PIEZAS.

Economy · Business

Nobody wanted to buy Netflix, and they had their reasons

Blockbuster could have bought it for $50 million. Yahoo negotiated for Google and walked away from the table. Kodak invented the digital camera and never pushed it. Microsoft laughed at the iPhone. Not one of them was an idiot.

In 2000, Netflix put itself up for sale for $50 million. It mailed DVDs in red envelopes, had been running for three years and was losing money. Its founder, Reed Hastings, was thirty-nine and had sold his previous software company for hundreds of millions.

He flew to Dallas to offer the company to John Antioco, Blockbuster’s chief executive, and the proposal went further than a sale: the Netflix team would build and run Blockbuster.com as the online rental division.

Marc Randolph, Netflix co-founder, described years later in his book what he saw on Antioco’s face when he heard the fifty million: the corner of his mouth turned up for an instant.

As soon as I saw it, I knew what was happening: John Antioco was struggling not to laugh.

Marc Randolph, Netflix co-founder, in That Will Never Work

Barry McCarthy, then Netflix’s chief financial officer, put it more bluntly afterwards: they were just about laughed out of the office.

Antioco denies it. He maintains he wasn’t at any meeting where selling Netflix came up, that the offer was off the cuff, and that it didn’t happen the way it’s told. It’s his version against Randolph’s and McCarthy’s, and it’s worth knowing that before going on.

Ten years later, Blockbuster filed for bankruptcy. Netflix no longer rents films: it produces them, wins film awards and, at the close on 3 November 2026, was worth some $260.4 billion on the stock market. Some 5,208 times what it was asking that morning in Dallas.

It’s tempting to read these stories as a parade of bad businessmen. They aren’t. Almost all of those who said no were competent managers and sovereign owners of their markets, deciding with the information they had at the time.

What’s interesting isn’t their clumsiness. It’s something considerably more uncomfortable: why disruption is almost impossible to see from inside the company that has already won, and why judging it twenty years later, knowing the ending, is the easiest and most dishonest way to tell it.

Who’s telling this

The framework comes from Uri Levine, who founded Waze and sold it to Google, and was later involved in Moovit, which Intel ended up buying. In Fall in Love with the Problem, Not the Solution he devotes a chapter to disruption and defines it in a way worth holding on to: it isn’t about technology, it’s about changing the market’s behaviour. And he adds a line about the road travelled by whoever arrives to break something:

First they laugh at us, then they ignore us, and then we win.

Uri Levine, founder of Waze

Levine has an unusual authority here: he’s one of the few who has sat on both sides of the table.

The four noes

Chart 1

The four noes, card by card

Each card carries the quote, the year and the figures detailed further down in the text.

Blockbuster 2000

Netflix was “a very small niche business”

As Netflix’s Barry McCarthy remembers it · Antioco denies it
Asking price $50M
Market value · 3 Nov 2026 ≈ $260.4B
×5,208 Times the
offer
Microsoft 2007

“There’s no chance the iPhone is going to get any significant market share”

Steve Ballmer · on the iPhone
His calculation · software share 60-80% of the 1.3 billion phones sold each year
Rather than · phone share 2-3% that Apple might scrape together
Kodak 1975

“It’s very pretty, but don’t tell anyone about it”

Line attributed to management · attribution not settled
The fact · December 1975 Steve Sasson, a 24-year-old engineer, takes the first digital photograph
Afterwards Patents the invention; the patent lapses in 1997. Files for bankruptcy in 2012
Yahoo 2002

Negotiated to buy Google and walked away from the table

On Google
2002 The negotiation broke down over price. There are two versions of the figures and neither is settled
Ending · 2017 Yahoo’s business ends up sold to Verizon for $4.48 billion

Quotes and data cited in the text · Netflix’s market value: Nasdaq close on 3 Nov 2026 times the shares outstanding at 30 Sep 2026 (Netflix 10-Q). The value and the 5,208× are my own calculation: 260,400 ÷ 50.

Why each “no” made sense

01 / 04

Blockbuster

Their logic

In 2000, Netflix had fewer than 300,000 subscribers, was burning cash and was on course to lose tens of millions that same year. Broadband was still a minority luxury and the business consisted of mailing envelopes. Blockbuster, meanwhile, had stores on every corner — more outlets than Starbucks at the time — and a scale that’s hard to picture today: Viacom had bought it in 1994 for around $8 billion.

What it didn’t see

Buying a loss-making company in a mail-order niche for $50 million wasn’t obvious: it was a bet.

Note on the figure

According to its 2000 annual report (10-K), Blockbuster had some 175 million shares, and in the third quarter of that year its stock traded between $8.44 and $12.06. On those figures it was worth between roughly $1.48 billion and $2.11 billion on the stock market. That’s my own calculation, and an approximate one: only the class A shares were listed; the class B shares all belonged to Viacom.

02 / 04

Microsoft

Their logic

Ballmer’s reasoning, which is almost never quoted in full, was this: he would rather have Microsoft software on 60, 70 or 80% of the 1.3 billion phones sold each year than the 2 or 3% market share Apple might scrape together, by his estimate. Seen that way, it isn’t foolish. It’s a correct calculation about a market he defined differently. Ballmer was measuring phone share.

What it didn’t see

Apple wasn’t coming to compete for phone share: it was coming to change what a phone was.

03 / 04

Kodak

Their logic

The most unfair of the four. Kodak didn’t hide the invention in a drawer: it patented it, and the patent lapsed on 20 May 1997. What it did was not accelerate. No executive in 1975 was going to cannibalise the product holding up the company for a heavy, slow, low-resolution device. The decision that looks suicidal today was, back then, protecting what already worked.

What it didn’t see

A camera without film wasn’t an opportunity for that business: it was the end of the company as they knew it.

04 / 04

Yahoo

Their logic

From June 2000, Yahoo used Google’s search technology on its own site, according to Google’s IPO prospectus (April 2004). In 2002 it negotiated to buy Google, and in both versions in circulation the talks broke down over price. Yahoo was a portal and its business was keeping the user.

What it didn’t see

That more money could be made by letting the user leave and charging for third-party ads.

Afterwards

By the time they understood it, it was late. They bought Inktomi to react, executed badly, and in 2017 the business ended up sold to Verizon for $4.48 billion — cut down from the $4.83 billion agreed, after two massive data breaches came to light.

The pattern

Four cases, three mechanisms that repeat:

  1. 1

    Fear of cannibalising the business that feeds you today. The film roll, the stores, the portal.

  2. 2

    Underrating the worse but cheaper or more convenient option. Disruption almost always comes in that way, with a product that is objectively inferior at first.

  3. 3

    Measuring the wrong market. Ballmer counted phones. Yahoo counted visits. Kodak counted film rolls. They all measured correctly; they measured the thing that was about to stop mattering.

Chart 2

The market each one was measuring

The unit each company counted, against the unit that ended up deciding the market.

What it measured What decided the game
Kodak Film rolls sold
Photographs taken per year
Microsoft Phone share
Minutes spent using the device
Yahoo Visits retained
Searches answered
Blockbuster Stores opened
Films watched at home

None of them measured badly. They measured correctly the thing that was about to stop mattering.

Diagram · own work, based on the cases documented in this piece.

The question almost nobody asks

And now the uncomfortable part, which Levine frames better than anyone because he had to live it.

He’s constantly asked whether selling Waze to Google in 2013 was a mistake. The company had 51 million users at the time, as Levine himself told Globes in June 2013. Today there’s no revenue figure to set beside it: Alphabet doesn’t break out Waze’s revenue. Would it have been worth more if he’d waited?

There are right decisions, and then there are the NON-decisions. When we make a decision and choose one road, nobody knows what would have happened had we chosen a different one.

Uri Levine

And then the turn that changes everything. Instead of asking what would have happened if Yahoo had bought Google, the better question is another one: would Google have become what it is today under Yahoo’s leadership and vision?

Probably not. Google is what it is partly because Yahoo said no. The same goes for Netflix: it might never have become Netflix inside Blockbuster, run as one more division. It’s the idea Levine closes that chapter with, in paraphrase: any claim about whether those decisions were right or wrong is irrelevant, because we can’t know what would have happened on the road not taken.

The theory, and its critics

The mechanism was described by Clayton Christensen in The Innovator’s Dilemma: leading companies fail not because they manage badly, but because they manage well according to the criteria that made them leaders.

It’s worth not buying it as dogma, though. The very concept of disruption has been sharply contested; the historian Jill Lepore criticised it in The New Yorker for its poor predictive value: it works very well for explaining the past and rather badly for anticipating the future, which is exactly what it promises. It’s a tool for understanding, not a crystal ball.

None of these companies said no out of stupidity. Blockbuster and Kodak went bankrupt, and Yahoo’s business ended up sold, for carrying on doing well what had always worked for them a little longer than they should have.

The lesson isn’t “look how foolish they were back then”. It’s much harder than that: learning to recognise, when your turn comes, that worse, cheaper, more awkward product everyone is laughing at.

How these figures were checked

These stories circulate in distorted form almost everywhere. It’s nobody’s fault in particular: it’s what happens to a story told a thousand times. It gets rounded off, simplified, and given whichever figure sounds best. Here are four of them, and what the sources actually say:

Kodak · 1975

1973 is the year Steve Sasson went to work for Kodak. The digital camera came two years later: Sasson, a twenty-four-year-old engineer, took the first digital photograph in December 1975.

Google · it wasn’t Yahoo

You’ll often read that Google offered itself to Yahoo for $1 million. It offered itself to Excite, which was a different company. In 1999 Page and Brin asked their chief executive, George Bell, for $1 million; they later came down to $750,000. Bell didn’t want it either.

Yahoo · negotiation broken off

You’ll often read that Yahoo refused to pay $5 billion for Google. There are two versions and they don’t match. The one that circulates on blogs says that in 2002 Terry Semel offered $3 billion and Google asked for $5 billion. Semel himself, interviewed by Ken Auletta in 2006, tells the opposite: that it was Google that first asked for $1 billion and then $3 billion, and he mentions no $5 billion at all. In both versions the negotiation breaks down over price, so it wasn’t a “no” to Google; the exact number remains in dispute.

Waze · $1.15 billion → $966 million

You’ll often read that Google paid $1.15 billion for Waze. That was the press figure. Google itself later declared $966 million.

That these discrepancies turn up even in books written by the protagonists of the story says a good deal about how hard it is to hold on to an exact figure once a case becomes legend.

4 November 2026

Correction. The following data were corrected:

Netflix’s value: it said around $333 billion at the end of August 2026 (25 August) and 6,660 times the asking price. That figure was the close of 24 August. It now says some $260.4 billion at the close on 3 November 2026 ($63.42 on the Nasdaq times 4,106,325,903 shares at 30 September, per Netflix’s 10-Q) and 5,208 times. My own calculation.

The Dallas meeting was at 11:30 in the morning, not in the afternoon (Randolph, as reported by Inc.).

Reed Hastings was 39 in September 2000, not 40.

Randolph describes the corner of Antioco’s mouth turning up, not his lip twitching.

McCarthy said Blockbuster “just about laughed us out of their office” (reported by Variety in 2013).

Antioco doesn’t say he only stopped by to say hello. He says he wasn’t at any meeting where a sale came up, that the offer was off the cuff and that there were no serious conversations (Next TV, 3 June 2022).

Blockbuster’s value in 2000: it said $5 billion to $8.4 billion depending on the source. On its 2000 10-K, its stock market value was roughly $1.48 billion to $2.11 billion in the third quarter (my own calculation).

Yahoo: the motive attributed to it (not wanting a search engine that sent traffic away) has no source and is withdrawn. According to Google’s IPO prospectus, Yahoo used its search technology from June 2000.

Kodak: withdrawn are the line “We’re in the paper and chemicals business”, which appears in no source, and the claim that it made money for years from patent US4131919, which is not documented.

First digital photograph: December 1975. The day, the 12th, is not in the sources.

1973 is the year Sasson joined Kodak. The claim that you’ll “often read” the digital camera dates from 1973 is withdrawn.

The version in circulation about Google and Yahoo mentions $1 million, not $2 million.

The 2002 version with $3 billion and $5 billion circulates on blogs; it was described as “the most widely circulated”.

Waze in 2013: it said $1 million in revenue and some 55 million users. Levine told Globes 51 million users on 19 June 2013; the revenue figure has no source and is withdrawn. Also withdrawn: that the Waze user numbers in circulation are third-party estimates, not company figures. Waze publishes its own on its website.

Uri Levine did not found Moovit. It was founded in 2012 by Nir Erez, Roy Bick and Yaron Evron. Planeta, his Spanish publisher, lists it among the companies Levine “has developed”.

Chart 1 showed “$50M $”; it now shows “$50M”.

Ending: it said the four companies died or shrank. That doesn’t hold for Microsoft, and Kodak went bankrupt in 2012 but is still listed on the New York Stock Exchange.

Sources
  1. Uri Levine, Fall in Love with the Problem, Not the Solution — chapter on disruption; definition of the concept and reflection on “non-decisions”.
  2. Clayton Christensen, The Innovator’s Dilemma — the mechanism by which leading companies fail while managing well.
  3. Jill Lepore, The New Yorker — critique of the predictive value of the concept of disruption.
  4. Marc Randolph, That Will Never Work — first-person account of the 2000 meeting in Dallas. The pull quote at the top comes from there.
  5. Statements by Barry McCarthy, then Netflix’s chief financial officer, about the same meeting, in a 2008 interview reported by Variety in 2013.
  6. John Antioco’s denial regarding that meeting, on LinkedIn, reported by Next TV on 3 June 2022: he maintains he wasn’t at any meeting where selling Netflix came up and that it didn’t happen as told. Version against version, not a proven refutation.
  7. Waze’s 51 million users in 2013: Uri Levine to Globes, 19 June 2013.
  8. Ken Auletta, interview with Terry Semel (2006) — Semel’s version of the negotiation with Google, without the $5 billion.
  9. Netflix market value at the close on 3 November 2026: Nasdaq closing price times the shares outstanding at 30 September 2026, per Netflix’s 10-Q. My own calculation.
  10. Kodak dates and figures (December 1975), Yahoo–Google (2002) and Yahoo–Verizon (2017), detailed in “How these figures were checked”.