Zayo Group
owns the road
your data drives on.

Right now, somewhere under a freeway, a strand of Zayo glass is carrying your afternoon.

You are not thinking about fiber. You never do. You open a video call, a model finishes training, a payment clears, a show buffers for half a second and then doesn't. All of it is light moving through hair-thin glass buried under roads, rivers and railbeds. A lot of that glass belongs to Zayo Group, and Zayo would prefer you keep not thinking about it. The company sells the one layer of the internet that only gets noticed when it breaks.

Zayo is a communications infrastructure company headquartered in Denver, Colorado. It owns and operates more than 150,000 route miles of fiber across North America and Europe after closing its acquisition of Crown Castle's fiber business in 2026. It does not make apps. It does not sell phones. It sells capacity - the right to move enormous amounts of data from one place to another, fast and reliably - to the companies whose names you do know.

In 2007, the smart money said the internet had too much fiber. Zayo's founders disagreed, loudly.

Rewind to the years after the dot-com crash. Telecom companies had laid fiber like it was going out of style, then watched demand fail to show up on schedule. The conventional wisdom hardened into a single word: glut. There was too much fiber, the story went, and there always would be.

Dan Caruso, John Scarano and Matt Erickson read the same market and saw the opposite. Caruso had helped build Level 3 Communications, so he knew the economics of glass better than most. Their thesis was simple and unfashionable: bandwidth demand does not run backward. It compounds. And while the headline numbers looked oversupplied, the fiber that actually mattered - metro routes into buildings, towers and data centers - was scarce, fragmented, and badly undervalued. Someone just had to go buy it before everyone else figured that out.

It is a slightly absurd thing to start a company on - the belief that the world will simply want more of an invisible thing forever. But that is the bet, and nearly two decades of traffic curves have been kind to it.

The plan was not to invent fiber. It was to buy everyone else's.

Zayo's early strategy looked less like a startup and more like a shopping list. In its first year alone it absorbed Memphis Networx, VoicePipe, Onvoy, PPL Telcom and Indiana Fiber Works. Then it kept going. Over the following years Zayo completed more than forty acquisitions, including AboveNet for roughly $2.2 billion and the storied long-haul operator 360networks. Each deal stitched another set of routes onto the map.

The financial markets eventually noticed. Zayo went public in 2014, raising about $600 million. Six years later, in 2020, the investment firms EQT and Digital Colony - now DigitalBridge - took it private again in a deal valued at roughly $14.3 billion. The buying never really stopped; it just changed who was writing the checks.

Forty companies, one map. Somewhere in Denver there is a spreadsheet that should be in a museum.

What Zayo actually sells, minus the jargon.

The flagship product has an ominous name and a boring reality. "Dark fiber" is just optical fiber that has been installed but not yet lit with electronics. Zayo leases those dark strands and lets the customer light them however they want - maximum capacity, full control, total privacy. Hyperscalers love it because it lets them run private highways between data centers without digging a single trench.

Around that sit the rest of the offerings: wavelengths (dedicated optical lanes now running up to 400G), Ethernet, dedicated internet access and IP transit, colocation in network-connected buildings, fiber to cell towers and small cells for wireless carriers, plus managed edge, SD-WAN and DDoS protection riding on the owned backbone. Different names, one underlying asset - glass that Zayo owns and rents by contract.

The case for Zayo isn't a slogan. It's a traffic curve that won't sit down.

Here is where the founding bet either pays off or it doesn't. For years the demand came from streaming, mobile and the migration to cloud. Now it comes from artificial intelligence, and AI has an appetite for fiber that borders on rude. Training and serving large models means shuttling staggering volumes of data between data centers, and that movement needs exactly what Zayo spent eighteen years accumulating: high-capacity, low-latency routes that already exist.

The numbers tell the story Zayo wants told. In 2025 the company added more than 1.3 million fiber miles globally and brought its entire network to 400G capability. In early 2026 it announced plans to build another 5,000 route miles of long-haul fiber specifically for AI demand. To pay for all that glass, Zayo has raised over $6 billion in asset-backed securitizations - financing as industrial as the product itself.

Bars are scaled for the eye, not the ledger - route miles and fiber miles measure different things. The point survives the footnote: the line goes up.

Who actually pays for all this? Wireless and national carriers. Hyperscale cloud and content providers. Banks, hospitals, broadcasters and manufacturers. Government agencies. Internet service providers who would rather rent capacity than dig for it. They are not Zayo's audience - they are its tenants, and the lease is the whole business.

"Connect what's next" is a tidy phrase for a deeply untidy job.

Zayo's stated purpose is to connect what's next - to keep building the physical layer that the digital economy quietly depends on. It is leadership with infrastructure in its bones. CEO Steve Smith took the top job in 2020 after running Equinix, the world's largest data-center company, which he grew from roughly $2 billion to $34 billion in market value. He understands the unglamorous economics of selling space, power and connectivity better than almost anyone alive.

The culture follows the product: operations-first, capital-heavy, patient in a way that software companies rarely are. You cannot ship a fiber route overnight. You permit it, trench it, splice it, light it, and then you wait years for the contracts to mature. It is a business for people who are comfortable being indispensable and invisible at the same time.

The next decade of computing is a logistics problem. Zayo is in the logistics business.

Strip away the hype and the future of AI, cloud and whatever the metaverse becomes is a question of moving data between places fast enough and cheaply enough to matter. That is a transport problem, and transport problems are won by whoever owns the network, not whoever rents it. Zayo's competitors - Lumen, Cogent, regional dark-fiber shops - are racing the same logic. The company that bought the most glass when it was cheapest now holds an asset that gets more valuable every time a new model ships.

There is risk in all of it. Debt is heavy, integration of forty-plus acquisitions is messy, and demand forecasts have humiliated smarter people. But the core wager - that the world will keep wanting more bandwidth than seems reasonable - has not been wrong yet.

You still aren't thinking about fiber. That's the whole point.

The call connects on the first ring. The model finishes a beat early. The payment clears before you've put your card away. None of it announces itself, because the best infrastructure is the kind you forget exists. Under the freeway, the strand of glass is still carrying your afternoon, and tomorrow it will carry a little more, and Zayo will have already bought the next route before anyone notices they needed it.

That was the bet in 2007. It is still the bet now. The only thing that has changed is how much of the world is riding on it.