February 21, 2011

Real Valley Stories: E-Commerce Analyst or Black Hat SEO?

Editor's Note: Part two in an irregular series of stories from my 12 years in Silicon Valley. The first one was posted in December about my first interview at my first job.

E-Commerce Analyst or Black Hat SEO?

After years of reading about people like Steve Jobs, Larry Ellison and Bill Gates in books, newspapers and magazines, I had finally gotten a real job in Silicon Valley in 1998. Lacking a computer science background, but being HTML savvy enough to administer a small Web site, it was clear I'd be doing a lot of writing, with some coding. But in the Wild West mentality of the late 1990s, at a company with a grand total of three people, myself included, the actual job was something of a work in progress. The company had big aspirations - helping other businesses recognize the full potential of the Internet, and to gain a top Web presence, especially when it came to the new world of search engines, led by names including Alta Vista, Excite and Lycos.

Our first projects surrounded chronicling Silicon Valley History, ranking the top 100 computer magazines through our set of practically arbitrary rules, and one of our biggest tasks, looking at the top 100 technology companies, ranked by market capitalization, and then contrasting the way Wall Street measured them with the way the Web did. Our theory, which was both right and broken at the same time, was that the more links a company had to its domain in the search engines, the more highly the Web perceived them, and how well they would perform in the future, especially to a tech-crazy stock market excited about page views and future growth rather than present realities.

As we ranked companies both by traditional means and new methods, we came up with a new benchmark we hoped would take hold, called IVB (the Internet Valley Benchmark), with a derivative called the iFactor (its Web value divided by sales value). If a company held, for example, the #10 position for both market cap and Web ranking, its iFactor would be 1.0. The real fun came when a company would be much higher in one ranking versus another. For example, if a company with a market capitalization of $1 billion had a Web ranking that would put it in the position of a company that would have a $15 billion market cap when given the same ranking in financial terms, its iFactor would be 15.0. Conversely, the opposite result would be a piddly 0.07. Not great.

This sounds like silliness now, but at the time, it was surprisingly insightful. Apple, which has always had a high following online, with many links (due to QuickTime back then, mostly) scored with a 3.54. In contrast, Compaq, Dell and IBM, Web laggards at the time, were less than one, while Adobe, Symantec and Microsoft had very high scores (thanks to PDF, anti-virus and IE, respectively).

A 21-Year-Old Picking Stocks

In time, with the dotcom bubble in full swing, I Was asked to take this same data and use it to look at Web companies looking to go public. WIth a CNBC-addicted public, including myself on days I worked from home, and my boss (who often came into the office after market close, mid-afternoon Pacific), this became one of the most exciting parts of my odd role. And to be honest, the predictions were very good. I could look ahead at the week's five or ten planned Internet IPOs and hand-pick the best two or three, accurately forecasting which ones would spike to a solid market cap in the first few days, and which would flop. My reports were printed out and handed around the office, including to our sister company's employees, who ate them up like candy.

Thus, the title of e-commerce analyst. My boss had grown enamored with this title being bandied about on CNBC by proclaimed experts, so at age 21, holding no degree, I was starting my day with classes at UC Berkeley, and driving across the Bay Bridge to play Net trends guru.

In parallel, my boss had cooked up some intriguing Web marketing magic, which was dramatically successful at propelling our content to the top of the search engines of the day. Our tactics, including multiple domains linking to one another, white text links and keywords on white backgrounds, massive amounts of interstitial landing pages chock full of keywords to aid search engines along, are head-slappingly obvious violations of Google's SEO guidelines these days, but in the late 90's, without clear rules in place, it felt like the Wild West. It looked like magic, and we thought we could change companies' businesses by applying the same tricks to their own site.

IBM Would Pay Billions for This!

That's where the two tracks got mixed. If we believed strongly in the power of Web links and search engine rankings as forecasting future market cap and sales, then the thought was we could bring our software to the big giants of tech that lacked the Web's leading position. We hoped to take our magic to someone like IBM, promise to add on $1 or $5 billion to their market cap, and simply ask for a percentage of that increase. You know... a few million here or there for our efforts. Of course, this didn't sell very well, especially when third parties instead told us our software was probably worth closer to a hundred bucks than a hundred million. That will lead to our next story, when our fun little experiment got voted down faster than a last-minute makeshift volcano at a science fair.