The global telecom industry generates over USD 1.15tn in annual service revenue and carries nearly 4.5 zetabytes of mobile data, yet it faces a widening structural gap between traffic growth and revenue growth. With data volumes set to nearly double by 2029 and mobile ARPU actually declining, operators are caught between falling returns on connectivity and a generational AI infrastructure investment cycle that could either save or bypass them entirely. We take a deep dive into the trends reshaping the sector and what they mean for European and CEE operators.

For years, the European telecom story has been easy to summarise and hard to get excited about: traffic keeps rising, capex stays elevated, competition remains rational only some of the time, and revenue growth rarely looks transformational. That backdrop has not changed. Global telecom service revenue is expected to increase from roughly USD1.15tn in 2024 to USD1.32tn in 2029, a modest 2.8% CAGR, while global mobile ARPU edges down from USD 6.32 to USD 6.20. In other words, usage is still compounding faster than pricing power, for investors, that has been the industry’s defining problem.

Global telecom service revenue vs. mobile data traffic (2020 – 2029E, USDtn, ZB)

Source: GSMA Intelligence, InterCapital Research

What is changing, in our view, is the strategic role of the network. Telecom is no longer just a connectivity utility. It is increasingly becoming a sovereign digital infrastructure. As AI workloads expand, more value is being created around where data is processed, how securely it moves, how low the latency is, and under whose rules it is stored. That matters because operators own assets that are difficult to replicate: fixed access networks, spectrum, metro fibre, local field infrastructure, enterprise relationships, and trusted national operating footprints. The next winners will not be the operators that merely sell more gigabytes, but the ones that monetise resilience, compliance, and service quality.

The most consequential development is the collision between two investment cycles moving in opposite directions. Telecom capex intensity has been stepping down, from around 27% of revenue in 2022 to roughly 23% in 2024, as many markets approach the end of initial 5G and first-wave fibre builds. Meanwhile, AI infrastructure spending is accelerating into a multi-year super-cycle: the USD 500bn Stargate Project in the US, the EU’s EUR 20bn InvestAI programme and Saudi Arabia’s USD100bn Project Transcendence. GSMA’s 2026 Network Transformation Survey confirms that data centre capacity is now the number one network investment priority for operators in every region surveyed. The question for operators is whether they will be providers of the connective tissue linking this compute economy together, or whether it will simply grow around the networks they built.

On network infrastructure, 5G Standalone is finally moving from promise to deployment at scale. As of September 2025, 78 operators had launched 5G SA in 42 markets, with Europe leading at 30 launched and 47 more planned. This matters because SA unlocks the capabilities like network slicing, mobile edge computing, and RedCap, that underpin the B2B monetisation thesis that justified 5G capex in the first place. GSMA consumer survey data shows users intending to upgrade to 5G are willing to pay on average 5% more than their current 4G plan. Ericsson’s latest forecasts project the global smartphone subscription base will reach approximately 8.5bn by 2030, with 5G accounting for roughly 6.3bn, or nearly 74%, of those connections, while legacy technologies virtually disappear.

Global smartphone subscriptions by technology (2020 – 2031E, in millions)

Source: Ericsson Mobility Report, InterCapital Research

That lens is especially useful in Central and Eastern Europe. Based on Ericsson regional forecasts, 5G accounts for only about 12% of smartphone connections in CEE in 2026, versus roughly 67% in Western Europe, and CEE does not cross the 50% mark until around 2030. The region still has a longer runway for 5G migration, but many operators will have to fund 5G standalone upgrades before the revenue case is fully visible.

CEE smartphone subscriptions by technology (2020 – 2031E, M)

Source: Ericsson Mobility Report, InterCapital Research

On fibre, however, the picture is more nuanced and more interesting. FTTH Council Europe data show that the broader European market has already moved beyond the “build at any cost” phase: EU39 fibre coverage reached 79.3% by September 2025, yet take-up was only 54.4% and penetration 43.1%. In other words, the strategic bottleneck is no longer rollout but monetisation. Many CEE markets are already ahead of the European average on coverage. Hungary stands at 82.3%, Poland at 83.1%, Slovenia at 85.0%, Bulgaria at 90.2%, and Romania at 97.0%. But the same countries do not all lead in take-up. Poland is at 41.2%, Hungary 43.4%, Bulgaria 43.9%, and Slovenia 50.5%, all below the EU27+UK average of 54.9%. The implication is straightforward: in CEE, the next phase is not about who can pass more homes, but who can migrate more customers, switch off legacy copper faster, and attach higher-value products to the connection.

European fiber: homes passed vs. take-up rate by country (2025, %)

Source: FTTH Council Europe, Eurostat, InterCapital Research

One of the most actionable near-term trends for our region is 5G Fixed Wireless Access. GSMA forecasts that 5G FWA net additions will accelerate in 54 of 72 tracked markets in 2026. Four markets will exceed 10% FWA penetration of total fixed broadband: Austria at 22%, the US at 13%, New Zealand at 12%, and Czechia at 11%. In CEE specifically, the numbers are meaningful: Poland is set to add approximately 266,000 5G FWA subscribers, the Czech Republic 164,000, Austria 154,280, Bulgaria 46,000, Slovakia 26,500, Croatia 15,000, Hungary 10,000, and Slovenia 10,000. For markets where last-mile fibre economics are challenging, FWA offers a credible broadband alternative as 5G SA coverage expands.

Projected 5G FWA net additions in CEE markets (2026E, 000s)

Source: GSMA Intelligence, InterCapital Research

The spectrum landscape also deserves attention from investors who tend to overlook it. Revenue per MHz per connection has declined by 67% over the past decade, from USD 0.37 in 2014 to USD 0.12 in 2024, according to GSMA’s spectrum pricing data. Licences in 43 markets are set to expire in 2026, with Europe accounting for the largest share. Meanwhile, the 6 GHz band is gaining traction for mobile, with 29 European markets engaging with it for IMT identification. In the satellite domain, direct-to-device connectivity is accelerating: 116 mobile operators had partnered with satellite providers by the end of October 2025, with an estimated ARPU uplift of 10-30% for mobile contract customers where charged.

Spectrum revenue per MHz per connection (2014 – 2024, USD)

Source: GSMA Intelligence, InterCapital Research

The enterprise opportunity may be even more important than any of the consumer-facing trends. GSMA’s 2026 industry outlook suggests AI adoption will increase enterprise demand for better network performance, edge processing, and security. At the same time, more enterprises expect AI workloads to sit on a hosted private cloud than on a public cloud, while telco-provided edge still has limited mindshare. That gap is precisely where operators can create value. In CEE, the opportunity is not to imitate hyperscalers, but to become the trusted local partner for regulated and latency-sensitive workloads across manufacturing, logistics, utilities, healthcare, and the public sector. In a region where data sovereignty, infrastructure resilience, and domestic control increasingly matter, local operators have an advantage that global platforms do not automatically enjoy.

One important structural shift reinforcing this thesis is the accelerating separation of passive infrastructure from service operations. Seven out of ten towers globally are now owned by TowerCos, which achieve average tenancy ratios of 2.4x versus 1.3x for operator-controlled towers. Listed TowerCos trade at 19-20x EV/EBITDA in the US, while most European integrated operators trade at 5-7x. That valuation gap is the single most powerful financial incentive for further carve-outs, and it is directly relevant to CEE, where operators sitting on portfolios of several thousand tower sites are holding assets that the market would value at a significant premium if separated.

On the ground in CEE, these themes are already playing out. Magyar Telekom and Hrvatski Telekom are the furthest along on 5G and fibre, with Magyar Telekom targeting 99% 5G coverage by the end of 2026 and HT having completed its 3G shutdown in January 2025 to refarm spectrum. Digi is the region’s most aggressive growth story, with nearly 100% FTTP/B coverage in Romania and an expanding Western European footprint after launching in Belgium and entering Portugal. Telekom Slovenije became Slovenia’s FTTH market leader and is pivoting toward ICT and cybersecurity, while A1 Telekom Austria remains the furthest along on infrastructure monetisation, having completed the EuroTeleSites tower spin-off in September 2023. Orange’s announced tower carve-out plans for Romania, Slovakia, and Moldova suggest the model could spread further across CEE.