When the end came, it was swift. The phone call asking me to go to the chief executive's office. The trek down the corridor that felt like the dreaded trip to the dean's office at Blackrock College years ago. The short conversation. "We're changing direction, and going back into product development mode. We're eliminating all the positions in the marketing department, and that includes you."

Thus ended my year to the day at a Silicon Valley dot.com. It was a wild ride, full of elation, despair, posturing, positioning, strutting and blind optimism. A typical day started one way and ended completely different. A typical month was nothing like the preceding or subsequent ones.

By the time I joined the company, which we'll call Newco, in August 1999 its executives had just completed the transition from making wild hand gestures and scribbling designs on paper. There was a fully-fledged demo and "beta" or test site.

Commercial partners were being signed up by the "business development team" - never call them salespeople. Advertising plans were being put in place and my job was to prepare for a November launch that would not only introduce Newco to its potential users, but would also get us visibility among potential investors for that all-important second round of venture funding.

Ah, funding. If you'd been reading the breathless coverage of Silicon Valley last year, you would be convinced that a one-page business plan and a photogenic executive team would set the venture capital spigot gushing. Not so in 95 per cent of the cases of course; it was the 5 per cent that was much more interesting to write about.

The Newco chief executive and other executives spent two months pitching the business to moneymen in Silicon Valley and on the East Coast. Several times we were really close to the holy grail of venture funding, the term sheet, but didn't get the money.

In the end, as typically happens for dot.coms, the scales tipped and we had a deluge of funding from six great investment houses and finished up with three times more money than we had set out to get. Given the awful drop in the Nasdaq in April this year, and the subsequent famine in new money, that was an incredible stroke of good fortune.

By December we had launched the site, got coverage in major national publications across the US and secured our funding.

We had a great launch party and we talked about how we'd manage our imminent wealth. Most of us wanted to pay off our frightening Valley mortgages, and those of us with kids wanted to completely fund their education right through university. The rest of the money would go to pet charities, needy family members and some conspicuous consumption.

Looking back, we completely ignored the "risk" part of the risk/reward balance. But it was fun to think about, and it made the daily stress and uncertainty very bearable.

Early this year we were achieving, to use the vernacular, "traction" in a big way. This meant we had hundreds and thousands of visitors to the site every month, we were getting the partners we needed and our investors were happy.

As if to prove our good fortune, a competitor of ours went public and got a stock market valuation of over $7 billion. This was a complete shock to us, and many of us secretly revised upwards the outcome of our own initial public offering, which we had started thinking of for later in 2000.

Then came April and the Nasdaq implosion. Valuations plummeted. And then dropped some more. One investment banker I talked to reckoned that 90 per cent of the aggregate valuations of Silicon Valley companies was wiped out within a 30day period. Pets.com, for example, lost 85 per cent of its valuation and joined the endangered list, having spent $28 million in advertising during 1999.

One magazine started a "Dot.Com Deathwatch" of companies that were running out of money rapidly. A widely watched, and frequently quoted, website, fuckedcompany.com, chronicled the despairing actions of chief executives and funders struggling to salvage their investments. Readers could bet on companies' prospects and get extra points when a company took a turn for the worse.

People's response when I told them I worked for a "dot.com" changed. A year ago, they would immediately ask if there was a way to get insider stock, known as "friends and family" stock. Now the response was as if I had told them I was the local branch chairman of the Donna Summer Disco Will Never Die fan club.

During April our executives and funders took a closer look at the traffic our site was generating. Up to that point, the game was to get as many possible site visitors or "eyeballs" and work out ways to keep them coming back. It didn't matter much what they did, although we had set up an extensive e-commerce section on the site so that visitors could make purchases.

After April, the mantra in the market became P2P, or Path to Profitability. This meant that, for us and many other websites, "eyeballs" were not enough; we needed their wallets too.

This realisation completely changed the game plan for Newco, and we began a major engineering effort to address it and drive more e-commerce traffic. Ultimately, although the four-month engineering effort produced some interesting new technology, some initial market research showed that it wouldn't result in more purchasing on the site. Newco executives therefore decided to go a completely new product direction, which made our marketing team redundant until the new product would be launched.

What went wrong? What happened to those IPO millions we were betting on? The Nasdaq meltdown completely changed the landscape and forced a reevaluation of the business model for us and every other technology company.

More broadly, though, I believe we were expecting our audience to move at the same frenetic pace that we were setting. This was a fundamental flaw in our logic, because our audience didn't have anything like as much of a stake in our success as we did. We didn't have time for gentle, step-by-step adoption cycles.

To make our business succeed, we needed all of them online, buying stuff, right now. This type of adoption works well when one technology company is selling to another, because each side recognises the pace of the other and can match it.

However, with several notable exceptions (eBay, Amazon, Yahoo come to mind), this doesn't work when it comes to websites like ours that require "real people" to change the way they do things. We were mightily successful attracting audiences, but we couldn't change their buying habits quickly enough for our business plan to work out.

And me? Thanks for asking. I'm not sure I want to be a dot.com guy again any time soon, so I'll either work for a more established company or go back to freelance public relations work. This past year has been a fascinating experience though, and even if I didn't make those Nasdaq millions, at least I got a story out of it.

Frank O'Mahony lives, and hopefully works, in Silicon Valley. He can be reached at frank@omahony.com