The million-dollar mystery of milk.com

Can you still make a fortune off a domain name?

An animated computer with a milk carton on its screen.

I don’t remember why I typed milk.com in my browser and hit enter — I just remember being delighted that I did. 

I’d expected the website of the dairy lobby, or perhaps some text that read, “This domain is for sale.” Instead I discovered what appeared to be a quirky personal website. It was full of jokes (“Your lactose pipeline // Established 1994”), random projects (Favorite Recipes, Hourly Kitten), and a link to the resume of one Dan Bornstein, which featured years of software development at companies like Slack, Medium, and Google.

Perusing milk.com, I thought of photos I’d seen of a two-acre wheat field growing two blocks from Wall Street and the World Trade Center. It was an artful act of gross inefficiency — even then, in 1982, the field was worth billions to developers. Milk.com is such a non-optimized, unmonetized use of high-profile digital real estate that it felt like a statement.

Milk.com’s homepage. (Screenshot via Milk.com)

Was milk.com a piece of performance art like the wheat field — perhaps a statement about the corporatization of the internet? Or an amusingly named project with the serious goal of returning to the open protocols of the early net? Or a case of domain squatting gone wrong, foiled by the devaluation of the website in the era of Facebook, Twitter, and TikTok’s walled gardens?

I had to know more. I sent Bornstein an email.

Digital Real Estate

Once upon a time, you applied for domains like you would a spot in a community vegetable garden. They were free, doled out by the Internet Assigned Numbers Authority, an organization within the US government, so long as you gave a good reason for wanting the domain. 

According to records, Bornstein registered milk.com during this era, in July of 1994. It was nine years after the first-ever dot com, symbolics.com, and just a few years after Microsoft claimed microsoft.com.

The Hustle

In 1998, the U.S. government transferred responsibility for managing domain names to a new nonprofit, ICANN, or the Internet Corporation for Assigned Names and Numbers, which was charged with transitioning to a system of private companies selling domain names. With the internet evolving from a research project into commercial infrastructure, the Clinton Administration wanted to encourage competition and global participation. And the domain-name fees funded ICANN’s budget. 

This is when domain broker Dave Evanson bought his first domain. Evanson then owned a consulting company that was working for an investment firm. His team’s PowerPoint had a slide about inheritances that needed embellishing. So Evanson Googled “inheritance.” (Or he may have AltaVista’d it.) He didn’t find much. The website inheritance.com was bare. 

Evanson finished the PowerPoint, but he kept thinking about inheritance.com. He found the guy who owned it and paid him around $5k for the domain. Then he bought several thousand more domains whose names had to do with inheritance, cars, and insurance. His plan was to develop three big, profitable sites on those topics. 

But he never really found time for it, and over the years, he got offers for his domains. The prices trended up. Five years after buying inheritance.com, he says, he resold it for $60k. In the early 2000s, he left consulting to become a domain-name broker. 

He regularly sells domains for six or seven figures, such as angel.com to Angel Studios for $2m. But the biggest sales involve NDAs. 

“One I was involved in was eight figures,” he says. “I can't talk about it.”

Similar to a real estate agent, Evanson has an exclusivity period, often six months, to sell the domain in exchange for a fee.

To gin up interest, Evanson posts about domain names on LinkedIn and Twitter. He reaches out to executives using contact info from databases — perhaps angel investors or VCs in the case of angel.com. (He also has a “pretty good rolodex” from his business career.) 

Young brokers, Evanson says, often sell quickly to investors who want to flip the domain. But Evanson sells to end-users. That’s who pays the most: the company or individual that can most profit or benefit from a one-word or two-word dot com by selling more product or more easily being discovered by potential customers and business partners.

“One of the things I say to companies is, sure, you could do a Super Bowl commercial for 30 seconds. And it'll cost you $7 million, and then you're done,” says Evanson. “Or you could buy this great domain name for $7 million, and then you have it. And you could resell it after you make money with it.”

Evanson doesn’t know of any authoritative estimates of how many premium domains are for sale. But he estimates that over half of domains in the $1m-$25m range are being used for commercial projects, while the rest are owned by investors or in the queue for development. 

Rob Schutz, founder of Snagged.com, largely agrees. From his domain-name work, he estimates that 30%-50% of premium, one-word domains are owned by investors and either listed for sale or awaiting “Godfather offers” that are so high the owners can’t refuse. 

Evanson says that closing a big deal is like solving a complex puzzle. He’s learned that an initial offer from a Chinese company is often a fraction of what they’re willing to pay, while Turkish companies tend to only go up 5%-10%. He’s learned not to discount small buyers, since they can move fast, whereas budget approval within a big company can take months. And he’s seen clients buy a house or put their kids through college with the money they get from selling a domain.

Some factors he considers when valuing a domain name are:

  • Number of industries: Is milk.com like energy.com, which would be valuable to a power company, but also to a company like Red Bull? Or are the buyers limited to the dairy industry? 
  • Comps: Comparisons are often apples-to-oranges, and many sales are not public. But since Evanson (and his employer Sedo) have sold many expensive domains, he can compare milk.com to some similar sales.
  • Global sales: Dairy products are a $1T industry, which makes milk.com likely much more valuable than saffron.com (annual sales of $500B).
  • Foreign use: If “milk” is widely used in major markets like China, rather than the local, non-English word, that can increase the domain’s value.

I ask Evanson how much he thinks milk.com is worth. 

“In the $5 to $10 million range,” he says, after some thought. He doesn’t know that its appeal goes far beyond dairy, but milk sales are large and global, and he says that he sold a four-letter domain recently for millions.

GettyImages-1151378 (1)Compaq’s purchase of Altavista.com was an early example of a premium domain name transaction. Altavista.com now re-routes to Yahoo. (Jeff Christensen via Getty Images)

But selling milk.com for that much would all come down to finding the perfect buyer. 

Like me, Evanson was surprised by milk.com and suggested a few theories for its current state. The owner, Bornstein, had a few free tech tools on the site, so maybe he would one day charge for premium versions? Or maybe he was sitting on a valuable asset, watching it appreciate, and waiting to sell until he needed the money.

Or maybe he was just wealthy and having fun.

Milk Boy

After a short period of what seemed to me like mysterious silence, I received a friendly reply from Dan Bornstein. He was game to talk.

The origin story of milk.com is a time warp. In 1994, personal email addresses weren’t common; Bornstein’s was provided by his employer. He was thinking about leaving the company, and he realized he’d have to email all his pals to say his email was changing. And then do that every time he changed jobs. Not ideal!

The obvious solution was to get his own domain name, a reasonably inexpensive internet connection (dial-up!), and send out that one email one time. He remembers that dan.com and daniel.com were available. But dan@dan.com sounded odd. 

A friend had nicknamed him Milk Boy (“for not-entirely-terrible reasons — I preferred chocolate milk to soda and still do”), so he applied for milk.com. He explained to the Internet Assigned Numbers Authority that he wanted a personal email address, and he got the domain.

In the 30 years since Bornstein applied for milk.com, the domain has served some other purposes:

    • Personal website: It’s a good home for his professional resume and various hobbies — another Dan has the domain danbornstein.com, and Bornstein likes milk.com.
  • Credibility: Within the tech world, people recognize that his ownership of milk.com means he applied for the domain very early in internet history.
  • A dairy empire? Bornstein has leaned into the milk theme by giving his tech projects dairy-themed names. So Evanson was partially right: While it isn’t the main reason Bornstein has the domain, he likes that he’d have a great domain if any of his projects turns into a company.

My elaborate theory of milk.com as performance art was mostly wrong. While Bornstein intentionally maintains the website’s early-nineties look and feel, he says he’s never thought of it as art or a statement. “But you're never supposed to believe what an artist says, right?” he adds cheekily.

Dan Bornstein. (Milk.com)

Instead his main motivation is still the original one: his personal email address. He likes the address, and it would be a pain to change. "You can't just change an email address overnight," he says. 

Occasionally someone will post about milk.com on Reddit or Hacker News. Bornstein sees these posts — and the people who call him a domain squatter: someone who sits on a domain name solely to sell it for big money. (The squatters prefer the term “investor.”) But Bornstein doesn’t worry about the criticism. He’s using the domain, not squatting. 

According to Bornstein, he kept getting emails from people claiming to be domain brokers. If he replied that he wasn’t interested in selling milk.com, they’d keep trying to convince him. In hopes of preventing brokers from emailing in the first place, he added a page to his site saying he wouldn’t sell for under $10m.

“It's actually unclear to me how much I could sell it for if I wanted to,” he says.

I tell him about my conversation with Evanson, and how he valued milk.com at well over $1m or even at $10m. I thought Bornstein might have a big reaction. I was ready for intense facial expressions. 

He was nonchalant: “It's nice to hear, I guess.”

We discuss how a domain broker has an interest in predicting a high price, and how the deep-pocketed “right buyer” may not materialize. Mostly, though, Bornstein feels he makes plenty of money as a software engineer. 

“It doesn't [all] have to be about money,” he says. 

The End of Domains?

In 2012, Google greatly reduced how much their algorithm valued premium domain names. It was a way to bury and disincentivize spammy, ad-heavy websites, but it reduced the SEO of all premium domains. In the years since, websites have been described as “dead” due to TikTok, YouTube, and Google monopolizing attention and keeping users within their ecosystems.

And today, chatbots and AI-boosted search engines from OpenAI, Google, and Perplexity respond to questions and queries not with a list of links, but with a conversational answer. They even encourage customers to audibly ask their chatbots questions rather than type in little boxes. 

Milk.com probably isn’t worth $1m in that future. 

In spite of these trends and Google’s algorithm changes, Evanson says, the price of premium domains has kept increasing as the internet gains more global users — 97m new users came online in 2023 — and more commerce and culture migrate online.

The Hustle

He expects the continued influx of eyeballs will counterbalance AI, at least for some time. But he also cautions clients that new developments may reduce the value of their domains.  

I ask Bornstein how he’d feel 10 years from now if milk.com is only worth a few hundred bucks.

“I think I'd be okay with it,” he replies. 

Rather than enjoy a payday, he prefers to keep the site. He’s an engineer who loved that Usenet, a precursor to the internet, enabled him to meet people he’d never otherwise meet. He loved the early internet’s sense of possibility. 

“I would love it if my legacy was that milk.com [remained] the way the internet should be or should have always been,” he says. “If it’s not my personal domain, as long as it’s part of those old-school ideas about how the internet should be, that would make me happier.”

New call-to-action

Get the 5-minute news brief keeping 2.5M+ innovators in the loop. Always free. 100% fresh. No bullsh*t.