The Launch and Year 1: May 1995-April 1996
Medscape was launched to the world on Monday, May 22, 1995. A few days later, Bill Seitz, SCP's manager of electronic medical information, posted the following notice on the Internet's "usenet" bulletin board/discussion system:
Medscape(sm) -- the online resource for better patient care.
A new, free Web site for health professionals and interested consumers. Practice-oriented information is peer-reviewed and edited by thought-leaders in AIDS, infectious diseases, urology, and surgery. Highly-structured articles and full-color graphics are supplemented with stored literature searches and annotated links to relevant Internet resources. From SCP Communications, Inc., one of the world's leading publishers of medical journals and medical education programs.
At SCP's regular Friday afternoon party at our 1915-era factory loft in Manhattan's Chelsea district (New York, NY), we raised a toast to Medscape and feasted on chips, beer, celery, and peanut butter. A few feet from the party table were our Medscape servers, a few Macintosh computers piled atop a desk with a bundle of exposed wires that the office cats, Boots, Felix, and Chelsea would occasionally play with. The computers seemed to be celebrating in their own robotic way: A few hundred people had discovered Medscape, and thousands of the articles that we posted had already been served to a new audience of readers -- Medscape's first members.
SCP had been on the Internet since 1992 using it for email, and on the Web since September 1993, when Marc Andreessen, a graduate student at the University of Illinois at Urbana-Champaign, released the first Web browser for the Macintosh. Prior to committing to building Medscape, the SCP editorial, sales, and technology team had spent a significant amount of time trying to identify ways to get into electronic distribution of its clinical content through strategic partnerships, before finally deciding that we were probably best off launching into the business on our own, and doing it on the Internet. These early efforts were not universally supported: One member of SCP's Board of Directors stated that we should "stop playing with computers," cut our losses, and focus on our core business. Following the hugely successful initial public offer (IPO) of Netscape a year later in August 1995, he changed his mind.
So without a formal dedicated budget but keeping track of our time and costs, the SCP skunks started working on Medscape in March 1995, completing the initial site in about 3 months and simultaneously preparing sales and marketing material, including an educational print newsletter, Internet Medical Marketing , for potential sponsors in the pharmaceutical industry.
Few potential commercial supporters of Medscape knew much about the Web site Medscape in 1995. So a print publication, Internet Medical Marketing , was created to educate the audience, and was published until 1998.
Few potential commercial supporters of Medscape knew much about the Web site Medscape in 1995. So a print publication, Internet Medical Marketing , was created to educate the audience, and was published until 1998.
Medscape in May 1995 was a shadow of the comprehensive site that it is now. But many of the features found on Medscape, 2005, were present from the start and are visible -- albeit much improved -- today. Here's what an early visitor to the site would experience:
The Web -- especially the medical Web in 1995 -- was a much more unreliable place than it is today. At the time, no major medical publisher had any meaningful Web presence. There were a few notable exceptions, but few searches on Yahoo! (incorporated by Stanford University [Stanford, California] graduate students David Filo and Jerry Yang 2 months prior to Medscape's launch) turned up content that was worth reading. The first Medscape homepage linked to dozens of peer-reviewed, full-text articles in 6 syllabu areas: AIDS, infectious diseases, managed care, surgery, urology, and oncology. A member could access content with a single click.
From day 1, Medscape was free, but visitors had to register as members and tell us who they were before they could use the site. Registration was one of the most hotly debated features of Medscape: It added significant technical complexity and was a barrier to people entering the site, as it took 3-5 minutes for users to complete. Nevertheless, we considered registration essential for editorial and business reasons. From an editorial point of view, we knew that in the near future we could use our registration database to automatically send readers to pages of interest, so psychiatrists and social workers would land on pages with mental health content, and Ob/Gyns and nurse midwives to pages with women's health information. And because registration provided a member's email address, we could soon send specialty editions of MedPulse eNewsletters to members -- creating, in essence, electronic minipublications. MedPulse was soon to become the most important driver of visitors to articles on the site.
From a commercial point of view, registration permitted us to create a business model that for the first time would permit us to verify that readers in a target audience had seen a page that contained an advertising message. Medscape would never reveal the identity of the member who had seen an ad -- and it never has -- but we could provide demographic details, such as "275 American urologists saw your BannerLink ad for UriWiz between May 30 and June 15, and 25 clicked on it and were referred to your Web site." In this respect, Medscape could offer a service that was more accountable than professional medical journals. Print journals can only provide advertisers with projected "advertising exposure" data based on survey samples; Medscape would have the advantage of being able to verify the genuine behavior of all users looking at every page.
Medscape was easy to remember, pronounce (in many languages), and spell, so it was relatively painless to type into a search engine or enter the Web address. It also didn't hurt us in the early years that the name sounded much like Netscape, and we sometimes would get callbacks from publishers, bankers, and thought leaders, all who thought that they were being contacted by a much more famous Web entity. Credit for the Medscape name went to recently hired SCP designer, Vincent Keane, a talented musician, graphic artist, and former Vermont ski mechanic, who won $50 in SCP's name-the-site contest.
It's worth remembering that while the nonprofit Pew Internet & American Life Project reports that since 2003 more Americans turn to the Web for healthcare information than any other media, back in 1995, the Pew project to collect data on the syllabu didn't even exist. Online usage was increasing but it was tiny by today's standards. America Online (AOL) had 2.5 million members in May 1995 and was not even providing direct Internet service.
The superiority of the Internet and Web over older technologies that displayed information to users on a computer screen was to be a decisive advantage for Medscape. We believed -- correctly, as it turned out -- that Internet technology would soon be ubiquitous, as it was technically superior and far less expensive to deploy and access than traditional online technologies (such as those used then by AOL and services, such as Physicians Online (POL), that had started the year before Medscape). Although AOL's 2.5 million users in 1995 were impressive for its time, just 10 years later it has been eclipsed by the Internet, used by 222 million in North America, and used by 900 million people worldwide. Today AOL is just another Internet service provider (with a few added features) and provides Internet access to less than 3% of the worldwide audience.
Medscape took its editorial cues from SCP and provided freely available, practical, peer-reviewed clinical material, leaving basic science content to others. We also embraced a concept radically different from other publishers who were putting content online. Most publishers saw the Web simply as a way to put their print magazine online on their dedicated Web site, where, they reasoned, they might eventually figure out how to "monetize the audience." (Translation: Make money though advertising, subscriptions, or both.). In contrast, Medscape, from the start, was going to be a megasite: a brand of which large numbers of visitors could quickly find content from a group of trusted publications alongside original Medscape content.
And so we started a partnership called the Medscape Publishers' Circle . Publishers would provide content to Medscape, and we would assume the costs of reformatting it, putting it online, and marketing it to our audience; we would then share revenue with the source-publishers when the content they owned was viewed. Inspiration for the concept came from practicing a history of the Associated Press (the "AP"), a confederation of newspaper publishers started in 1848 to share content that still thrives today. It was a difficult idea: Most medical publishers today -- like the newspaper publishers of 1848 -- are fiercely proud, independent, and competitive. Many were (and still are) threatened by the notion of computer technology, use it poorly, and would make references to "having printer's ink in our blood." When Medscape launched, there were no members of the Circle . But within a year, we had persuaded 14 journals to participate. Today, 20 societies and over 20 publishers and data suppliers have joined the Circle to provide over 120 publications and databases.
Medscape launched with a pool of about 100 full-text articles, but it seemed like a lot more. This is because each specialty syllabu area could link to any article in the pool. We assumed correctly that in print publications, clinically relevant articles only reach a fraction of the audience that could make use of them. We used the Web's ability to access articles from many different paths to overcome this limitation. By presenting collections of articles by specialty syllabu area (rather than try and create a journal for each syllabu area), editors could select articles for audiences based on relevance and from a variety of sources. So pediatricians, for example, could have easy access to articles on urogenital or infectious diseases that they would likely have never seen in their specialty publications. The Publishers' Circle allowed us to put this cross-distribution idea on steroids. Medscape, of course, embraces this principle today, and provides editors of its specialty syllabu areas a robust selection of articles and features from many sources.
Web design and the nascent field of what is now known as "usability" were also carefully considered. We understood that our audience would only use Medscape if it was fast -- faster than producing an answer with paper and phone-based information retrieval. And Medscape had to be easy to use: Healthcare professionals are not computer geeks. Many doctors couldn't type and found the transition to point-and-click difficult. Among the design principles used at the start and still largely followed today was that no article link featured on the homepage would be more than 1 click away from the article itself. Future research was to show that with each additional click needed to get to a desired piece of information, 50% of Web users leave the site.
One critical and controversial usability principle started at the launch and followed by Medscape today is that every article was to be presented in a single, identical format, with highly structured content links to each section. The "consistent article" mandate eliminated the need for users to learn a new interface depending on the source of an article. Although the benefit may seem obvious to users, most publishers at the time insisted that their online content look like their print magazines. Most publishers avoided the Web altogether, chose CD-ROMs for their "electronic strategy," and attempted to sell image files (most commonly PDF files) of their print publications -- an expensive proposition that found few customers. Some publishers put their PDF files on the Web, resulting in a slow and clumsy experience. Medscape rejected this approach (as well as the PDF files); it was designed from the ground up to present content in a format optimized for fast Web browsing and searching on a relatively small laptop screen.
We also wanted to take advantage of the Web's ability to display pictures. This desire was in conflict with our wish to keep the site fast, because compared with text, picture files are large and display slowly. Our solution: Use only graphics that contributed to clinical understanding, and present small "thumbnails" that were zoomable -- enlarged with a click of the mouse. Many Web sites fail because they are bloated with large graphic files that load slowly. Medscape was relatively snappy from the start.
To take advantage of the interactivity offered by the Web without compromising our desire for speedy loading, we developed a quiz called PicTours, in which users could click on different parts of a clinical image to make a diagnosis.
Medscape's primary target audience was and is American clinicians: a business decision made prior to launch. SCP understood the information needs of healthcare professionals and had success in interesting potential sponsors who wanted to reach them. But significantly -- and controversially -- Medscape was also designed to be open to anyone who registered. Physicians, consumers, or any combination thereof could register and have full access to the site, whether they lived in Portland, Oregon, or Prague, Czech Republic.
Our open-door strategy to get as many people interested in Medscape as possible was in contrast to POL, the leader in the field at the time whose tagline, "for physicians, by physicians," embodied a clear commitment to prevent anyone but an American MD from entering their site. POL was so determined about its "physicians-only" strategy that in its first years, it avoided the Internet and Web altogether: To access it, doctors had to install special POL software on their computers and dial into a private network to verify their identity.
Just as it does today, Medscape at launch welcomed everyone interested in professional-level medical information for both commercial and philosophic reasons: We liked the idea that consumers and non-MDs could finally get the same information their doctors read. As it turned out, and in contrast to the conventional wisdom at the time, many doctors liked that, too.
From a business point of view, we also believed that by acquiring the loyalty of nonphysician and non-US members, we had the opportunity to develop commercial services for them in the future, at little or no incremental cost at the start. The first efforts to do this began shortly after launch, when we started discussions with European publishers, which later developed into a full-fledged effort to start Medscape International -- a dedicated attempt to develop content and services relevant to specific non-US countries and regions.
Because it was on the Web from the start, Medscape from the start had an enthusiastic international audience. In the first 5 years, some 15 international partners joined in to translate and manage local editions of Medscape in several languages.
Medscape was a low-budget operation that had the benefit of years of work spent at SCP developing software and work practices designed to help people collaborate in creating content. To manage the workflow of articles in 1983 -- 12 years before the launch of Medscape -- SCP developed an Article Tracking System (ATS) that permitted articles from various sources to be pooled, and then prepared by different groups of people for a number of different publications. The ATS was adapted to Medscape's needs, and was still in use through 2001.
The initial site was served on Macintosh computers, as Windows NT was in its infancy and we had no expertise in UNIX, the only other alternative at the time. Unlike modern Macs, pre-2001 Macs were not suitable for large-scale Web serving. But they were ideal for our purposes of Web-prototyping and testing our thinly financed business model. Once scalable performance and reliability became significant issues for Medscape, and we realized that we would want to be able to buy certain high-end capabilities (ad servers to dynamically insert ads according to business rules rather than just randomly, template-based middleware for creating sophisticated search capabilities), we knew we had to switch to another computer platform.
Medscape also took an approach to content management that was anathema to publishers at the time. We used human editors to sort, prioritize, annotate, and link content. Instead of leaning on technology at launch (ie, a software-based publishing engine), Medscape was a hand-coded site all the way through 1999-2000. It was economically viable and gave a growing cadre of editors an amazing amount of control on what and how content could be most beneficially presented to their readership.
Medscape's first Web software was a 1994-era Macintosh product called WebSTAR. It had a wonderful feature that permitted administrators to watch, in real time, the progress of users visiting links on the site: You could see when someone entered the site, what articles they visited, what picture-thumbnails they "zoomed," and if they answered the question we asked of all new registered members: "How did you discover Medscape?" Being a fly on the wall and watching what and how people read was hypnotic, instructive, and permitted us to continually Improve the site based on our observations.
There is nothing more important to a writer and publisher than to understand what's of interest to their audience. Twenty years earlier, the editorial premises for Hospital Physician , the first medical journal I edited, was based on my watching housestaff read journals in the cafeteria of Bellevue Hospital (New York, NY), rounding with them, and seeing them rise from a broken sleep on a ward library to fix a stopped intravenous. Now we could watch not 1 but hundreds of Medscape users in real time, see what articles they clicked on, and learn at what point they left the site for other Web surfing.
Registration on the site numbered a few hundred after a week, and increased every day. Of special interest was heavy traffic from new members in Hong Kong, China. We emailed one of the members to learn why they used the site, and what they liked and didn't like about it.
The traffic was emanating from Queen Mary Hospital, one of the largest acute regional medical centers in the southeastern part of China, and a teaching hospital and research arm of the medical school of The University of Hong Kong. The hospital sported high-speed access to the Internet -- unusual for 1995 -- and made it available to several medical departments.
One of the physicians at the hospital had read Bill Seitz's Usenet post announcing Medscape and circulated the news to others in the hospital by email. The department of infectious diseases was especially pleased to discover Medscape and the 3 infectious disease-oriented journals from SCP on the site.
I noted with interest the name of another user early in 1995: George Lundberg, MD, editor of JAMA and the prime mover behind the American Medical Association's (AMA's) nascent Web efforts. The AMA's site started a few months after Medscape was interesting, but you had to work hard to find it, entering the awkward Web address https://www.ama-assn.org/ . If you entered "AMA" in the Yahoo! search engine, the first results would be for the American Management Association or the American Marketing Association.
Medscape's efforts to reach early Internet users were low-key but effective. Office interns were hired to scour the Web for disease and health-related sites. The interns emailed managers and Webmasters, invited them to evaluate Medscape, and asked them to include a link to Medscape if they liked what they saw. Among the hundreds that obliged was Jerry Filo, founder of Yahoo! By the year 2000, some 15,000 other sites linked to Medscape, all at no cost. In addition to the links encouraged by interns, MedPulse was made freely available, and soon became a regular publication on many small medical Web sites (complete with its links to full-text articles on Medscape).
As the first year drew to a close, membership passed 40,000, and some 700 visitors were signing in to the site daily, practicing an average of 9 articles per visit.
By the start of its second year, Medscape had registered more than 75,000 members, had 20 people working on the site, and had 14 publications in its Publishers' Circle . Some critical features, like MEDLINE searching, were not yet available on Medscape or anywhere else on the Web.
Some 1800 new members per week (including 700 physicians) were registering. Medscape had 10 syllabu areas, a daily news feed, 14 members of the Publishers' Circle , and 400 full-text articles -- a tiny number by today's standards, but meaningful in its day. In its first 12 months, 3.4 million searches were conducted and 32.2 million articles viewed.
The data were so good it was intoxicating: The small experiment could indeed change the world. Everyone was working long days and weekends, but the results were clear, and company morale was high. If membership increased at its current rate, Medscape would surpass 1 million members within 2 years, far exceeding the reach of any medical journal in history and most consumer publications. ( The New York Times , for example, has a daily print circulation of less than 1 million.) In Silicon Valley, the influential San Jose Mercury News stated, "(Medscape) offers a wealth of healthcare information to satisfy the cravings of even the most savvy medical news surfers." The nascent industry of Web review sites, most of which had ratings and awards, also noticed, and Medscape got top billings from sites, such as Magellan, InfoSeek, Physicians' Choice, Point Review, Scout, and Six Senses -- sites born in the early enthusiastic days of the Web, but unlike Medscape itself, are gone today.
In our first year, we had done much, and on a financial shoestring did what the giant medical publishers, such as Elsevier, McGraw-Hill, Wolters Kluwer, and the Massachusetts Medical Society and its New England Journal had chosen to ignore. We proved that Medscape had a valuable and loyal audience that numbered in the tens of thousands -- already more than most print medical journals. We created the start of a business model. We released our first advertisers' rate card and were soliciting sponsorship. Moreover, we were enthusiastic about the work and the culture of the company, even if occasionally it meant sleeping over at the office or needing to break the pressure with an extemporaneously arranged game of ping-pong on a table shoehorned into our cramped space.
The decision to require registration was richly rewarded in our ability to learn about members, communicate with them, and provide annonomized demographic information to potential sponsors and potential advertisers.
On the business fronts, we had many appointments to demo the site and provide education to potential sponsors. But no genuine sales were made in this period. Companies that traditionally advertise to doctors -- pharmaceutical and medical device companies -- didn't have a Web presence or Web advertising, so it was difficult for them to support the site, even if they wanted to.
The success and challenges of the first year clarified a number of business issues that caused us to think differently about Medscape the site, and Medscape the business. If Medscape was to continue to grow and innovate, it needed a dedicated financing effort to sustain its start-up losses, and the money needed was beyond SCP's resources. Building sales was going to be a slow process, and Medscape was going to have to come up with another source of money to support the effort. Medscape also needed a full-time staff that went beyond the initial "skunks." As the site's first year came to an end, plans were made to run Medscape in its own company and raise its own money. The decision would have many implications for its future.