice will run over it. This one fact saves more wasted deployments than anything else on this page.
The other half of failure 1 is where the number lives. A number tied to a physical line at your premises dies with the premises. A number that terminates in the network can be pointed somewhere else in seconds, from a phone, standing in a paddock. That is a property of how the service is provisioned, not an add-on feature, and it is worth confirming you have it before you need it — what a cloud phone system actually is covers the distinction.
Failure 2: The Shed
Signal in the yard. Nothing in the shed. The shed is where the work happens. This is the single most common regional complaint and the cheapest thing on this page to fix.
A steel building is close to a perfect radio shield. No amount of network investment changes that — a tower two kilometres away and a tower two hundred metres away both fail to get through a steel wall in the same way. Coverage obligations framed around outdoor coverage will not fix it either, which is worth noting given how the 2027 proposal is currently framed.
What does fix it: put Wi-Fi in the building, and use a phone system app that treats Wi-Fi as just another internet connection.
- Get internet to the building. Cable if the run is short, a point-to-point wireless bridge if it is not. Ordinary outdoor wireless bridge kit covers a few hundred metres reliably and is not expensive.
- Put an access point inside. One is usually enough for a shed; a large workshop with internal walls may need two.
- Install the phone system app on the phones people already carry. No second device, no new number.
- Test the handover. Walk from the yard into the shed on a live call and confirm it survives the transition from mobile data to Wi-Fi.
Why this beats a mobile repeater
Repeaters and boosters are heavily regulated in Australia and must be authorised for use on a specific network — unauthorised devices interfere with the network and are illegal to operate. Wi-Fi calling through your phone system app sidesteps the entire question: it is not touching the mobile network, so there is nothing to authorise. It also gives you a business number in the shed instead of an assortment of personal mobiles, which matters for records, continuity and what happens when someone leaves.
Failure 3: Genuinely Out of Range
The honest section. If your people are in a vehicle or on a property with no network of any kind, nothing available today carries a voice call to them. Not satellite messaging, not roaming if it is eventually declared, not Wi-Fi. That is the current state of the technology and no amount of product selection changes it.
What you can do is make sure the business does not lose anything because of it.
Nothing is dropped
Every missed call logged and visible to someone at base, voicemail transcribed to text, and a callback list that a human works through. The call is not lost, only delayed.
An automatic reply
An SMS that fires on a missed call — "we're on site and will call back today" — converts an unanswered call from a dead end into a held position.
Base answers, not the ute
The main number rings the office first and the field second, not the reverse. Coverage-dependent people should never be the first line on a customer-facing number.
Satellite for safety
Direct-to-mobile texting means a worker in a dead zone can confirm they are alright. That is a genuine improvement in a category that had nothing.
The design principle is simple: never make an out-of-range person the first ring on a number a customer calls. It is the mistake almost every small regional business makes, because the owner's mobile was the original business number and it never got restructured. The pattern, and what it costs, is the same one documented in what missed calls cost tradies.
Failure 4: The Link Itself Drops
Weather takes the power out. A contractor cuts a cable. The provider has an outage. The site is fine; the path to it is not.
The fix is layered failover, and the important word is configured. Failover that requires a person to notice and act is not failover, because the person is usually the one dealing with the actual emergency.
| Layer | What it is | Switches when |
|---|---|---|
| 1 | Primary internet at the site | Normal operation |
| 2 | A second path on a different technology — mobile broadband if the primary is fixed, or the reverse | Automatically, on loss of the primary |
| 3 | Staff mobile apps over mobile data, anywhere | Automatically, if the site is unreachable entirely |
| 4 | Divert the number to another site, a mobile, or an answering arrangement | Automatically, after a set time with no answer |
Layer 2 deserves a note: the second path must be a different technology, not a second service. Two providers reselling the same infrastructure fail together, which is a lesson Australian businesses have relearned during every major outage. The broader design question is covered in business phone redundancy.
The test nobody runs
Pull the primary internet out during business hours, deliberately, on a quiet Tuesday, and ring your own number. Almost every business believes failover works. A small minority have watched it work. The usual findings: the backup was never actually provisioned, the divert points at a number that changed two years ago, or the whole arrangement rests on one person noticing. Twenty minutes on a Tuesday beats finding out during a storm.
What Satellite Actually Gives You
Two different satellite stories get told as one, so here they are separately.
| Satellite broadband to a site | Direct-to-mobile satellite | |
|---|---|---|
| What it is | A dish at the premises providing internet | An ordinary handset talking to a satellite with no tower involved |
| Carries voice calls | ✓ Yes on low earth orbit; ✗ poorly on geostationary | ✗ No |
| Carries data | ✓ Yes | ✗ No general mobile data |
| Carries text | ✓ Via the internet | ✓ Yes — this is its purpose |
| Works indoors | ✓ The dish is outside; everything inside works normally | ✗ Needs open sky |
| Use it for | Running the site, including the phone system | Safety and coordination for people beyond any network |
Direct-to-mobile launched in Australia during 2026, initially for text and then extending to a small curated set of satellite-optimised apps, across mainland Australia and Tasmania and excluding the Australian Radio Quiet Zone in Western Australia. Compatible handsets only, outdoors, with a clear view of the sky. A second carrier service has been indicated for 2026 subject to approvals.
Tell your field staff this explicitly
Satellite messaging does not place a voice call to Triple Zero. It is easy for someone to conclude that a phone showing a satellite indicator can call for help, and that assumption is dangerous in exactly the environments where the feature is most used. Say it out loud during induction. Triple Zero from a cloud phone covers the emergency calling rules that do apply.
The Half Day That Replaces the Coverage Map
Coverage maps model outdoor signal. They know nothing about your buildings, your terrain at ground level, or the time of day your local tower gets busy. Half a day of walking your own site produces a document worth more than any map, and it is the thing to hand any provider you are talking to.
- Walk it with a phone. Place a real call from every spot where a person works — office, shed, workshop, cool room, yard, the far gate. Note the address of each failure, not just that it failed.
- Repeat on all three networks. Borrow handsets for a day. This separates "there is no coverage here" from "our current provider has no coverage here", and it is frequently the second.
- Test Wi-Fi calling from the same spots. Using the phone system app over site Wi-Fi. Count how many of your blackspots disappear — usually most of the indoor ones.
- Kill the internet. Deliberately, during business hours, and watch an inbound call. This is the test that finds the broken assumption.
- Ring yourself after hours. 5:05pm, 7pm, Saturday morning. A large share of regional enquiries arrive outside business hours and go straight into whatever you have not configured.
What you end up with
A one-page map of where your business is genuinely unreachable, separated from where it merely feels unreachable. In most cases the genuinely-unreachable list is much shorter than expected, and the remainder is solved with an access point, a routing change and a second internet path. That is a week of work, not a policy outcome.
And the Policy, Briefly
For completeness, since it is being widely reported and you will be asked about it.
| What happened | What it means for you |
|---|---|
| The ACCC launched a mobile services inquiry in August 2026, running about twelve months, considering whether to declare wholesale mobile services including domestic roaming and a radio access network service | A declaration would let your handset use whichever network is present rather than showing no service. But this is an inquiry into whether to decide — and if a declaration follows, terms, systems and launches follow that. Years, not months |
| A discussion paper is due shortly, and forums will be held in regional and remote Australia | If coverage genuinely constrains your business, this is a rare chance to say so on the record. Worth an hour |
| A Universal Outdoor Mobile Obligation is proposed for 2027, and the ACCC weighed it when deciding to proceed | Note the word outdoor. It will not put signal in your shed. Failure 2 remains yours to solve |
| The ACCC declined to declare roaming after its 2017 inquiry, on the basis that mandated access reduces the incentive to invest in regional infrastructure | The same argument is being made again. Plan as though nothing changes and treat a declaration as upside |
The conclusion is not that policy does not matter. It is that the five things in this article pay for themselves under every possible outcome, and none of them is made redundant by a roaming declaration. A business that spends the next twelve months becoming reachable will be better off than one that spends it following a consultation — and it will still get the benefit of the consultation if it lands.
If you want the wider connectivity picture, how to read mobile coverage maps covers what the published categories actually promise, and the 6G roadmap covers the longer arc including satellite integration into future standards.
Frequently Asked Questions
My site has almost no mobile signal. Can I have a proper business phone system?
Almost certainly yes, and this is the most common misunderstanding in regional telephony. A cloud phone system runs over an internet connection and does not touch the mobile network at all. If the site has fixed broadband, fixed wireless or satellite internet, desk handsets and computer applications work exactly as they would in a city office, and the mobile signal outside becomes irrelevant to them. What voice needs is modest: roughly 100 kilobits per second per call in each direction, so ten simultaneous calls is about one megabit each way. The real constraint is almost never bandwidth but consistency — jitter under about 30 milliseconds, packet loss under one per cent, one-way latency under about 150 milliseconds. A fast connection with erratic timing sounds worse than a slow steady one, which is why speed test results are close to useless as a predictor of call quality. The second half of the answer is where your number lives: a number tied to a physical line at the premises dies with the premises, whereas a number terminating in the network can be redirected in seconds from a phone while you stand in a paddock.
Is satellite internet good enough for phone calls?
It depends entirely on which kind, and this distinction catches more people than any other. Satellite internet is two different products for voice purposes. Low earth orbit services operate close enough to the ground that latency lands in the same range as terrestrial connections, and voice works acceptably — the practical limitation is brief interruptions during satellite handover rather than the delay itself. Geostationary services sit around 36,000 kilometres up, and the round trip imposes a delay that no equipment, codec or configuration can remove, producing the talk-over effect familiar from old satellite phone calls where both parties keep interrupting each other. If a site is on satellite internet, establish which type it is before assuming voice will run over it. This single fact prevents more wasted deployments than anything else in regional telephony, because the two products are marketed with the same word and behave nothing alike once a conversation is involved.
Can I make calls over satellite direct to my mobile?
No. Direct-to-mobile satellite launched in Australia during 2026, initially carrying text messages and then extending to a small curated set of satellite-optimised applications covering messaging, navigation and weather, across mainland Australia and Tasmania and excluding the Australian Radio Quiet Zone in Western Australia. A second carrier service has been indicated for 2026 subject to regulatory approvals. The limits are firm: no voice calls, no general mobile data, no MMS or ordinary internet messaging apps, compatible handsets only, and you must be outdoors with a clear view of the sky — it does not work inside a shed, in a vehicle or under tree cover. Treat it as a genuine improvement in a category that previously had nothing: a worker in a dead zone can confirm they are alright, a driver can report a delay, a crew can be redirected. One thing to say out loud during induction: satellite messaging does not place a voice call to Triple Zero. It is easy for someone to see a satellite indicator on their phone and conclude they can call for help, and that assumption is most dangerous in exactly the environments where the feature gets used.
There is signal in the yard but none in our shed. What fixes that?
Wi-Fi, and it is the cheapest fix on the list. A steel building is close to a perfect radio shield, so no amount of network investment changes it — a tower two hundred metres away fails to get through a steel wall in the same way one two kilometres away does. Coverage obligations framed around outdoor coverage will not address it either. The fix is four steps. Get internet to the building, using cable if the run is short or a point-to-point wireless bridge if it is not, since ordinary outdoor bridge kit covers a few hundred metres reliably and is inexpensive. Put an access point inside — one is usually enough for a shed, while a large workshop with internal walls may want two. Install the phone system app on the phones people already carry, so there is no second device and no new number. Then walk from the yard into the shed on a live call and confirm it survives the handover. This is preferable to a mobile repeater: boosters are heavily regulated in Australia and must be authorised for use on a specific network, whereas Wi-Fi calling through your phone system app never touches the mobile network so there is nothing to authorise. It also puts a business number in the shed rather than an assortment of personal mobiles.
What can I do about staff who are genuinely out of range?
Nothing available today carries a voice call to someone with no network of any kind — not satellite messaging, not roaming if it is eventually declared, not Wi-Fi. What you can do is make sure the business loses nothing because of it. Log every missed call somewhere a person at base can see it, transcribe voicemail to text so it can be read and actioned quickly, and work a callback list so the call is delayed rather than lost. Fire an automatic SMS on a missed call — something as simple as a note that you are on site and will call back today converts a dead end into a held position. Use direct-to-mobile satellite messaging for safety check-ins where it is available. And apply the one design principle that matters most: never make an out-of-range person the first ring on a number a customer calls. The main number should ring the office first and the field second, not the reverse. Almost every small regional business gets this wrong, because the owner's mobile was the original business number and nobody ever restructured it.
How do I test whether my failover actually works?
Pull the primary internet out during business hours, deliberately, on a quiet Tuesday, and ring your own number while watching what happens. Almost every business believes failover works and a small minority have actually watched it work. The usual discoveries are that the backup service was never really provisioned, that the divert points at a number which changed two years ago, or that the entire arrangement depends on one specific person noticing and doing something. Twenty minutes on a Tuesday is a great deal better than finding out during a storm. While you are at it, check that your second path is a genuinely different technology rather than a second service — mobile broadband behind a fixed primary, or the reverse. Two providers reselling the same underlying infrastructure fail together, which Australian businesses have relearned during every major outage of the last few years. A properly layered arrangement runs primary link, then a different-technology second path, then staff mobile apps over mobile data anywhere, then an automatic divert to another site or number after a set time with no answer — all of it configured in advance so no human decision is required at the moment it matters.
Should I wait for the ACCC roaming inquiry before making changes?
No. The ACCC launched a mobile services inquiry in August 2026, running approximately twelve months, examining regional coverage and whether to declare wholesale mobile services including domestic roaming and a radio access network service. A discussion paper is due shortly and forums will be held across regional and remote Australia, which is a genuine opportunity to put your experience on the record if coverage constrains your business. But be realistic about what the timeline produces: this is an inquiry into whether to decide, and if a declaration follows, terms and pricing must be settled, systems built and services launched. The distance to a practical difference is measured in years. It is also worth knowing that the ACCC ran a similar inquiry in 2017 and declined to declare, reasoning that mandated access at regulated terms reduces the return on building regional infrastructure. The same argument is being made again, and the proposed Universal Outdoor Mobile Obligation for 2027 complicates the roaming case further. Plan as though nothing changes and treat a declaration as upside — every practical step described here pays for itself under either outcome, and none of them is made redundant if roaming is eventually mandated.
