From CommsDay of 14 August 2026
We have already sent you, yesterday, our Story of the Week. However, also in yesterday’s issue of CommsDay, was an analysis of the Telstra annual results that were announced on Thursday and which were well covered in the financial and general media as well as in CommsDay. The interesting aspect of Grahame Lynch’s analysis is that it contrasts the general perception of Telsta – what it is and what it should do – that reflects the past, with what is happening within the organisation now. Telstra has evolved to address changing sector and economic realities and opportunities. It is definitely not the organisation that many of us once worked for. I think we recognise that, but the expectations formed in the past, even in the past before privatisation, still linger.
ANALYSIS BY GRAHAME LYNCH
Telstra adopts a narrower vision of where future value lies
The most interesting thing about Telstra’s FY26 result is the shape of the company that is emerging underneath it.
For those who have followed Telstra for a couple of decades, there is a certain historical irony here. Revenue is now effectively back at FY06 levels and almost 20% below its FY18 peak. The old vertically integrated Telstra grew by selling more telecommunications services to more people. The emerging Telstra is being forced to think much harder about which parts of that enormous asset base actually deserve capital.
The analyst discussion yesterday showed where the pressure sits. Postpaid mobile growth is softer. Meanwhile Telstra is directing more capital and management attention towards the collection of infrastructure assets now gathered under Telstra Digital Infrastructure.
Anyone who has sat through enough Telstra results briefings will recognise the first part of this movie. Analysts invariably start with mobile because mobile still pays the bills.
But the interesting part of Telstra’s response is that management is increasingly refusing to play the old subscriber market-share game on those terms. It wants investors to look at the whole portfolio: Telstra postpaid, prepaid, Boost, Belong and wholesale. That is logical, but it also tells you something about where the Australian mobile market has arrived. Growth at the bottom end is stronger than at the top.
This is why Telstra’s network-as-a -product strategy matters more than the terminology might suggest.
For years, the justification for paying the Telstra premium was simple: the network was better, and particularly outside metropolitan
areas it was often materially better. With large data allowances now commonplace and cheaper brands offering increasingly capable network experiences, Telstra needs other ways to justify premium pricing.
Hence 5G Standalone, APIs, differentiated quality of service and the attempt to turn network experience into something customers can perceive and pay for.
Fixed consumer tells a related story. Telstra continues to lose services but is making more money from those it retains. Brad Whitcomb [Group Executive, Telstra Consumer] made clear the company would not chase subscriber numbers “at any cost”. Excluding legacy copper, the business is now generating more than $500 million in annual profit.
That sounds almost heretical if your mental model of telecommunications is still based on dominant market share.
It is also a not-so-subtle rebuke to those in policy circles who still seem to want Telstra to be the ubiquitous caring, sharing national telco: maintaining enormous reach, competing everywhere, carrying marginal customers and accepting lower returns in the broader national interest.
Sorry, but that is what you created the NBN for.
That structural separation was supposed to end the era in which Telstra carried both the commercial burden and the social welfare expectations of being the telco of last resort. It is therefore a little rich to lament when the company acts as a for profit company: allocating capital where returns are strongest and declining to pursue uneconomic scale for its own sake.
The more consequential change, though, is Telstra Digital Infrastructure.
InfraCo Fixed already produces close to $2.8 billion of income. Telstra has now put InfraCo, International and Field together under Steven Worrall. Inside that perimeter sit fibre routes, ducts, subsea capacity, landing assets, physical sites, towers, satellite infrastructure and data centre-related assets.
In other words, much of the physical machinery required by everybody else’s AI boom.
The industry risks making a conceptual mistake by treating AI infrastructure as synonymous with data centres.
The data centre may be where the GPUs live, but those machines are not much use without power and enormous quantities of connectivity.
That connectivity has to run somewhere.
The hyperscalers can build data centres, finance submarine cables and increasingly own significant network infrastructure themselves. But Australia still requires terrestrial fibre, diverse routes, landing points, backhaul and access to physical sites.
Telstra already owns a great deal of it.
AURA NOT JUST ANOTHER PROJECT: This is where Aura becomes more interesting than just another large Telstra capital project.
Its original business case dates from 2022, before ChatGPT existed. Since then, data centre demand has shifted towards much larger AI developments, including gigawattscale “AI factories”. Google has become Aura’s largest customer, and Telstra says the pipeline has accelerated significantly.
There is an obvious reason analysts remain sceptical given previous telco bubbles. The extra $200 million of Aura expenditure sharpened that question yesterday.
Michael Ackland’s [Telstra’s CFO] answer was essentially that cash starts appearing as serviceready routes fill over the next couple of years, while the economics should be judged over a slightly longer asset life: a double-digit IRR and roughly nine-year payback.
That brings Telstra back to an older version of itself.
Telstra’s previous strategic strength was the simple fact that it owned things that were difϐicult and expensive for anybody else to replicate.
Liberalisation, regulatory declaration and the NBN progressively changed the value attached to many of those assets.
AI may now be changing the equation again.
Ducts that looked like boring legacy infrastructure suddenly matter if somebody needs multiple high-capacity routes between huge data centres. Fibre corridors matter more when AI trafϐic becomes material. Subsea capacity matters when workloads move between Australian and overseas compute regions.
This does not mean Telstra has discovered an AI goldmine. It does not lack competition either, as Vocus, FibreconX, SUBCO and others demonstrate. What’s more, the hyperscalers are formidable counterparties and increasingly formidable infrastructure owners. Telstra still has to prove it can earn attractive returns rather than merely provide another input into somebody else’s more proϐitable business.
And there is still plenty of risk around the future operating environment.
Potential mobile performance standards could divert capital spending. There are unresolved questions around roaming, possible declaration of the RAN and the future cost and availability of spectrum. Add the still-unknown cost of effectively insuring customers against scams, plus a regulatory regime in which an honest operational error can now create exposure to nine-digit fines, and the range of possible outcomes around future returns is wider than management would probably like.
That is the real risk in the strategy. Telstra is trying to become more disciplined about where it puts capital at exactly the same time as policymakers are creating more ways to tell it where capital must be spent and more ways to punish it when something goes wrong.
Telstra enters FY27 with two stories running at once. The first is familiar: mobile remains the earnings engine, fixed is being managed for value rather than share, and enterprise and international are still being cleaned up.
The second is more interesting.
Telstra is starting to look less like the sprawling telecommunications conglomerate of old and more like a mature mobile business sitting alongside a strategically important digital infrastructure platform. Revenue may have travelled all the way back to where it was 20 years ago, but the company underneath that number is becoming very different.
Grahame Lynch
IN YESTERDAY’S COMMSDAY (Friday 14 August 2026)
Mobile remained a key earnings driver for Telstra in FY26, with the telco yesterday announcing a 3.2% increase in attributable net profit to $2.24 billion and 3% growth in EBITDA after leases to $8.2bn, despite a 0.9% decline in total revenue to $23.4bn.
Telstra is seeing “very strong demand signals” for its digital infrastructure assets, with chief executive Vicki Brady pointing to an accelerating sales pipeline for capacity on its intercity Aura Network.
Telstra says it is moving beyond experimentation with artificial intelligence and is increasingly optimising technology, customer service and network operations around where AI can deliver measurable returns.
Telstra is looking to 5G Standalone and its network as a product (NaaP) strategy to create clearer differentiation between premium and lower-priced mobile offerings as growth increasingly shifts towards prepaid, wholesale and sub-brand services.
GRAHAME LYNCH COMMENTS: The most interesting thing about Telstra’s FY26 result is the shape of the company that is emerging underneath it.
The Telecommunications Industry Ombudsman has called for a “root-and-branch review” of the telecommunications regulatory framework following a systemic investigation into the experiences of telco consumers living in regional, rural and remote Australia.
Australian Telecommunications Alliance CEO Luke Coleman has warned that Australia’s ambitions to build a sovereign AI industry risk being undermined unless governments address weak returns on telecom investment, planning barriers and looming spectrum demands.
Centuria Capital Group and its joint venture ResetData has made progress with plans to develop a pipeline of AI factory deployments to meet demand for AI compute capacity.
New Zealand’s Commerce Commission has proposed keeping point-to-point, transport and co-location and interconnection fibre services regulated after finding competition remains too weak to sufficiently constrain the country’s incumbent local fibre companies.
New Zealand’s Commerce Commission has pushed back on calls to change its cost of capital methodology to reduce price shocks at regulatory resets, finding that adopting a trailing average cost of debt would deliver only modest smoothing benefits.
Inligo Networks has appointed former 5G Networks executive and co-founder Garry White as group chief commercial officer, with responsibility for global sales and commercial transactions.
Assistant industry minister Andrew Charlton says the government will move from its recently released Statement on Space to “concrete and tangible actions” aimed at commercialising more Australian technology and growing the sector’s international footprint.
Datagrid New Zealand has signed a pre-purchase agreement with Transpower New Zealand for a grid-connected substation to be built on its Southland AI factory site.
Keppel has secured a facilities-based operator licence from Singapore’s Infocomm Media Development Authority for its planned Kruger Cable System, allowing it to own and operate telecoms infrastructure for the project.
Oracle has entered a multi-year partnership with Quantinuum to make the latter’s Helios quantum computer available through Oracle Cloud
Infrastructure and explore integrated quantum, AI and high-performance computing services.
Singapore-headquartered Lightstorm has secured a 2,500 crore (US$262 million) debt facility to fund its participation in the I-2SEA submarine cable system.
Nokia has partnered with OpenNebula Systems to develop a technology stack for sovereign AI infrastructure in Europe.
Kreios Space plans to demonstrate an air-breathing electric propulsion system designed to support long-duration satellite operations in very low Earth orbit using Kongsberg NanoAvionics’ MP42 microsatellite bus.
_______________________________
CommsDay is published by Decisive Publishing, S704 6A Glen St Milsons Point NSW
ACN: 065 084960 Mailing Address: PO Box 490 Milsons Point NSW 1565 Australia
TO SUBMIT EDITORIAL FEEDBACK OR INQUIRIES: gr**********@**********il.com
TO ENQUIRE ABOUT A SUBSCRIPTION: vi***@**********il.com
