Telecom Advertising: Why Bots Don’t Have SIM Cards
What a decade of chasing relevance through the browser has cost advertisers, and why telecom advertising offers better signals.
Talk to media planners for long enough and the conversation usually lands in the same place. Targeting tools keep getting more sophisticated, and yet the results get harder to defend in front of a CFO every year.
Most of the industry has responded with better algorithms: smarter bidding, sharper lookalikes, new attribution models. I understand the instinct, but I think it misreads the problem. Almost all these tools still work from the same raw material, the traces people leave behind in browsers and apps. When that material degrades, a cleverer model gives you a more confident guess and little else. Very few planners have looked at what telecom advertising can offer instead.
What Ad Fraud and Wasted Impressions Cost Advertisers
Classic display regularly produces click-through rates below 1%. IAS data suggests that around a third of paid impressions never meet the IAB/MRC viewability standard. Juniper Research estimates that 22% of digital ad spend is lost to ad fraud, and the ANA’s programmatic transparency study found that only 36 cents of every dollar entering a DSP reaches the consumer.
Taken together, a large share of every media plan funds intermediaries, invisible placements and bots before a real person sees anything.
Meanwhile, the signals behind targeting keep getting weaker. Cookies are fragmenting and mobile identifiers depend on permissions set by Apple and Google. Feeds contain too much content. Many younger, tech-savvy consumers also run ad blockers out of habit. As a result, brands struggle to reach the audiences they care about most.
A large share of every media plan funds intermediaries, invisible placements and bots before a real person sees anything.
The Limits of Guessing
Most digital targeting is probabilistic. A platform sees that you watched three sailing videos and decides you like sailing. It then serves you more sailing content, because that keeps you on the platform. What it rarely knows is whether you are about to buy anything, and its picture of you goes stale as your life moves on.
For brands, the cost adds up quietly. Acquisition costs creep up quarter after quarter. We estimated the reach figures using a model. And because competitors buy the same inferred audiences on the same platforms, targeting gives nobody an edge for long. Everyone pays the same premium at every auction.
Telecom Signals as a New Source of Relevance
At VOX Solutions we came at this from an unusual angle. We have spent years working with mobile operators around the world on messaging and network monetisation, and from that side of the industry we could see something advertisers rarely do.
Telecom networks generate technical signals all the time. They record that something happened (a phone connecting abroad, activity in a commercial area, recurring interest in a category) without revealing who the person is or what they said. Classified carefully and used at segment level, inside the operator’s own environment, those signals have real commercial value.
The relationship between an operator and a subscriber also has qualities advertisers struggle to find anywhere else. It rests on a contract, the subscriber is verified, and it tends to last. Consent and marketing preferences are agreed directly between the two, under local law.
For performance budgets, that has practical consequences.
Every message lands with a person. Bots don’t have SIM cards, and a message sent through an operator channel either reaches an active subscription or doesn’t arrive at all.
Signals also describe events that actually happened, which is what makes deterministic targeting possible. A business traveller and a tourist look quite different once you see how often, and in which seasons, someone roams. Segment size is known before budget is committed, and that number is a count of eligible users.
Then there is timing. Building a retargeting audience the usual way takes hours or days. A real-time trigger, such as someone landing in a new country, can open a window for travel insurance or car rental within minutes. In categories where intent is short-lived, that gap shows up directly in the cost per acquisition.
Ad blockers have nothing to act on either. Delivery happens in native messaging, push notifications and operator apps, with no third-party tags or scripts to load, so brands keep the audiences that ad blockers would otherwise take away.
Bots don’t have SIM cards.
A message sent through an operator channel either reaches an active subscription or doesn’t arrive at all.
How TrueSignal Works in Practice
TrueSignal is the telecom advertising platform we built around this idea. The operator remains in control: technical identifiers are anonymised locally, consent is checked before activation, and personal data stays with the operator. Brands work with verified audiences and campaign results.
Segments can get very specific. For a luxury watch brand, we might combine signals linked to sailing, golf, trading and luxury shopping in cities such as Dubai, Paris and Abu Dhabi.
Campaigns run through operator channels including rich media messaging (VMS), push notifications, in-app display and video, RCS, and conversational voice.
The results so far are encouraging. A beauty retailer targeting a segment of active beauty buyers reached a 13.8% click-through rate against a 9.5% category benchmark, with 83% more interactions than demographic targeting and a 3.8% conversion rate. Gaming campaigns in Bolivia and Ecuador have delivered click-through rates of 15 to 18%.
Voice is where the numbers stand out most. In fintech, insurance and real estate, an AI agent can speak with eligible users in their own language and follow up by SMS. Response rates sit between 35% and 55%, and 4 to 8% of contacted leads convert. I usually sum it up as voice qualifying the lead and SMS closing it.
We price TrueSignal on outcomes – cost per action, lead or click – and we have never sold it on CPM. Before a campaign starts, we agree what counts as a qualified lead and how conversions will be measured, so both sides work from one set of numbers.
With AI voice agents, response rates sit between 35% and 55%, and 4 to 8% of contacted leads convert.
What Telecom Advertising Means for Brands and Agencies
Social, video and display remain very good at building awareness, and I would keep investing in them for that job. TrueSignal was built for another part of the funnel: finding people who are not yet in a brand’s CRM and reaching them while their intent is current.
For agencies, we have kept the workflow simple. The agency brings the brief and the ideal customer profile. We build and size the segments, recommend formats and a pricing model, and produce the creative at no extra cost, including training the voice agent. The agency keeps final approval over the message. Many campaigns go live within hours of a signed brief.
The main threshold is audience volume. Below roughly 30,000 to 50,000 users in a segment, there is too little data to learn anything useful, so we widen or combine segments at the brief stage, before any money is spent.
In Romania, we reach an estimated 40% of subscribers eligible for marketing communications, and the same model runs with operators in Latin America, the GCC, MENA and Asia Pacific.
Advertisers who start testing verified telecom signals this year will build up campaign data, benchmarks and creative learning while their competitors keep paying more for the same crowded inventory. That kind of head start is very hard to close later. The market is already moving in this direction, and it will keep moving with or without the brands that decide to wait.
Read the full article, in Romanian, here.
Next year’s media budgets are being decided now. Some brands will pay for real customers. Others will keep paying bots.
Which one will you be?
