← all stories
1995–1997acquired by Microsoft · 1997

Hotmail

The first mass-market free web email service: a mailbox that opened in any browser, with no internet provider of your own. It became famous not for the product but for the way it grew, since every message carried an advertisement for the service to whoever received it. From zero to millions of users in a year and a half, and at the end of 1997 Microsoft bought the company.

Founders Sabeer Bhatia · Jack Smith
Domains hotmail.com
emailviral-marketingfree-tieracquisitionspam

In brief

Hotmail was the first genuinely mass-market free web email service: an address and a mailbox you could open from any browser, without an internet provider of your own and without paying a dollar to use it. It is known less for the product itself (web email was an idea anyone could copy) than for HOW it grew: every message sent became an advertisement for the service to whoever received it. From zero to several million users in a year and a half, and in December 1997 the founders sold the company to Microsoft; the sum was never officially disclosed, but a figure of about $400 million went around the market estimate. Along the way the company's investors literally coined the term "viral marketing," or at any rate were the first to write it down.

How it started (the founders)

Sabeer Bhatia and Jack Smith met at Apple and then moved together to Firepower Systems, a maker of PowerPC workstations and a Canon subsidiary estimate. Bhatia finished BITS Pilani, won a transfer scholarship to Caltech and a master's in electrical engineering at Stanford; he himself wrote later that he had gotten into Apple on his grades and had built Hotmail while learning on the fly. Smith designed chips for workstations and invented one of the first accelerator cards for web servers: a hardware engineer, not a marketer estimate.

Their first idea together had nothing to do with email. The company was conceived as JavaSoft, a web store for personal files, photos, and documents reachable from a browser anywhere, in effect a proto-cloud a decade and a half before the word entered common use estimate. While they were writing the JavaSoft business plan on Firepower's computers, the corporate firewall blocked their personal email, and that irritation suggested the idea: if the mail itself lived on a web page, the firewall would not tell it apart from ordinary web traffic estimate. By retellings of Bhatia's interview for "Founders at Work," the specific trigger was a call from Smith in his car; Bhatia, worried that the cellular line was being listened to, asked him to call back from a home phone before going on estimate.

The money came hard: by various retellings, between 18 and 20 funds said no before Draper Fisher Jurvetson (DFJ) agreed to invest estimate. The first round closed at $315,000, of which $300,000 came from DFJ, for 15% of the company at a valuation of about $2 million [the $300,000 from DFJ is a fact, confirmed by a third independent source: a Business Week quotation of August 25, 1997 inside Jurvetson and Draper's own essay, found during verification. The 15% stake and the ≈$2 million valuation are still an estimate, resting only on two independent retellings of the Harvard Business School case]. Separately, a retelling of the book "Viral Loop" (2009) mentions a later bridge tranche of $50,000; we could not establish whether this is the same tranche as the $15,000 from Rex Smith in the HBS case or a separate event estimate. The company got its financing in mid-February 1996 estimate and launched on July 4, American Independence Day, played on as independence from provider-tied email estimate. The name was a choice too: Bhatia went through options ending in "-mail" and settled on Hotmail because the letters H-T-M-L hide inside it, and the first spelling of the brand deliberately showed that in capitals: HoTMaiL.

Year-by-year timeline

Lesser-known but significant facts

  1. The term "viral marketing" was coined by the company's own investors, not by journalists after the fact. Tim Draper and Steve Jurvetson of DFJ described Hotmail's mechanic as "viral marketing" in their own essay in the niche Netscape M-Files back in May 1997, seven months before the sale to Microsoft, when the outcome of the investment was still unknown. An expanded version came out in Business 2.0 in November 1998.
  2. The service that the largest maker of Windows ended up buying ran for three years without a single line of Microsoft technology. Hotmail ran on Solaris and Apache on top of FreeBSD; after the purchase Microsoft spent years moving the infrastructure to its own stack.
  3. The idea came out of paranoia about a car phone being tapped, not out of a marketing session. Hearing Smith's first words from the car, Bhatia cut the conversation short himself and asked him to call back on a landline, by his own retelling for "Founders at Work" estimate.
  4. Its own success nearly drowned the company in legal costs. By the spring of 1998 Hotmail was already having to sue spammers who were forging [email protected] return addresses en masse for mailings of pornography and get-rich-quick schemes; the volume of parasitic complaints forced the company to hire staff purely to dig through other people's mail. Virality produced growth and reputational damage together.
  5. The first idea was not about email at all. Before the firewall and web email, Bhatia and Smith were writing a business plan for JavaSoft, a web store for personal files and photos, in effect the idea of cloud storage a decade and a half before the term entered common use estimate.

The first growth lever

The mechanic was extremely simple and cost almost nothing at the time: a line advertising the service, with a working link to sign up, was added to the signature of every outgoing message. Verification found the primary source, an essay by Tim Draper and Steve Jurvetson themselves (Hotmail's co-investors and the authors of the term "viral marketing"), which was hosted on the DFJ site. By their account, Draper proposed the postscript "P.S. Get your free email at Hotmail": shorter, and WITHOUT "I love you"; the team found the idea "very contentious," and in the end it was the "P.S." that was dropped, leaving the advertising line without it. The full phrase that circulated later, "PS: I love you. Get your free e-mail at Hotmail" (as what Draper supposedly proposed and what supposedly ran in production), appears only in retellings from 2009 and after and diverges from the earlier and more primary source; it looks like later mythmaking added in the retelling rather than what actually happened estimate. The idea came from investor Tim Draper, when he asked the founders how they would spread the word about a free product with no money for billboards and radio, and this is now confirmed not only by retellings but by Draper's own first-person essay as co-author. The key property of the mechanic was targeted exposure rather than advertising to all comers: the recipient of a message had almost always already been corresponding with an existing user, and so resembled that user in profile, but was not signed up yet, and what he saw was not a banner but a personal message from someone he knew, with a working link at the end.

The before-and-after figures for switching the signature on differ in detail but agree on order of magnitude: before, a hundred or two hundred sign-ups a day; after, about 3,000 a day, with hockey-stick growth within hours estimate. Then more than 100,000 users in the first month estimate, about 1 million roughly six months after launch, the second million five weeks after that estimate, and more than 9 million at the moment the Microsoft deal was announced in December 1997, which is Microsoft's own official figure rather than a retelling. The peak rate of growth is given in different retellings as either 20,000 or 60,000 new accounts a day, and we could not reconcile those figures against dates.

A refinement to the legend that they grew with practically no budget. In the first pass the figure of ≈$500K on marketing looked like a retelling with nothing behind it. Verification found the primary source: Draper and Jurvetson themselves write, in the essay on viral marketing, the exact words "from company launch to 12 million users, Hotmail spent less than $500K on marketing, advertising and promotion." So $500K is a real figure, documented by the investors themselves, but it covers a narrow line, "marketing, advertising and promotion," and a specific milestone, 12 million users, around the end of 1997. Separately, per a court document of April 1998, several months later and with a broader wording of "marketing, promoting, and distributing," the company claimed ≈$10 million. These are not a mutually exclusive myth and fact but two primary sources defined differently and dated differently: the narrow advertising and promotion line stayed low, and it suited the investors to show precisely that as proof of the viral marketing thesis, while the broader marketing plus promotion plus distribution line, which naturally takes in infrastructure, support, business development, and legal defense against spammers, had grown by an order of magnitude by the following spring. The lesson is not that the legend burst but that a headline figure has almost always already been filtered through someone's definition of the term: the signature in the message really was free as a media channel, but the company itself, the support of a growing infrastructure, PR, business development, and, as a separate line, the legal defense of the brand against spammers cost real money.

Parallels today (projects from the catalog)

What a builder can take from this in 2026

  1. "No budget" almost never means "free," only one specific line of the budget. At Hotmail the spending on advertising and promotion in the narrow sense really was tiny, less than $500K over the first 1.5 years, and the investors themselves confirm it, but the $10 million on support, PR, BD, and defending the brand in court by the following spring did not go anywhere. The zero-budget legend usually quotes the numerator of a single expense line, the narrowest one; budget not for advertising but for what grows along with your users.
  2. The best distribution channel is the product's own exhaust, not a separate marketing initiative. The signature in the message was not a campaign; it was built into what the user was already doing every day. Look for what your product produces by default and turn that into a channel.
  3. Successful virality attracts abuse in proportion. The faster and more freely a brand grows, the more attractive it is to spammers and fraudsters, which follows from the same mechanism that produces the growth. Budget for trust and safety along with growth, not after the scandal.
  4. A curve of viral users is not a business model. Hotmail had millions of mailboxes and effectively zero revenue; the exit was the sale of the company, not profitability. Decide in advance whether reach without revenue is a strategy aimed at a sale or a phase before monetization.
  5. Small craft details in a product pay off as free PR. "HoTMaiL" deliberately showed the letters HTML, a small thing that thirty years later still lands in articles as a ready-made hook for the story.

Discrepancies and what we could not verify

Sources (primary first)

Secondary (context, used with confidence markers):

The same thing, about today

The same breakdowns, but of projects launching right now: what the product is, where the first users came from, how they charge. The card is free, the full dossier is $5 (the dossier itself is written in Russian).

Browse the dossier catalog →