Verizon Left the Dow. The Telecom Opportunity Didn’t.

Alphabet’s promotion to the Dow isn’t a new story. It’s the same one I’ve been writing about all year — just delivered by an index committee instead of a keynote stage.

Today, Alphabet joins the Dow Jones Industrial Average, replacing Verizon Communications. It’s a small mechanical change to a 128-year-old index, but it’s being read, correctly, as a signal: the market no longer sees telecom as the face of digital infrastructure. AI, cloud, and compute have taken that seat.

I’ve spent 25 years on the wholesale and carrier side of this industry, building dark fiber, wavelength, and IRU programs with the same carriers and hyperscalers now driving that signal. So I’ll say plainly what the headline misses: this isn’t a new story. It’s the same story I’ve been writing about here all year, this time delivered by the Dow Jones index committee instead of a hyperscaler CEO on a keynote stage.

A Pattern I’ve Been Tracking All Year

Go back through what I’ve written here over the past few months and Alphabet’s seat on the Dow doesn’t create the pattern, it confirms it.

I wrote about why dark fiber is the infrastructure play of the decade: hyperscalers stopped wanting to lease capacity and started wanting to own the glass, and the routes worth owning take years to build and are running out. Alphabet joining the Dow is the index catching up to a scarcity story that’s already playing out on the ground.

I wrote about the hyperscaler land grab: how Amazon, Google, Microsoft, and Meta stopped being carriers’ customers and became carriers’ competitors, buying dark fiber, submarine capacity, and data center real estate faster than the market has priced in. Alphabet’s promotion is that land grab getting a seat on the index that used to define industrial America.

I wrote about the battery replacement wave sweeping US data centers, tied to hyperscaler capex that topped $350B in 2025 and is forecast past $600B in 2026. That same capex line is what funds the data centers, servers, and power infrastructure behind Alphabet’s $180-190B guidance this year.

And I wrote about “scale across”, Marvell’s term for AI clusters that have already outgrown a single building and now need a connective fiber layer linking campuses together. That’s the same physical layer the Dow just put a market capitalization and a board seat behind.

Four different pieces, written from four different angles, over four different months. None of them were predictions. They were descriptions of a shift that was already underway. The Dow just made it official.

The Index Changed. The Physics Didn’t.

Alphabet’s capex guidance for this year is $180 to $190 billion, up from $91.45 billion last year. That capital is going almost entirely into data centers, servers, networking equipment, and power. None of it materializes without physical infrastructure: someone has to permit it, build it, fiber it, and power it. The Dow reclassified where it sees economic leadership. It did not eliminate the requirement for someone to lay the cable.

That’s the part I’d push back on gently in this week’s piece: Verizon’s exit doesn’t mean networks become less important, and the article gets that right. But it undersells how much more important they’re about to become. AI-era data centers consume orders of magnitude more bandwidth and power density than the enterprise and consumer traffic patterns telecom networks were originally built for. The fiber routes, the duct space, the substations, the interconnection points; all of it now has to be re-architected for a completely different load profile, on a timeline measured in quarters, not decades.

Where the Money Actually Lands

Having built wholesale carrier and hyperscaler programs from scratch twice — one a dark fiber and IRU program, the other built around wavelength, ethernet, and IP capacity — I’ve watched this pattern up close: every dollar a hyperscaler spends on compute creates a multiple of demand for the infrastructure layer underneath it. Dark fiber and wavelength capacity between data center campuses. Last-mile connectivity into new builds. Power procurement and interconnection. None of that shows up in Alphabet’s earnings narrative, but all of it is where wholesale carriers, fiber owners, and infrastructure operators are picking up the work the hyperscalers don’t want to own directly.

I built a version of that connective layer myself, leading a multi-carrier fiber alliance across the US-Mexico corridor that generated $80M+ in contracted revenue by giving carriers and hyperscalers one accountable partner instead of three carriers pointing across a border. The same gap exists today inside the US, between gigawatt-scale campuses in Texas, Virginia, Indiana, and Louisiana that were never designed to operate as one network. Someone is going to own that handoff. It’s the highest-value seat at this table, and it isn’t the hyperscaler’s seat to fill.

This is also exactly why the NeoCloud category has emerged so fast over the past two years: a new layer of infrastructure-focused operators built specifically to absorb this demand, sitting between the hyperscalers and the physical world of power, fiber, and real estate. They didn’t exist as a distinct category five years ago. Now they’re some of the most active buyers of wholesale capacity in the market.

The Real Shift for Telecom

The opportunity for telecom companies isn’t disappearing, it’s relocating. It’s moving away from being the default, branded face of connectivity for consumers and enterprises, and toward being the specialized, trusted infrastructure partner that hyperscalers and NeoClouds can’t scale without. That’s a smaller, less visible seat than the one Verizon is giving up at the Dow. It is also, for the operators who reposition fast enough, a far more valuable one, because the buyers on the other side of that table are deploying capital at a pace and scale telecom has never sold into before.

Railways, steel, and oil each took decades to reshape the economy. Telecom itself took decades to build out. AI infrastructure is compressing a similar transformation into a few years. The companies that win in this cycle, whether they’re hyperscalers, NeoClouds, or the wholesale carriers underneath them, will be the ones that read the speed correctly and move their commercial strategy at the same pace the capital is moving.

Verizon left the Dow. The demand for what telecom actually builds didn’t go anywhere. It just found a new customer with a much bigger checkbook, and it’s the same customer I’ve been writing about all year.