Telecom coverage claims: the -115 dBm line and evidence files

November 9, 2026 · 12 min · Free, with full transcript

How should telecom marketers defend speed and coverage claims now that regulators in Australia, the US and the UK use standard yardsticks? The episode argues the evidence file comes before the copy. Telstra's three million square kilometre claim, measured at -122 dBm, shrinks to about 1.9 million on a plain handset at ACMA's -115 dBm line. Listeners learn a one-line claim ledger: exact words, broadest reasonable reading, proving metric, expiry date and a rival-ready proof, with the rule that if the reading is wider than the proof, the words change. Cases cover Vodafone UK at the ASA, T-Mobile, Verizon and Charter at NAD, Ookla and RootMetrics awards, two-tier evidence and MVNO resellers.

Listen to the podcast

11 min

Chapters

Key takeaways

  • Give every claim a ledger row with its exact words, the broadest reasonable reading, the proving metric, an expiry date and a version of the proof you could hand a rival.
  • If the customer's reading is wider than the proof, change the words instead of finding a friendlier dataset.
  • Keep statistics away from brand taglines, because nearby numbers turn a positioning line into a superiority claim, as Vodafone UK found at the ASA.
  • Claim only the category your evidence covers, so speed data supports a speed claim and nothing broader like best network.
  • Build two tiers of proof: a shareable file with method, dates, sample sizes and totals, and a locked file with site-level engineering data.
Read the full transcript

Host:You're listening to MBA Training. Today's subject: The -115 dBm line and the evidence file every telecom coverage claim now needs. Picture a sign-off meeting in Sydney, years back. Somebody's holding a storyboard: ute on a red dirt road, one bar of signal, huge number on screen. Nobody in the room asks what that one bar actually means.

Expert:And nobody has to, because back then nobody could check. That's the whole story, really.

Host:The number being Telstra's three million square kilometres of coverage.

Expert:Three million. Telstra being Australia's biggest mobile operator. They ran that claim for years, a rival pushed the ACCC, the Australian competition regulator, to go after it, and the ACCC said no.

Host:Wait, they let it go?

Expert:They let it go. Their reasoning was "technical complexity and evidentiary challenges." In plain English, there was no consistent way to measure mobile coverage, so nobody could prove the claim was false.

Host:Which is a gloriously Australian outcome. "Mate, who's to say?"

Expert:Pretty much. Except the "who's to say" is gone now. ACMA, the communications regulator, has a coverage map standard. Any signal weaker than minus 115 dBm gets labelled "no coverage."

Host:Define dBm before someone emails us.

Expert:Decibel-milliwatts. It's signal strength. The more negative the number, the fainter the signal. Minus 115 is roughly where a normal phone stops being a phone and turns into a very expensive torch.

Host:And Telstra's three million was measured...

Expert:...at a weaker threshold, minus 122. That counts flaky, in-and-out signal. On a plain handset, with no external antenna bolted to your roof or your ute, analysis of Telstra's own figures to the ACCC puts it at around 1.9 million.

Host:Hang on. So the gap is...

Expert:About a third of the headline. Roughly a third of that famous number assumes you own kit that most customers have never seen.

Host:A third of the country, conditional on a ladder.

Expert:Conditional on a ladder, yes. That's the number I keep coming back to, because what matters about it is how it fell apart, more than how big it was. Nobody lied about square kilometres. They picked a yardstick, and then the official yardstick turned up and it was a different one.

Host:Let me push on that, though. Telstra still has the biggest network in Australia by a mile. Optus and TPG are way behind even on the handset number. So what did they actually lose?

Expert:The multiplier. The lead survives. The size of the lead, the bit the ad was bragging about, shrinks a lot once you measure on a handset. It's still a great story. It's just a smaller one.

Host:And smaller stories don't get the dramatic drone shot.

Expert:Tragic. The drone operator will have to find other work.

Host:Okay, I have to drop something in here, because it rhymes. The Agile Brand Guide reported this week that Visa ran AI shopping agents through a benchmark, and they failed 42.8% of the runs.

Expert:Nearly half. Of the shopping. Which is the one job they exist to do.

Host:And same day, an FIS survey found half of US adults want to approve every single transaction their AI assistant makes.

Expert:Honestly, sensible people. It's the same pattern, though. The agent vendors made a big capability claim, then someone built a standard test and the claim met the test. Same thing that happened to Telstra.

Host:Except nobody's regulating "my agent buys your groceries" yet.

Expert:Give it time. Telecom just got there first because coverage is physical. You can stand in a paddock and check.

Host:Right, so zoom out. What's the durable lesson? Because "don't exaggerate" isn't a lesson. My nan knew that.

Expert:The lesson is that the evidence file comes before the copy. The best operators treat a claim as a sentence the evidence has earned. It's never a sentence the brand team liked and then went shopping for proof.

Host:That sounds very nice on a slide.

Expert:It's brutal in practice. There's a one-line version I love, sort of a ledger: every claim gets a row. Exact words, then the broadest reading a reasonable customer could take, then the metric that proves that reading, then an expiry date, then a version of the proof you could hand to a rival. And the rule is, if the reading is wider than the proof, the words change.

Host:The words change. Not the proof gets massaged.

Expert:Correct. You don't go find a friendlier dataset.

Host:Here's where I think you're being naive. Brand lines aren't claims. "The Nation's Network" is a vibe. It's a feeling with a logo.

Expert:Vodafone UK made exactly that argument this January. Said it was a corporate positioning line, not capable of objective proof. The ASA, the UK advertising regulator, said consumers would read it as superiority. More customers, better coverage, more reliable.

Host:Because people are literal?

Expert:Because the ads sat it next to "99% population coverage" and "London's Best Network." The nearby copy turned the vibe into a promise.

Host:Okay, that's self-inflicted. Put a statistic next to your tagline and the tagline stops being poetry.

Expert:And Vodafone had actually won an earlier ruling on an older execution. They leaned on it. New creative, new reading, different result.

Host:That one stings a bit, honestly. Somebody in that building thought they were safe because of a ruling with their name on it.

Expert:Yeah. It's the worst kind of loss, the one where you did your homework for the previous exam.

Host:But I'll still defend the vibe. Brands need room to be bold. If every line has to survive a lawyer, you get ads that sound like terms and conditions.

Expert:I don't buy that. The bold ones are the narrow ones. T-Mobile's fibre business took on Charter recently with "doesn't slow down during peak hours." NAD, the US ad industry's self-regulator, upheld it, because the claim was tight and the evidence covered exactly that.

Host:"Doesn't slow down at peak" is bold?

Expert:It's a fight-starter aimed straight at the cable company's weakest moment, eight p.m. when everyone's streaming. That's bolder than "best network," which is just a word wearing a crown.

Host:Fine, but you're picking your winner. Most marketers don't have a claim that neat sitting around.

Expert:Then they've got a product problem dressed up as a copy problem.

Host:Oh, come on. That's a very consultant thing to say.

Expert:It's a very true thing to say.

Host:We're not settling this.

Expert:We are not.

Host:Words, then. Where do they break?

Expert:Telecom superlatives are a trap because each one maps to different engineering. "Fastest" is download speed. "Largest" could be land, population, or subscribers. Verizon said "America's largest network" and meant connections. T-Mobile said, in this industry, "largest" means geography or population. NAD called it ambiguous.

Host:So Verizon was technically right about something nobody thought they meant.

Expert:Which in advertising is the same as being wrong, just with more paperwork.

Host:And awards? Everyone's got a trophy now.

Expert:Everyone's got a trophy because the trophy-makers disagree. Ookla, and worth saying they sell network testing data, so it's a commercial outfit, crowned T-Mobile on speed. RootMetrics, which Ookla also owns, gave Verizon the overall award. Same parent, two winners.

Host:That's like a school giving out two "best in class" ribbons and telling both parents it's a big deal.

Expert:And both parents put it on the fridge. The rule there is each operator can only claim its own category. Speed proof supports a speed claim. T-Mobile tried stretching speed data to cover "Best Unlimited Network," and NAD said even if the data's good, it's not as broad as the words.

Host:Here's my uncomfortable one. T-Mobile also had a claim, 5G capacity, and just refused to show the evidence.

Expert:Last December. They said the claim was true but they had confidentiality worries about sharing data in a case AT&T brought. NAD referred them to the FTC and state attorneys general.

Host:And part of me's sympathetic. Why hand your spectrum map to your biggest rival?

Expert:Totally fair instinct. But if your proof only works when nobody can see it, it doesn't work. The smart setup is two tiers. One you could give a rival's lawyers: method, dates, sample sizes, totals. And a locked one with the site-level engineering stuff.

Host:So the cagey move was really a filing problem.

Expert:Pretty much. Nobody had built the version of the proof you're allowed to show.

Host:And the people who pay are the media team who already spent the budget.

Expert:That's the part I find genuinely grim. Rulings land after the flight's run. The planners did nothing wrong and they're the ones explaining it to finance.

Host:Back to Telstra, then. Did they have anything?

Expert:Actually, yes. They pointed to a huge drive-testing program, cars covering the country, showing usable service even at weaker signal. That's real evidence. The fix is labelling it honestly: this proof supports this threshold, with this equipment.

Host:With a ladder.

Expert:With a ladder. Put it in the footnote. "Ladder sold separately."

Host:What about the resellers? The little brands riding on the big networks?

Expert:MVNOs, the virtual operators who rent capacity, repeat your claims in their own ads. In Australia the big three now have to give partners the maps. So the yardstick follows the claim wherever it gets repeated.

Host:Which means your claim is out there in places your legal team has never seen.

Expert:On some reseller's app store listing, at three in the morning, reading slightly wider every time.

Host:Last thing. Does all this make telecom marketing duller?

Expert:I think it makes it more honest about where the fight is. When anyone can measure, the advantage goes to whoever actually built the better thing in a specific place, at a specific hour.

Host:I'm still not convinced the vibe is dead.

Expert:The vibe's fine. It's the vibe sitting next to a number that gets you.

Host:So for Telstra, the real question was never whether three million was big enough.

Expert:No. It was whether they'd have picked that number if they'd known someone else would be holding the ruler.

Host:Drawn from Acs, ABC News and Notatechguy, links in the show notes. That's it from us. The reading continues at mba-training.com, new CMO analysis every week.

Read and go deeper