The 149-year-old Swedish firm that helped write the rules of 5G now has to prove those networks can grow again - under a new CEO and an AI-driven cost squeeze.
You have almost certainly never bought anything from Ericsson. There is no Ericsson app on your phone, no Ericsson box on your desk, no checkout page with its logo at the bottom. And yet when your phone jumps from "4G" to "5G" in the corner of the screen, there is a decent chance the equipment doing that handoff was built by a Swedish company that opened as a telegraph-repair workshop in 1876 - the same year the telephone was patented.
That is the strange position Ericsson occupies. It is one of the largest telecom-infrastructure companies on earth, with roughly 89,000 employees and business in about 180 countries, but it lives one layer below the consumer. Its customers are the carriers - more than 100 mobile network operators - and the carriers sell to you. Owning the layer under everyone else's customer is an old, patient, unglamorous strategy. Ericsson has been running it for 149 years.
What it actually does
Strip away the acronyms and Ericsson does one core thing: it builds the radio access network, or RAN. These are the base stations, antennas and basebands bolted to towers and rooftops that turn your phone's signal into internet traffic and back again. Around that sit two more layers - the "transport" that moves the data, and the core software that routes calls, authenticates SIM cards and bills the customer. Increasingly, that core runs as cloud-native software rather than dedicated hardware, which is a quiet but important shift in how the whole business works.
The reason this matters is scale. A single operator upgrading a country to 5G is a multi-year, multi-billion-dollar project involving tens of thousands of sites. There are only a handful of companies on the planet that can deliver equipment at that scale, keep it running, and guarantee it will interoperate with everyone else's gear. Ericsson is one of them.
Who actually buys it
The buyers are the big carriers - the household telecom brands in North America, Europe, the Middle East and Asia. When one of them announces a nationwide 5G rollout, Ericsson is often on the shortlist of vendors doing the work, and it frequently runs the networks it builds through long managed-services contracts. That last part is worth pausing on: a lot of Ericsson's value is not the box on the tower but the promise to keep millions of connections alive at 3am, which is why so much of its workforce is engineers on the operations side rather than salespeople.
Increasingly the customer list runs wider than carriers. Through Cradlepoint, a factory can run its own private 5G network for robots and sensors; through Vonage, a software team can wire calling and messaging straight into an app. The common thread is connectivity someone else has to make reliable - and that is the problem Ericsson has spent nearly 150 years learning to solve.
The hidden advantage
Here is the part that separates Ericsson from an ordinary hardware vendor. Cellular networks only work because every phone and every tower, from every manufacturer, agrees on the same technical standard. Those standards are set by a body called 3GPP, which Ericsson helped found in 1999. Since then the company says it has contributed more than 80,000 technical inputs shaping everything from 2G to 5G - and it holds one of the industry's largest patent portfolios, with tens of thousands of granted patents.
That produces a second, less visible business: licensing. When rival manufacturers build equipment that uses those standards, they often pay Ericsson royalties. It is a high-margin income stream that keeps flowing regardless of who wins a given hardware contract. Selling the gear is the loud business. Owning the patents is the quiet one.
The competition
The market for network equipment is not crowded - it is concentrated. The main rivals are Finland's Nokia, China's Huawei and South Korea's Samsung, with ZTE also in the mix. For years Huawei was the low-cost giant taking share worldwide, until security concerns pushed several Western governments to restrict or ban its gear. That reshuffled the board and handed Ericsson and Nokia openings in markets they might otherwise have lost.
So how is Ericsson different? Partly it is the standards and patent depth described above. Partly it is geography - a Nordic vendor with deep roots in North America and Europe is politically easier to buy from than a Chinese one in many capitals. And partly it is the pivot toward software and services, where the pitch is not "here is a better antenna" but "here is a network you can program."
It helps to see how the money actually arrives. There are three streams: one-off equipment sales tied to build cycles, which swing up and down; recurring software and managed-services contracts, which are steadier; and patent licensing, which keeps flowing no matter whose hardware wins. The strategic project of the past few years has been to lean harder on the second and third - the predictable parts - so the company is less hostage to the boom-and-bust rhythm of hardware orders. That is the whole logic behind the Cradlepoint and Vonage deals, and behind Narvinger's job description.
The reinvention
The trouble with selling 5G hardware is that the boom eventually flattens. Once a country is built out, the giant orders slow down. Ericsson's response has been to expand what it sells and to whom. In 2020 it bought Cradlepoint, a US wireless-networking firm, for around $1.1 billion, moving into enterprise wireless and private 5G - dedicated networks for factories, ports, stadiums and warehouses that do not want to share public airwaves.
Then in 2022 came the big swing: a $6.2 billion acquisition of Vonage, its largest deal ever. Vonage runs a cloud-communications platform used by well over 100,000 businesses, and it comes with something Ericsson wanted badly - a way to expose network features through APIs that ordinary software developers can build against. The bet is that a phone network can become a programmable platform, the way cloud computing did. Whether that pays off is one of the open questions hanging over the company.
The moment
All of this is happening during a leadership change. Börje Ekholm, who ran Ericsson for nine years - through a $1.2 billion US bribery-case settlement in 2019, the Vonage deal, and the entire 5G build - is stepping down. In his final quarterly results the company reported revenue down about 6% to roughly SEK 52.7 billion and warned that prices may have to rise, because the AI boom is pushing up the cost of the custom chips that go inside radio equipment. It is an unexpected knock-on effect: the same silicon crunch feeding data centers is now reaching cell towers.
His replacement, effective October 2026, is Per Narvinger - an Ericsson insider who joined in 1997 and most recently ran the Networks business. That is a telling choice. It signals continuity rather than a hard reset: the job is less about a new vision and more about delivery, margins and proving the software story can actually grow the top line.
The long view
It is worth remembering how far back this goes. Lars Magnus Ericsson started as a mechanic fixing telegraph instruments before building his own telephones. The company later gave the world the AXE digital exchange, deployed in over 100 countries, and the one-piece Ericofon that now sits in design-museum collections. It even spent a decade making consumer handsets through the Sony Ericsson joint venture before walking away from phones entirely to focus on the networks underneath them.
Now the research labs are already working on 6G, and the Stockholm headquarters itself is set to move from Kista to a new home in Hagastaden over several years starting around 2028. A company that has outlived the telephone, the mobile phone and the smartphone is betting that the next act is not a device at all, but the invisible, programmable network that everything else runs on.