A year ago, the people I knew at dot-com startups were ordering Aeron chairs and arguing about which espresso machine to put in the kitchen. This year, the kitchens are empty and the chairs are listed on eBay. Pets.com is gone. Webvan is gone. eToys is gone. The list of obituaries is long enough that you stop reading them after a while, the way you stop reading the names in a war memorial when you realize the wall keeps going.

I work at Knight Ridder Digital in San Jose, where I run the engineering organization that builds and operates technology for Knight Ridder’s newspaper websites, philly.com among them. I am writing this from inside an industry that did not get rich during the bubble and is not getting wiped out now, but is getting rattled hard. I want to put down what I am seeing while it is fresh, because I think we are about to spend the next several years rewriting the story of the last five, and the version we end up telling will not be the version that is true.

The view from a newsroom is different from the view from Sand Hill Road

When the trade press writes about the crash, they write about it from the perspective of venture capital. Burn rates, runway, down rounds, founders walking out of buildings carrying boxes. That is one true story. There is another true story, which is what happened to the people who were doing internet work at companies that already had revenue.

We were not getting stock options worth a fortune. We were also not depending on a Series B to make payroll. The newspaper paid the bills with classifieds and display ads and a loyal subscriber base, and the online operation was funded out of that. When the bubble was inflating, our biggest problem was that we kept losing engineers to startups that promised options and free lunches. When it popped, our biggest problem became something else entirely: the online ad market collapsed, our parent company got nervous, and the strategic conversation shifted from “how do we grow this” to “how do we justify this.”

In some ways the engineers who came back to us from failed startups in 2001 were a gift. They had learned things at high speed under pressure that would have taken years to learn inside a stable company. They had also learned, often the hard way, that a great idea badly executed is worth less than a mediocre idea shipped to real users.

What actually killed the companies that died

I have spent a lot of time this year reading post-mortems and talking to friends whose companies cratered. The pattern is more boring than the press makes it. The companies that died did not die because the internet failed. They died because they spent more than they made, for too long, on things that were not going to ever earn it back.

Pets.com was selling forty pound bags of dog food at a loss and paying to ship them. Webvan was building automated warehouses in cities where they had no customers yet. eToys was doing television advertising during the Super Bowl on the assumption that the next two Christmases would scale to make the math work. None of these were stupid people. They were smart people running a strategy that only paid off if growth never slowed, and then growth slowed.

The deeper lesson is the one Warren Buffett has been muttering about for years and that nobody listened to during the run-up. A business has to eventually produce more cash than it consumes. The internet did not repeal that. It just hid it for a while behind a wave of capital that was looking for somewhere to go.

What survives is what was useful before the money showed up

The websites that are still here in late 2001, and that will probably still be here in 2005, are the ones that solved a real problem for real users in a way the users would have paid for, or that ad supported, even if the IPO market had never existed. eBay. Amazon. Google, which is small but quietly making money. The newspaper sites, including ours, which have actual readers who keep coming back because we publish things they want to read.

I think the same will be true of the internal tools we are building. At Knight Ridder we have been working on Cofax, an open source content management system that drives the publishing workflow for several of our newsroom sites. Cofax did not get built because someone raised a round to build it. It got built because we needed it, and our sister papers needed it, and we were tired of every property reinventing the same wheel. That is the kind of work that will outlast the noise. Useful things that people would have built anyway, and that get used by people who actually need them.

The lessons I want to remember when the next wave comes

Some of this is going to sound obvious in five years and is not obvious now. So I am writing it down.

First, every speculative wave ends. The web did not end, the wave around it did. There will be other waves. Broadband is rolling out, devices are getting cheaper, the underlying capacity for what people can do online keeps growing. The next wave will look different and the same people will say it is different this time. It will not be different this time.

Second, the companies that get through a downturn intact are the ones that were sober during the boom. Sobriety is not glamorous and gets mocked when everyone else is having a great quarter. It pays the rent later.

Third, talent comes back. The engineers and designers who left for startups during the bubble are calling around now, and the ones we get back are better than the ones we lost. People learn more from a failed venture than from a steady job, if they are paying attention.

Fourth, there is a personal version of all of this. The way you act toward people during the boom is the way they remember you during the bust. I have a phrase I have started saying out loud to myself, that victory is winning people over, not defeating others. Startups that treated their staff and partners like disposable inputs during the run-up are finding it very hard to get help now. The ones that built relationships still have a network to draw on.

The dot-com crash is not the end of the internet. It is the end of a particular financial story we were telling ourselves about the internet. The work continues. The websites continue. The people who were here doing it before the money showed up are still here, and a lot of the people who showed up because of the money are leaving.

That is not a tragedy. That is the tide going out, and the rocks underneath being visible again, and the people who actually want to be here looking at each other and saying, alright, now back to work.