The Numbers That Landed This Month
Three separate sets of figures, from three organisations, none of which was written for a business buying phone services. Together they describe the cost base your invoice eventually comes out of.
| What | When | The headline number |
|---|---|---|
| Telstra full-year results (FY26) | 13 August 2026 | Total income $23.405 billion; net profit after tax $2.406 billion; total dividend 21.0 cents; $1 billion buyback |
| ACMA final spectrum valuation | Settled May 2026, applications opened 18 June 2026 | $7.32 billion to renew mobile and fixed wireless licences expiring 2028–2032 |
| NBN Co fibre eligibility change | From July 2026 | Roughly 600,000 fibre-to-the-curb premises now eligible for full fibre without ordering a high-speed tier |
What Telstra’s FY26 Actually Says
Telstra reported its full-year results for the year ended 30 June 2026 on 13 August 2026. The figures, as published:
$23.405b
Total income, down 0.9%
$2.406b
Net profit after tax, up 2.7%
$8.2b
EBITDAaL
21.0c
Total dividend per share
Operating profit rose 1.7% to $4 billion and earnings per share rose 5.3% to 19.9 cents. Cash earnings grew 11.6% to $2.9 billion with cash earnings per share up 13.8% to 25.5 cents. A final dividend of 10.5 cents was declared, and a $1 billion share buyback announced. Mobile service revenue grew 3.2% and the company added more than 270,000 mobile services over the year. All of it sits under the Connected Future 30 strategy, whose stated ambition is to be the number one choice for connectivity in Australia.
What a carrier result does and does not tell you
It tells you the direction of the largest player's cost discipline and where it is putting capital — in this case network resilience, 5G and digital infrastructure. It does not predict retail pricing, and reading it that way is a common mistake. Total income fell while profit rose, which is a cost story rather than a price story. A carrier can post a strong result while its retail prices go up, down or nowhere, because the two are only loosely connected in a market this concentrated.
The genuinely useful signal for a business is subtler: 270,000 added mobile services in a year is a market where the incumbent is still winning volume. That is not the profile of a market under heavy price pressure, and it is a reasonable input into how hard you should expect to have to negotiate.
The $7.32 Billion Bill
This is the item with the longest tail and the least coverage, and it is worth understanding because it sets the input cost of Australian mobile networks for the next fifteen years.
Mobile spectrum is licensed, not owned. A large tranche of Australian licences expires between 2028 and 2032, and rather than re-auction them, the ACMA set a renewal price. After a preliminary figure of $7.34 billion in December 2025, the final valuation landed at $7.32 billion across seven bands — 700MHz, 850MHz, 1800MHz, 2GHz, 2.3GHz, 2.5GHz and 3.4GHz. Those licences underpin fixed wireless broadband and more than 30 million mobile services in Australia.
| Detail | What it means |
|---|---|
| Who pays | Telstra, Optus, TPG and NBN Co — the four holders of the expiring licences |
| What it covers | The frequencies carrying 4G and 5G mobile, and NBN Co's fixed wireless service to regional premises |
| Timetable | Renewal applications opened 18 June 2026, starting with the 850MHz and 1800MHz bands, with nine-month application windows |
| How the price was set | Per megahertz per head of population, band by band. Lower bands were revised down slightly and mid-band prices were revised up |
| When it hits accounts | As licences expire from 2028 onward, not now. This is a cost being planned for rather than a cost being incurred |
Why a business should care about a 2028 licence renewal
Because capital planning happens years ahead of the payment. A carrier that knows it owes several billion dollars in 2028 makes different decisions in 2026 about what to build, where to build it and what to charge. That is the actual transmission mechanism, and it is far more important than the headline figure. It is also the reason the industry fought the valuation as hard as it did.
The Argument About Whether It Is Too High
The industry position is not subtle and it deserves to be stated properly rather than dismissed, because it may well be right.
Telstra argued the ACMA had overvalued the spectrum by $3.3 billion and urged Treasury to cap the total at $3.9 billion — roughly half the figure the regulator settled on. The Australian Telecommunications Alliance warned that higher spectrum costs would mean “less investment, or higher prices or both”. The telcos were unable to overturn the valuation.
| The industry argument | The counter-argument |
|---|---|
| Spectrum is an input cost, and inflated input costs are recovered from customers or taken out of network investment | Spectrum is a scarce public resource and underpricing it transfers value from the public to shareholders rather than to customers |
| Regional coverage is the first thing cut when capital is tight, and regional coverage is already the acknowledged weak point | Regional coverage has been the acknowledged weak point through periods of both high and low spectrum pricing, which weakens the causal link |
| A price set administratively rather than by auction is a guess, and this guess is a large one | An auction on expiring incumbent-held licences has its own well-documented problems, which is why the renewal mechanism exists |
We have no position on who is right, and businesses reading this should be sceptical of anyone who does. What matters practically is that the number is settled, it is large, it lands from 2028, and every carrier now has it in their planning assumptions. Whether it was the correct number is an argument for economists; that it is a number carriers must fund is the fact you can actually use.
What NBN Co Changed, and Who It Helps
Two NBN Co changes matter to businesses, and one of them removes an obstacle that has frustrated a lot of them.
Fibre to the curb, unlocked
From July 2026 NBN Co removed the requirement to order a high-speed tier in order to be eligible to upgrade from fibre to the curb to full fibre. Roughly 600,000 single dwelling premises become eligible without buying a faster plan they may not want.
Speed uplifts already applied
Under the Accelerate Great programme, wholesale speeds were lifted on eligible FTTP and HFC services: 100/20 and 100/40 to 500/50, 250/25 to 750/50, and 1000/50 to 1000/100. On the business side, Business 100 at 100/40 was uplifted to 250/100.
Consumption keeps climbing
Average monthly data per premises rose from 460GB in June 2024 to 508GB in June 2025, up more than ten per cent in a single year. That is the pressure behind the uplifts, and it does not look like slowing.
Why upload is the number that matters
For voice, the download figure is largely decorative. Upload headroom is what determines call quality when eight people are on calls and someone starts a video meeting. The uplift from 40 to 100 on business services is the meaningful part of that table.
The practical action
If your business is on fibre to the curb, check your eligibility for a full fibre upgrade now. The condition that previously blocked a lot of businesses — having to commit to a speed tier above what they needed — has been removed. And if you are still on a business plan with 40Mbps upload, check whether an uplift has been applied to your service, because a surprising number of businesses are entitled to a faster service they have never been told about.
The Spectrum Story Underneath the Spectrum Story
One more item, because it is the piece most likely to change the picture materially within five years and it is getting almost no attention from business media.
The ACMA is reallocating 2GHz radiofrequency spectrum for mobile satellite services, with an auction expected later in 2026. Meanwhile the ACCC is examining the long-term role of low earth orbit satellite services — Starlink, which has partnered with both Telstra and Optus, and Amazon's Project Kuiper, favoured by NBN Co — as part of the mobile services inquiry it launched on 5 August.
What it might change, stated cautiously
Direct-to-device satellite has the potential to alter what “no coverage” means in Australia, which would affect regional pricing, the roaming debate and the value of terrestrial spectrum all at once. It is also early, capacity-constrained and not a substitute for a terrestrial network for anything volume-dependent. Treat it as a real development to watch and not as a reason to defer a decision about connectivity you need this year.
The Chain From Auction to Invoice
Here is the honest transmission path, with the lag at each step, because the gap between “spectrum got expensive” and “my bill went up” is where most commentary quietly stops.
| Step | What happens | Lag |
|---|---|---|
| 1. Spectrum price set | $7.32 billion settled for licences expiring 2028–2032 | Done |
| 2. Capital plans adjust | Carriers rebalance what to build and where, knowing the payment is coming | Immediate and invisible |
| 3. Licence payments | Cash goes out as each licence renews | 2028 onward |
| 4. Wholesale rates | Input costs feed into what wholesale customers and resellers pay | Slow, and partly absorbed |
| 5. Retail pricing | What appears on an invoice | Slowest, least predictable, and shaped more by competition than by cost |
The conclusion this leads to
Nothing in the spectrum decision will change your bill this year, and treating it as though it will is how businesses get talked into signing longer terms. “Lock in now before prices rise” is a sales line, not a forecast. The costs are real and they are years out, and by then your contract will have come up for renewal at least once regardless.
What Actually Drives What You Pay
If spectrum, carrier results and wholesale economics are not what moves your invoice in the next twelve months, what is? Five things, in order of how much money is usually sitting in each.
| Driver | Typical size | What to do about it |
|---|---|---|
| 1. What you are paying for twice | Usually the largest single item | Conferencing, SMS tools, answering services and call recording bought separately from a platform that already includes them. Audit the card statement, not the phone bill |
| 2. Seats you are not using | Frequently 10–20% of a per-user bill | Licences for departed staff, duplicate accounts, and extensions on desks nobody sits at. Nobody audits this and providers rarely volunteer it |
| 3. Your contract terms | Large, and entirely negotiable | Term length, exit costs, price review clauses and what happens at renewal. This is where the real money is decided and where least attention goes |
| 4. Call spend versus plan structure | Smaller than most expect | Unlimited call packs versus metered. Worth checking once, then leaving alone. It is rarely the problem |
| 5. Actual wholesale input costs | Real, slow, largely outside your control | Spectrum, backhaul, NBN wholesale. Worth understanding so you can tell a genuine cost pass-through from an opportunistic increase |
The test that separates the two kinds of price rise
When a provider raises your price and cites industry costs, ask one question: is this applied to all customers on published rates, or just to mine? A genuine input-cost increase shows up in the published price list. An opportunistic one arrives as an individual letter and is negotiable. That single question resolves most of them, and asking it politely costs nothing.
How to Buy Against This
Practical positioning for a period where input costs are rising slowly and the timing is uncertain.
Prefer published pricing
A provider with a public price list is constrained by it. Bespoke pricing feels like a win and removes your reference point at renewal, which is exactly when you need one.
Be sceptical of long terms sold on price fear
Five years to hedge against a cost that lands in 2028 is a trade worth examining carefully. The provider is not the one taking the risk in that arrangement.
Read the price review clause
Almost every contract has one and almost nobody reads it. Find out what your provider is permitted to raise, by how much, and with what notice. Then decide whether the term length is still attractive.
Audit before you renegotiate
Duplicates and unused seats first, price per seat second. Negotiating a discount on services you should have cancelled is a poor use of the leverage you only get once a term.
The summary
Telstra had a solid year on falling revenue and rising cash. The spectrum bill is settled at $7.32 billion and lands from 2028, over strenuous industry objection. NBN Co has removed a real obstacle to full fibre for about 600,000 premises. None of that changes your invoice this year. What changes your invoice is the duplication in your stack, the seats you are not using, and the contract clause you have never read — and all three of those are entirely within your control.
Related reading: the July 2026 price rises for the retail side, full fibre upgrades and your business phone for the NBN change in detail, and what a business phone system costs for the pricing structure underneath all of it.