Telcos aren’t broken, but they’re stuck. EBITDA looks fine, yet growth crawls, ROIC slides, and M&A papers over cracks.
- Growth crawls: service revenue up just 2.9% CAGR through 2028; slower than inflation.
- Returns slip: $300B extra CapEx since 2018, ROIC down 10–15%.
- CapEx bites: peaked at 15.1% of revenue, forcing asset divestments to free up cash.
- Consolidation everywhere: in Europe, M&A became the default “spreadsheet fix.”
One colleague put it bluntly: “The biggest challenge is balancing CapEx demands with customer expectations.” That tension sits at the heart of the industry.
And then there’s the waiting game. Non-standalone 5G was a stopgap, standalone is still delayed, and now the industry waits for 6G, or “6Godot,” as another colleague coined brilliantly.
So what do telcos reach for? Efficiency. It buys time, freeing cash and stabilizing margins. But efficiency is a pause button, not a growth story. And that’s why the AI conversation matters: not as pixie dust, but as a way to build things customers will actually pay for.
The megatrends (you can’t ignore!)
PwC highlights five forces reshaping every industry: climate, demographics, tech breakthroughs, shifting global power, and urbanization. These megatrends are turning efficiency into table stakes, and telcos are no exception. The pressure to reinvent is bigger than ever: smarter, greener networks, real personalization in crowded markets, faster innovation cycles, and more open ecosystems.
The dual mandate for AI
This is where AI enters the picture with a clear dual mandate: defend margins today, unlock revenue tomorrow.
- Efficiency now: cut waste, predict failures, stop fraud. AT&T saves millions through predictive fleet maintenance; Vodafone’s TOBi handles 45M chats a month, slashing response times by 60% and cutting service costs by a third.
- Revenue next: the bigger prize. B2B AI revenue is small today (about $4B in 2025), but it’s set to grow at a steep 65% CAGR through 2030. That’s not hype; that’s the upside.
From the basement to the revenue engine
Today, for many telcos, AI is stuck in the pilot phase or lives in the basement, doing ops work customers never see. The next step is moving it upstairs, into the revenue engine:
- Predictive SLAs: enterprises will pay premiums for quantifiable uptime.
- AI-powered APIs: Telco networks exposed via APIs (see GSMA’s Open Gateway) plus AI turn static catalogs into living marketplaces with dynamic pricing, discovery, fraud detection, orchestration.
- Hyper-personalization: every interaction tailored. McKinsey estimates +8% revenue, –15% cost-to-serve, +20-40 CSAT points in customer satisfaction.
The GTM problem (and the TechCo jump)
Let’s be honest. Telcos don’t lose on networks; they lose on go-to-market.
Telcos still sell minutes, megabytes and coverage. TechCos sell outcomes: fewer outages, less fraud, faster integration. Customers don’t buy “AI” or “5G.” They buy results.
The commercial playbook has to change:
- Start with customer problems, not internal KPIs,
- Sell solutions, not minutes or megabytes,
- Build ecosystems, not just networks.
In a recent LinkedIn poll I published, respondents split their votes across four GTM levers:
- AI-driven personalization
- Ecosystem partnerships (B2B2X)
- Adjacent services (cybersecurity, IoT, cloud)
- Reinventing the core (simpler offers, faster launches)
No single silver bullet. But the energy around ecosystems and AI-driven personalization stood out. One voice put it bluntly: “Ecosystem is definitely the most lucrative option, if operators can finally move from the shelf to the market table.”
And here’s the good news: telcos already have enviable assets: strong brands, huge customer bases, rich data, vast infrastructure, and skilled people. The opportunity, as McKinsey puts it, is to reinvent, whichever path you take. Other industries already have:
- Finance → FinTech: APIs forced banks to compete on UX, not just licenses.
- Energy → EnergyTech: utilities moved from selling kWh to smart grids and EV charging, turning efficiency mandates into monetized services.
- Healthcare → HealthTech: shift from reactive care to remote monitoring and AI diagnostics.
- Media → Streaming: Platforms like Netflix owned the interface. Pipes without product lost. (Sound familiar?).
Why now?
MWC25 made it clear: AI is everywhere. But most telcos still use it for efficiency only. Fair enough. Just don’t stop there. Use the savings to build the revenue layer.
The tech innovation is already there. What’s missing is the courage to innovate go-to-market. As one industry colleague said: “Telcos are prisoners of their legacy systems. They alienate consumers with bundles no one wants, while OTTs engage with users daily.”
That’s the GTM gap in plain words.
Final thought
AI is a dual catalyst: defend margins today, unlock new revenue tomorrow. The rest is execution.
So I’ll leave you with the same LinkedIn poll question:
Which GTM lever will really move the needle for telcos: personalization, ecosystems, adjacent services, or reinventing the core?
