(This is chapter 16 of my ongoing memoir of the Internet industry.)
Pathfinder Personal Edition, launched in November 1996, turned out to be the biggest and most expensive debacle in early web history. The dimensions of the failure were immediately obvious to Time Warner management, who cut off all plans to promote or continue the service in early 1997. Now all that was left was to fire everyone responsible for the mess.
It was fascinating for me to watch the machinations and weak attempts at defense that now took place. I observed a desperate rear-guard action by a few Pathfinder executives to blame the whole fiasco on Marie Blue, director of publicity and marketing. It’s true that her Personal Edition ads were insipid (they featured a dog who would “fetch your news”) and Marie Blue was quickly fired. But Time Warner top management knew Pathfinder’s problems went deeper than bad marketing. Other heads would have to roll.
As a middle-manager in the tech team, I was too low on the totem pole to be blamed for anything. The big brass weren’t going to fire me; they didn’t even know my name. We in the soldiering ranks of the tech team basement knew our jobs were safe, because the Internet was still burning-up hot and Time Warner was clearly committed to succeeding on the web. It was unclear, though, whether or not Time Inc. New Media or Pathfinder.com would continue to exist in their current form. And it was totally clear that most of our bosses were going to get fired.
They went one by one. Bruce Judson simply disappeared one quiet day. Paul Sagan announced that he was quitting “to spend time with his family”, though he somehow also found the time to co-found the important web technology firm Akamai (like Jim Kinsella’s, Paul Sagan’s luck improved greatly as soon as he left Pathfinder). Technology director Oliver Knowlton, the man who’d hired me, hung on for a while, became the general manager of the wounded organization and then made a deft sideways move across the street to Sports Illustrated, leaving Vicki Zilaitis to run the tech team by herself. I welcomed this change.
Oliver was a genial man and a fair and honorable manager. I appreciated the way he once supported me in an internal battle against another developer who, I knew, he considered a closer friend. As one of the original founders of Pathfinder, I imagine he felt heartbroken by the Personal Edition disaster, but I also believe that he made some of the key mistakes that doomed the effort. One of the biggest mistakes was the decision to build a gigantic infrastructure for e-commerce long before we had any actual customers wanting to spend money on our site.
Perhaps the single best lesson I’ve learned during many difficult web development efforts is this: don’t build stuff you don’t need. It doesn’t matter if you think you’ll need it soon; you shouldn’t build it until you need it now. If you have 50 paying customers, you should maybe build an e-commerce architecture for 500 customers. Once you get 500 customers, it’s time to build an architecture for 5000.
But Pathfinder launched Personal Edition with an expensive architecture for 500,000 paying customers. This left us badly exposed when the customers didn’t show up.
In fact, there’s no shame in launching a bad product, or a product that doesn’t catch on with consumers. Media companies do it all the time. The shame — and the shocking failure of our supposedly sharp executive team — was that we decided to go all-in on Personal Edition’s version 1.0. If we had kept expectations low, test-launched some features, done some focus groups and felt our way through, who knows? Maybe we could have nursed the product along to eventual success. But our management got too excited by itself, and lost their sense of proportion.
Another early mistake was to fall for a big sales pitch by some well-dressed MBAs from a Cambridge, Massachusetts start-up called Open Market. These slick execs must have had a great PowerPoint presentation, because the contract we signed with them mandated that they would provide our entire software environment, from web server to application server to commerce server. This locked us in to a single corporate vendor at a moment when innovation was exploding everywhere else. Netscape quickly emerged as the leader in high-end web software, but we were locked in with Open Market. An open source web server called Apache became Netscape’s biggest competition in the high-performance web server space, but we couldn’t benefit from the great work developers around the world were doing on the Apache platform, because we were locked in with Open Market.
Ironically, Open Market used its relationship with Time Warner (we were their first major customer) to establish itself as a major player in the Internet marketplace. As we suffered and struggled to keep our sites from choking on their clunky out-of-date software, they had a very successful initial public offering. Their stock rose 120% on its first day.
IPO’s were all the rage by 1997. It’s on IPO day that stock options turn into real money, and if a company sets its price at $15 and the market decides on the morning of the offering that it’s worth $25, many individuals involved with that company become instantly wealthy.
As I nursed my own disgust and disappointment over the evident failure of the big media venture I had so excitedly joined two years before, I kept myself optimistic by following the Internet stock market and thinking about my future. A lot of people were doing the same thing in 1997.
The Internet stock market craze, opened by brave innovators like Netscape and Yahoo, was halfway through its second year, and was showing no signs of slowing down. The offerings were getting better, deeper, more clearly useful. Jeff Bezos’s Amazon’s IPO on May 12, 1997 was a big success. The stock soared 30% over its opening price, and now the loud doubters on Wall Street were starting to quiet down as more and more investors clamored to get in.
Since every investor and mutual fund and Sunday golf player now wanted a piece of an Internet company, every Internet company was racing to establish itself to the point where an IPO became possible. That’s why even chump outfits like Open Market were able to get in on the game.
Most of the Internet IPOs in these days were from San Francisco, staffed with talent from Stanford University, Berkeley and the San Jose/Silicon Valley industrial parks. The first New York City-based Internet company to make it onto the IPO map was DoubleClick, an advertising software company whose success surprised me. They competed with NetGravity, whose software we used (and I had helped to select) at Pathinder.
DoubleClick’s angle was ad targeting. Their servers would aggregate all the bits of information they could find about each visitor — domain name, IP address, provider — and use this information to target specific ads based on presumed visitor demographics.
In fact, individual-level ad targeting had always been Bruce Judson’s dream, but I was skeptical. DoubleClick’s founder Kevin O’Connor came to Pathfinder and tried to persuade us to switch from NetGravity, but I advised Pathfinder management that DoubleClick’s ad targeting was just another fad of the day, soon to be forgotten. It turned out I was wrong, because DoubleClick’s approach prevailed, and they would eventually become the central player in New York City’s “Silicon Alley”.
Maybe I wanted to believe NetGravity’s plans were better than DoubleClick’s because I liked NetGravity’s people better. NetGravity’s CEO John Danner was a jovial, approachable and endearingly wonky guy. DoubleClick’s Kevin O’Connor, as far as I could tell, had no personality at all. His face had one expression: blank. He answered questions with single words. I made an emotional judgment that DoubleClick could not succeed because they didn’t have a swell guy running the show, and my judgment turned out to be wrong. DoubleClick eventually acquired NetGravity, then got acquired by Google, and remains today the major player in online advertising services.
Even though my ability to “pick ’em” was not always great, I was catching the fever, and I wanted to play the IPO game. As I heard more and more stories about people cashing in on their stock options, I began to resent the fact that Time Warner was already a public company and thus couldn’t do an initial public offering (we got stock options, as Time Warner employees, but couldn’t benefit from the great price fluctuations that occurred when a new company launched a stock). However, it was also possible to benefit from an IPO by being offered a “friends and family” gift, and a friend of mine had made over fifty thousand dollars by doing a consulting job for Yahoo and getting some “friends and family” shares. I also heard rumors that several Pathfinder managers had benefited from “friends and family” stock when Open Market went public. Where were my friends and family?
Even though NetGravity had fallen behind DoubleClick’s strategic direction, they were also growing fast and were headed for an IPO in a year or so. I’d certainly helped this company grow, especially since they used a quote from me on their ads:
Actually it was pretty clear by 1997 that DoubleClick defined the state of the art in Internet advertising, with NetGravity a close second, but what can you do? I wanted a piece of the action, anyway, so I called my contact at NetGravity. “Can I get in on your friends and family?”
“Sure, we’ll work something out.”
The truth is, I didn’t even think about making money. I just wanted to play the game.
I did some job-hunting. My friend Ken Jordan was the creative director of Word.com, a literary journal run as a loss-leader by a New Jersey tech consulting firm called Icon. He urged me to go to New Jersey where I had a very pleasurable long talk with Icon’s web visionary, Dan Pelson. I liked Dan a lot, and soon afterwards I visited the Broadway and 54th Street office of Word.com and was thrilled when the elevator stopped two floors short of the Word offices and I found myself staring at a large brass statue of Alfred E. Neuman. Apparently Word.com was in the same office building as Mad Magazine.
It was hard to turn down a job where I’d be sharing an elevator with the staff of Mad Magazine, but I felt wishy-washy about the whole idea. Word.com was hardly a runaway success, and Icon wasn’t likely to pull of an IPO anytime soon. Dan Pelson was probably thinking the same thing, because as I contemplated the job change he called and told me to forget about Icon — he was leaving himself to start a new company, a community website targeted at teenagers. He had a good understanding of web advertising and felt (correctly) that this was an audience segment sponsors would want to reach.
I was excited by the idea, and went to meet Dan’s business partner in this venture. But this meeting went very badly. Dan’s partner reminded me a whole lot of Oliver Knowlton and Bruce Judson and the other starry-eyed “build-it-big and build-it-fast” high rollers at Pathfinder. He was full of himself, confident of quick success, eager to do everything at once: an e-commerce platform, a message board platform, a chat service, personal home pages, polls and quizzes, email accounts. When I asked him how he planned to use these services he looked at me like I was crazy. It was a question he didn’t understand.
I hated to do it, but I told Dan Pelson I’d decided to stay at Pathfinder, passing on the chance to become the CTO of Bolt.com. They were a moderate success, though I imagine they must have learned some hard lessons about technology along the way.
I was beginning to feel directionless as a software developer, and worried that I was losing my edge amidst all the silly distractions of the dot-com biz. The only challenging technical project I was working on was my moonlighting gig, BobDylan.com. The most difficult but also most original part of this project was a complete hypertext catalog of every song and album Dylan had ever recorded or written, using Perl and Sybase. Developing this was great fun, and I spent hours on it every day at work (my “moonlighting” didn’t involve much moonlight). Dan Levy and I got along well, and I appreciated his practical and earthbound approach to planning. We had the site ready to go well in advance of launch of the major new Dylan album Time Out of Mind, and I got a nice paycheck for my work.
I’m afraid I didn’t contribute much to BobDylan.com creatively, though. Dan always welcomed my brainstorming, but I was obsessed with my data models and regexps and found, somewhat to my dismay, that the music-listening and code-writing centers in my brain had little interaction. Once the novelty of developing for Bob Dylan wore off, code was code. When the going got tough, I even began to resent Bob Dylan for making me do this work. When the album came out, I found myself getting slightly irritated at its popularity, since it only reminded me of Perl.
I had a similar experience when I met a fascinating person, Hanuman Books publisher Raymond Foye, and ended up building him a website for his books (which I loved). What I loved about his books is that they were tiny, matchbox-sized (it turns out they were printed by a manufacturer of small religious tracts in India) and contained short poems and prose by a unique set of writers representing a cross section of underground culture including Beat poetry (Gregory Corso, Herbert Huncke), punk rock (Richard Hell, Patti Smith), postmodernism (Rene Ricard), Deadhead culture (Robert Hunter), Abstract Expressionism (Willem DeKooning) and Andy Warhol’s factory (Candy Darling). I somehow ended up building a website for Hanuman Books, but I ran out of steam halfway through — between LitKicks, Pathfinder and Bob Dylan I had nothing left to give — and taught Meg HTML so she could pitch in and take over. Luckily, she caught on quick and did a great job.
I was simply burned out. I did feel a lot of pride, though, when I caught Bob Dylan’s show at Jones Beach featuring Ani DiFranco. Enjoying his intense performance, I felt like telling people, “Yo, I built that guy’s website”. Like anybody would care. I didn’t get a backstage pass.
There was mystery in the air in the spring of 1997. Allen Ginsberg died on April 5, and this was quickly followed by a blissful explosion of appreciation on the BEAT-L mailing list and elsewhere on the Internet. I collected a bunch of these posts and assembled them into a page on Literary Kicks, and this has always been one of my favorite pages. Soon after I put this up, the Beat/underground publisher Jeff Weinberg of Water Row Books asked me if I wanted to turn the Ginsberg tribute into a book. I kicked myself later, but I said no. I was too busy with Coffeehouse.
Coffeehouse: Writings From The Web was the name Christian and I came up with for our literary anthology. We thought it was a pretty good title. We’d sent an early 450 page draft to Marjan Bace, and he hated it and sent it back asking us to cut 200 pages. We cried in protest and got busy making cuts (looking back, we should have made more).
But we ended up with an impressive collection, I think. The layout was very attractive, and the cover was pretty good. Some of the pieces are truly classic, and many of them are worth remembering. The book’s thematic presentation was one of the things I thought Christian and I did very well. Here’s the table of contents:
1. Unbroken Chain (stories about parents and children)
— Last Night’s Violent Math by Janan Platt
— Family Lore by Dan Brodnitz
— Seven Times 7 by Galinsky
— The Ice Cream Man Myth by John Labovitz
— Remember the Woolworth’s by Spike Gillespie
— Getting the Hang of Yin and Yang by Gael Morlan
— Life With Father by Joseph Squier
2. Insignificant Others (stories about relationships)
— Blind Isn’t Love by Mia Lipner
— Cat in a Tree by Ben Cohen
— Amy Beauty Rose by Xander Mellish
— Jerry Was by Jamie Fristrom
— Almost Awake by Malcolm Humes
3. The Center Cannot Hold (hypertext/experimental)
— Notes on Sinking by Clay Shirky
— No Bird But An Invisible Thing by Christian Crumlish
— Flux (a collage of Usenet and antiweb posts put together by me and Christian)
— Like Magic by Freeman Ng
— Bigamy in the Desert by Christy Sheffield Sanford
— 8 Minutes by Martha Conway
4. Tangents (four related pieces)
— Gravity by Jason Snell
— The Damnation of Richard Gillman by Greg Knauss
— The Valentine Thief by Carl Steadman
— Intervention by Walter Miller
5. A Thousand Words (visual pieces)
— Touch by Annette Loudon
— Contacting Sun Ra by Scot Hacker
— Very Short Stories by Steve Seebol
— Negative Space by Mark Napier
6. Place (stories about place)
— Water Business by Peter Crumlish
— Art World Rag by Gregory Severance
— A Geography of Smell by Briggs Nisbet
— The Understanding (2525) by David Keith
— Why My Town Is Redneck by Mike Watt
7. Alone (stories about isolation)
— Nigga What Couldn’t Find No Work by Scott W. Williams
— Some People by John Humpal
— Lotto by Alison Dorfman
— The Weight by Steve Silberman
— Fragmented Geographies by Nick Meriwether
— Double Truth On The Second Level by David Alexander
— Dry Popcorn by Sudama Adam Rice
— Farewell Letter by Gregory Pleshaw
8. Can Music Save Your Mortal Soul? (stories about music)
— Maybe Reality Bites But I’ll Take It by Tabitha Rasa
— Keepers by Lee Ranaldo
— Fragments From Africa by Meg Wise-Lawrence
— The Second to Last Night at Skanks by Michael McCullough
— Double Sun by Robert Hunter
— Apparition by Levi Asher
— A Personal History of Alternative Music and the Internet by E. Stephen Mack
APPENDIX A: Finding Good Writing On The Web
APPENDIX B: How To Write For The Web
I had a nagging sense that the raw material was better than the package, but it was out of our hands now. The publication date would be sometime in mid-August 1997.
While we waited for a box of the books to show up at our home in Forest Hills, Meg and I took the kids on a day trip to the waterfalls in Paterson, New Jersey with Greg Severance, his wife Doreen Cragnotti and their kids (another online poet friend named W. Luther Jett also met us there). Our goal was to stand on the bridge over the Paterson falls, the bridge you can see on the cover of the poetry book Paterson by William Carlos Williams, and recite a few lines from the book.
It was a nice cool drizzly day, and we read the William Carlos Williams even though the book got completely wet from the windy light rain. On the car radio on the way up, we’d heard that the great Beat/postmodern/experimental author William S. Burroughs had died. It felt somehow significant to be there with literary friends on this unexpected day, and portents were thick in the air.
Many moving tributes were posted online for William S. Burroughs, as they had earlier been for Allen Ginsberg. Maybe Christian and I would have dedicated Coffeehouse to William S. Burroughs, if Ginsberg hadn’t already taken the spot.