(This is chapter 13 of my ongoing memoir of the Internet industry.)

By early 1996, Pathfinder's day-to-day advertising operations had become a gigantic impossible mess. This was because we were, unfortunately, very good at ad sales.

It was the only thing we were good at. The Pathfinder.com site was a public laughing stock for its stale content and unhip design. Our top management was weak ever since our leader Walter Isaacson ran for the hills (leaving Paul Sagan to keep the train chugging along), and our clunky site was also famous for slow delivery and sudden crashes. People even made fun of our URLs, which contained an early attempt at cookie-like persistence via session IDs, making them look like this:

http://www.pathfinder.com/@@YgCb20QASf3HvD/Welcome.html

The ad team, however, was part of Time Inc.'s larger and world-renowned advertising juggernaut, and they put together a young but powerful sales team for Pathfinder, led by Linda McCutcheon and Charlie Thomas, who immediately began executing on the plan to lure Time Inc.'s blue-chip clients like AT&T, Ford and Merrill Lynch to the web.

Advertising was really the only area where Pathfinder was living up to its name and truly blazing new trails in the Internet economy. Along with our California peer HotWired, we were ahead of the pack in pushing six-figure deals. These deals were no easy sell for either HotWired or us. HotWired based their sales pitches on hype and community, because that's what they were good at.

The Time Inc. division of the Time Warner corporation, on the other hand, had a reputation as an industry leader in precision marketing. The two-way nature of the web seemed to make it a Time Inc. marketing executive's dream come true (and this was why Time Inc. was investing in the Internet). Pathfinder's goal was to sell expensive ad packages to major corporations on the promise of futuristic opportunities like individual-level ad targeting and mass customer feedback. This was all the brainchild of our marketing director Bruce Judson, who I worked closely with and had very mixed feelings about. He was the one who had come up with Pathfinder's annoying session-tracking URL scheme, because he was obsessed with observing user behavior. Bruce had a vision of an ever-flowing fountain of live marketing data -- this, to Bruce, was the entire meaning of the World Wide Web. I found his passion and expertise fascinating, even though his market-control visions often felt repellent.

I gradually made peace with the fact that I had become the technical slave for a mad marketing scientist. I would have felt morally troubled if we were actually doing anything with precision-marketing, but in fact we weren't. Bruce and Linda and Charlie would hint about it in advertiser presentations, but we never actually made anything work. The truth is, it was all we could do to keep our web servers from crashing every day, and our web serving platform had nowhere near the sophistication or the stability to perform advanced marketing techniques.

Here's the dirty little secret about Pathfinder's infamous "session ID" URLs: we never used any of that data. None. I know this for a fact because it was my job to process our log files. Hell, most of the time we couldn't even find our log files.

Pathfinder tech was truly a mess in the early days, though this was through no lack of effort on the part of myself and director of business operations Eduardo Samame. We competed with other parts of the company for attention and head-count, but the production/content team was having even bigger problems than us, and so was network operations. The members of the editorial team, meanwhile, were frenetic and nervous, not sure whether to jump off the boat (like Walter Isaacson had) or hang on for dear life. Chief editor Jim Kinsella had been unceremoniously removed, and now a spirited woman with a Hillary Clinton haircut named Meg Siesfeld took over. I made several friends among the editors -- Craig Bromberg, Kevin McKean, Steve Baldwin, Chris Peacock, Mac McKean, Jack Mason, Noah Robischon, Josh Quittner -- but they were mostly a surly lot, hyper-competitive and dismally pessimistic about Pathfinder's future.

The tech team was an even surlier lot than the edit staff, and amidst all the misery I found to my surprise that the people I liked best at Time Inc. New Media were the ad sales people. I could sit in Charlie Thomas's office and shoot the shit (about browser wars, about marketing, about music or football or anything) for hours. Folks like David Keith, Abigail Hornik, Jeff Huter, Angela Penny, Vanessa Reyes, Jean Cho, Mary Bentley Houk, Margo Demski, Andy Sussman, Christy Cook and Rick Gruber were invariably gracious, fun to be around and appreciative of my work. I didn't see much "appreciative" around the editorial or tech department hallways.

I didn't even deserve the appreciation, because the sales reports we were turning out were invariably erroneous and late (though they had pretty covers). In desperation, I got authorization to hire a junior programmer whose only job would be to keep our log files crunching and make sure the advertisers got their weekly reports on time with a halfway reasonable standard of accuracy. I needed to find a person with a special kind of skill for this position: he couldn't be very ambitious or talented, and he needed to be able to endure an incredible amount of tedium.

I finally established some calm by hiring Alvin Travers, a young Brit fresh from the London pub scene, who was a strange creature. He was obsessed with British 80s technopop and considered the Human League's "Dare" album the pinnacle of musical genius, while the mention of American bands like Nirvana or Pearl Jam could throw him into a petulant fury. He didn't really fit in at Pathfinder, but I admired the unique solution he came up with to the log file workload problem. Whenever he started falling behind, he'd sit around the office all day doing little work, waiting for everybody to go home. By 9 pm or so, the office would be empty, and he'd crank up Human League, settle in with a bag of potato chips and a quart of Coca-Cola and god knows what other stimulants, and have a private rave with the log files until the wee hours of the morning. The next day we'd find him gone, a bag of potato chips and a quart of Coca-Cola in the garbage, and our weekly reports complete. I never knew exactly what Andy did during these long nights, but to this day I can't hear a Human League song without wondering whether the ad sales reports are done.



There had to be a better way, and Eduardo convinced me to have an open mind and hear a sales pitch from a Silicon Valley startup he'd found called NetGravity. They pledged to build the web industry's first ad management software package. I was skeptical because Pathfinder was already suffering through a relationship with a pathetic Cambridge startup called Open Market who'd sold Bruce Judson and Oliver Knowlton some of the worst software I'd ever seen (our ridiculous URL session IDs, which required 20 characters when 6 would have sufficed, were an example of Open Market's work, as was our terrible web server software).

We also heard constant sales pitches from other venture-funded startups who knew a client like Time Warner would ensure their next stage funding. We weren't very popular with industry critics, but we were very popular with wannabe CEOs. By early 1996 it got to the point where I wanted to run when I saw a business plan coming.

But Eduardo had a good feeling about this company called NetGravity, and their CEO John Danner made a good pitch. We signed with them, and a few months later I sat in Eduardo's office with database manager Mike Stoeckel and Eduardo and ran the Unix 'mv' command that turned one of the world's very first full-featured web advertising systems on. It was an exciting moment, and I have to admit that NetGravity came through for us.

We had gotten our advertising act together, but the rough events leading up to this had left me with some permanent conflicts inside the tech group. One loyal and highly trusted young programmer began to criticize my work ethic to others, stating that I did not go all out to solve the ad team's problems as quickly as I could have. He and many others in the tech group had a "sacrifice your soul and body" attitude towards their responsibilities at Pathfinder. They didn't like my cynical and sometimes detached approach, and they especially didn't like that I left every day at 5:30, no matter how bad a crisis was going on. I did this because I liked to spend time with my family, and I was upset to find that I had enemies inside the tech team. This was the first time I'd ever had a conflict with a co-worker that cut so deep we could not solve it by talking it out.

One reason this person disliked my attitude is that I was very dismissive of the big project the tech and editorial departments were taking on for the fall of 1996. Pathfinder Personal Edition was our next attempt to realize Bruce Judson's dream of individual-level targeting: we would charge customers $4.95 a month for the right to customize their own news/sports/entertainment service. While Mike and Eduardo and I slaved away on the advertising system, a much larger team began to tackle the gigantic task of building the web's first personalized news delivery system. I wanted no part of this project -- it sounded very lame to me, and I already had my hands full with the NetGravity software to manage.

As all this swirled around me, my own web sites were continuing to get attention. I'd long been annoyed at the fact that Wired Magazine -- the canonical magazine of the Internet era -- had never mentioned Literary Kicks, so I was pleased when they finally got around to it in their July 1996 issue.



Two weeks later, I was surprised when my friend Christian Crumlish's Enterzone got a great write-up in the New Yorker that also singled out a work I'd created, "Chicken Wire Mother". This was an art/poetry experiment I put together with my old C++ buddy Mark Napier, who was making a name for himself as a digital artist. It was wonderful to read my name -- my new name -- in the New Yorker, and this mention also put a personal spin for me on a long-running family joke. My father Eli Stein is a successful cartoonist whose work has appeared constantly in top publications like the Wall Street Journal, Good Housekeeping and the Chronicle of Higher Education, but he has never been published in the New Yorker, much to his vocal chagrin. So, even though my name just turned up in the fine print here, it seemed to mean something positive for the family legacy. Maybe we'd broken the curse.



With Enterzone in the New Yorker and Literary Kicks in Wired, Christian and I felt pretty hip, and the time was definitely ripe for us to get busy on our book deal. The deal had now been closed, and Christian flew in from California to New York City and we drove up to Greenwich, Connecticut to meet with Marjan Bace, the genteel editor of Manning Books, who was paying us a $7500 advance to produce an anthology of the best writing on the web.

We felt like we were on top of the world, though I wonder if we made a mistake by underestimating the difficulty of what lay ahead. What would we call the book? We didn't know. Could we find enough great web writers to fill a book? Sure ... right? Would the writers give us reprint rights even though we can't pay them? Um ... sure, right? How long would it take to complete? Ahh, we'll knock it out in 3 months via email. Be done by the end of the summer.

Christian and I spent most of the rest of his New York visit celebrating, talking, partying and avoiding work. We probably should have gotten a jump on it.