The first system Geoff Hollingworth worked on did not fit in a pocket. At CERN, before his long career in mobile communications, he wrote software for accelerator beam controls. The assignment placed him close to a moment that would later acquire the neatness of technological folklore: Tim Berners-Lee was inventing the web at the same institution. Hollingworth's own work was less visible. It was code attached to machinery, where precision mattered and abstractions eventually had to meet physical reality.
That origin explains something about the executive he became. Hollingworth is chief marketing officer of Rakuten Symphony, but his public arguments rarely begin with advertising. They begin with the system underneath: who owns it, how quickly it can change, what the organization measures, and whether a customer receives anything worth staying for. He speaks like an engineer who wandered into the CMO chair and decided the boundary between architecture and story was mostly imaginary.
In 1991 he joined Ericsson just as 2G was beginning. More than three decades later, he is still in the same broad industry, though almost every layer has been rebuilt. He has worked in software research and development, network rollouts, business development, consulting, marketing and strategic communications. His postings have included London, Stockholm, Dallas and Palo Alto. The résumé moves between disciplines, but the recurring job is translation: taking a complicated technical change and asking what it permits people or businesses to do differently.
The race that escaped television
A sailing race provided an early demonstration. Hollingworth led multimedia work for Ericsson around the Volvo Ocean Race, a round-the-world competition whose drama was trapped by distribution. He had worked on several editions and later admitted that he had never seen one of the half-hour television shows. Sailing did not command prime slots. The program existed, but the interested viewer could rarely find it.
For the 2008-09 race, Ericsson helped turn the event into a three-screen experience. Live in-port racing reached the web and mobile devices, while recordings remained available for television. High-definition cameras traveled on the boats. Media specialists became part sailor, part cinematographer, packaging footage from an environment where even staying upright was work. Hollingworth could finally watch the show on a handheld, or simply listen to it while moving.
The anecdote is charming because the problem feels quaint now. Of course niche sports should be available on demand. But that obviousness arrived through decisions about networks, compression, interfaces and rights. The project's useful lesson was not that mobile video was impressive. It was that distribution could release value already present in the content. Infrastructure mattered because it removed the obstacle between an audience and the experience it wanted.
“Our success starts and ends with the enjoyment of the audience of the Volvo Ocean Race.”Geoff Hollingworth, during the 2008-09 race
Inside the network, outside the habit
By 2012, Hollingworth was Ericsson's head of IP Services Strategy for North America and one of the company's Networked Society evangelists. He also became embedded with AT&T, helping its Silicon Valley innovation initiatives. That position put him between a global equipment vendor, a major operator and the culture of startups. It is an awkward but revealing place to work. Every side speaks about innovation. Each means something different by the word.
His next prominent chapter made that tension explicit. MobiledgeX, founded by Deutsche Telekom in 2018, aimed to make distributed telecom computing useful to developers. The company's pitch was not simply that servers could sit closer to users. Hollingworth described an attempt to abstract fragmented operator infrastructure behind common interfaces. He once compared the model to Airbnb for infrastructure: make scattered capacity discoverable and usable without forcing every developer to negotiate the peculiarities of every network.
The framing exposes a Hollingworth pattern. He treats technical architecture as an incentive structure. A fragmented interface discourages builders. A common one invites them. A long procurement cycle narrows who can participate. A data layer that operators cannot control limits what they can learn. The diagrams may belong to engineers, but their consequences arrive as market behavior.
A CMO argues against the easy sell
Hollingworth joined Rakuten Symphony as CMO in 2021. The company packages technology and operating experience from Rakuten Mobile's software-based, cloud-native network in Japan for other operators. It sells into a market that loves generational promises. Hollingworth's more interesting move has been to question the industry's habit of waiting for technology to solve a business problem.
His criticism is not anti-technology. It is the irritation of somebody who knows the technology too well to grant it magical powers. Open RAN can create choice, but installing new components through an old procurement model will reproduce old constraints. AI can analyze unified data and automate network reactions, but it cannot select the business an operator should pursue. A 6G standard may improve performance, but it cannot invent a reason for customers to care.
“If AI changes your business strategy, then something was wrong with your business strategy.”Geoff Hollingworth at MWC Las Vegas, 2024
At a 2024 debate about 6G investment, he compressed the argument into a joke about the “six remaining customers” an operator might have after chasing a distant standard while neglecting present needs. Then he returned to basics: performance, operational simplicity and integration with other technologies. Investment choices for the next 18 to 24 months, he argued, should be based on what people actually need.
The style matters. Hollingworth is willing to be playful, sometimes cheeky, because telecom language can become anesthetic. His posts pull in Clash Royale, Ryan Reynolds, airlines and fast-food analogies. The references are not decoration. They drag abstract network arguments back into ordinary behavior: losing a game when connectivity fails, switching to a cheaper carrier, staying loyal for points, or mistaking the measurement of consumption for the value that caused it.
The choice hiding beneath the network
In Rakuten Symphony's 2026 Industry Growth Report, Hollingworth sharpens the issue into a choice. Operators can behave as utilities, ecosystems or platforms. Each path can be coherent. The expensive position is to carry the cost structure and ambition of a technology company while earning commodity-like returns from connectivity, without committing to a model that changes the economics.
His comparison is the airline business. Flying became fiercely competitive, while loyalty and financial-service ecosystems created a different source of value. Rakuten's own mobile proposition sits inside a wider collection of shopping, payments, banking and travel services connected by identity and loyalty points. The lesson is not that every operator should copy Rakuten. It is that a network's business design determines what the network is for.
This also changes the meaning of measurement. Subscriber count made sense during a long era of human subscriber growth. When that pool stops expanding, old metrics can reward motion without progress. Hollingworth's line is blunt: “You are what you measure.” If the business wants lifetime value, faster experimentation or ecosystem participation, then architecture, teams and dashboards must make those outcomes visible.
Name the business model before buying the transformation program.
Keep enough architectural control to change direction and learn.
Use narrow automation wins to build data quality and trust.
Track the customer and economic outcome, not the fashionable input.
What the engineer keeps asking
Colleagues who have recommended Hollingworth publicly describe him as curious, resilient, loyal, persistent and unusually able to move from business strategy to marketing programs and technical architecture. Those observations fit the visible work. He debates because definitions matter. He repeats the customer question because large systems naturally turn inward. He distrusts a silver bullet because he began close enough to machinery to know that no component operates alone.
His aspiration for telecom is not modest, but it is concrete. Operators should reclaim architectural ownership, run networks with the iteration speed of software, automate routine operations and spend more human effort on customer problems. They should use AI to extend sound judgment, not to disguise its absence. They should offer developers and partners interfaces that reduce friction. And they should know whether they want to be a utility, an ecosystem or a platform.
The thread from CERN to Rakuten is a demand for alignment. At the accelerator, software had to control the beam. On the ocean, distribution had to find the audience. At the edge, infrastructure had to become usable to developers. In a mobile operator, technology, organization and measurement have to serve the same business choice.
Telecom's next chapter will contain more AI, more software and eventually another G. Hollingworth's contribution is to make those nouns answer a verb: what will the operator do for a customer that it could not do before? The question is almost stubbornly plain. After three decades of increasingly elaborate networks, plain may be exactly the point.