Verizon Won the Network 35 Times Running. Then It Tried to Be Google.

Twice a year since 2003, J.D. Power surveys tens of thousands of Americans about dropped calls, dead zones, and the spinning wheel of a video that won't load. Twice a year, in region after region, Verizon comes out on top. By September 2025 it had done this 35 times in a row — Volume 1, then Volume 2, then Volume 1 again, an unbroken line of #1 finishes stretching back more than two decades.⁷ That is not a marketing slogan. It is one of the longest winning streaks in American consumer business, and it is the entire reason Verizon can charge a premium for a service that, to most people, looks identical to its rivals'.

So here is the puzzle. A company that owns the single most durable quality moat in telecom looked at that moat, decided it wasn't enough, and spent roughly $8.88 billion trying to become an internet advertising company instead.

The official story is that Verizon was building 'the third option in mobile advertising' — a media business to sit alongside Google and Facebook, monetizing all those phone-screen eyeballs riding on its network. The real story is simpler and harsher: a capital-intensive infrastructure company tried to buy its way into a winner-take-most market it had no structural right to win, and the market took its money.

The moat that compounds, and the one that can't be bought

Network quality is a strange kind of asset. You build it with cash — spectrum, towers, fiber, the endless engineering of squeezing more bars out of the same air — and once it's built, it doesn't sit still. Every year a rival closes part of the gap, Verizon pours in more capital and reopens it. The 2024 study, fielded across 26,725 customers, named Verizon highest-ranked in all six U.S. regions.⁸ The mechanism is dull and powerful: the company that already wins keeps winning because the people most willing to pay for reliability cluster on the best network, fund the next upgrade, and widen the gap that brought them there. The award streak isn't a trophy. It's the visible exhaust of a flywheel that runs on its own reputation.

Advertising is the opposite kind of asset. Its moat isn't built with capital — it's built with data, scale, and the ruthless network effects of a two-sided marketplace. Google and Facebook don't win because they spend more; they win because every advertiser must go where every user already is, and every user is already there. You cannot pour money into that gap and close it, because the gap isn't made of money. Verizon's core competence — deploying enormous capital to build physical infrastructure — was precisely the wrong tool. It brought a checkbook to a knife fight over attention.

The price of buying into a fight you couldn't win

The shopping spree came in two trips. In May 2015 Verizon agreed to buy AOL for $50 a share — roughly $4.4 billion — for its programmatic ad-tech stack and its content brands.¹ Then in February 2017 it closed on Yahoo's operating business. The headline number you'll see is '$4.5 billion,' but the SEC-filed amended agreement is exact: $4,475,800,000, after Verizon negotiated a $350 million discount when Yahoo disclosed two enormous data breaches mid-deal.² The discount itself was a tell — Verizon was paying full freight for an asset whose chief value, user trust, was actively leaking out the side.

Bolted together, the two became Oath, a unit pitched with the ambition of reaching two billion users — a figure attributed to its chief, Tim Armstrong, rather than any filing, and one it never came close to hitting. By the fourth quarter of 2018, less than two years after the combination, the accountants had seen enough. Verizon disclosed a $4.6 billion goodwill impairment on Oath, citing 'increased competitive and market pressures' and 'lower than expected revenues and earnings.'³

It's worth being precise about what that charge meant, because the headlines weren't. A wave of coverage suggested Verizon had written Oath down to $200 million — i.e., to near-zero. Yahoo Finance had to append an explicit correction: Verizon wrote the goodwill portion of its investment down to about $200 million, but the unit still held roughly $5 billion in other assets.⁴ That distinction is the whole lesson. Goodwill is accounting's name for the value a deal is supposed to create — the premium you pay above what the parts are worth, on the bet that you'll make them worth more together. Verizon erased nearly all of it. The assets survived; the thesis didn't.

The exit, and the math nobody wants on a slide

In May 2021 Verizon gave up. It agreed to sell the media business — by then renamed Verizon Media — to Apollo Global Management at a $5 billion enterprise value. But 'sold for $5 billion' is another comforting rounding. Verizon's own release lays out the terms: it received $4.25 billion in cash, plus $750 million in preferred interests, and kept a 10% stake.⁵⁶ Set the cash Verizon got out against the cash it put in, and the picture is plain: against roughly $8.88 billion in purchase prices, the realized proceeds were a fraction of that, and the net destruction lands somewhere around $4 billion. The unit reverted to a familiar name — Yahoo — owned by a private-equity firm, as if the whole six-year detour had been quietly reversed.

Wasn't this a reasonable bet that just didn't pay off?

The fair objection is that hindsight is cheap. In 2015, the logic looked sound: mobile data was exploding, every byte of it crossed Verizon's network, and capturing the advertising value of that attention — rather than just renting out the pipe — was the dream of every telecom on earth. AT&T was making the same bet, in larger numbers, with Time Warner. Diversifying away from a slow-growing, capital-hungry utility business into high-margin advertising wasn't stupid; it was the consensus play. And the assets weren't junk — there really were ~$5 billion of them left at the end.

But the steelman is exactly where the error shows clearest. The bet wasn't lost to bad luck or bad execution; it was lost to structure, and structure was knowable in advance. Advertising was already a winner-take-most market with two entrenched winners by 2015 — that was visible in the share data, not just in the rear-view mirror. Buying the third- and fourth-place properties and stapling them together does not create a third winner; it creates a larger also-ran. Verizon's distinctive skill, deploying capital to build the best physical network, was non-transferable into a market where capital wasn't the constraint. The detour didn't fail because the dream was foolish. It failed because Verizon mistook a market it could buy into for a market it could buy its way to the top of.

There's a quiet symmetry in the timeline. While Verizon spent six years and roughly $4 billion learning that it couldn't out-buy Google, its network kept doing the unglamorous thing it has always done — winning the J.D. Power study, again and again, 32 times, then 35.⁷⁸ The advertising adventure was loud, strategic, and forgettable. The award streak was dull, narrow, and worth everything. The lesson Verizon paid $4 billion to relearn is the one its own balance sheet had been telling it all along: the moat you can't buy is the one that's actually yours. Spend on widening it, not on escaping it.