Top Ten Lessons from the Internet Shakeout

24 februari 2003, 22:07

The past boom and bust of the Internet sector is one of the biggest business events of the past several decades. In the interest of finding lessons that help us avoid similar debacles in the future, here are ten observations about the dot com shakeout.

1) Nothing changes overnight. The single most fatal miscalculation investors made regarding the Internet was to massively overestimate the speed at which the marketplace would adopt dot com innovations. That assumption of speed dictated the rapid pace and scale of investment by both VCs and public investors – and the resulting over-investment led to the inevitable bubble and bust. We somehow believed it was different this time. It wasn’t. It will always simply take time and lots of it for people to integrate innovations into the way they do things.

2) New stuff doesn’t replace old stuff. History tells us repeatedly that innovations almost never replace existing products but rather typically worm their way into the mix and inhabit their own niche. Yet, many dot coms and their funders persisted in modeling businesses that assumed a zero-sum game in which, say, online retailing displaces a significant percentage of existing retailing. In retrospect, all we had to do was look at the history of catalog marketing to predict that e-tailing might wriggle its way into some minority of purchases, eventually reaching its natural saturation point. Recognition of historic precedent could have spared some large and costly investments.

3) Too early? Too bad. Timing issues continually pop up in the post-mortem of the dot com shakeout. Many of the web’s wrecks came to market with high-cost products well before the infrastructure was ready to receive them. The digital entertainment category is one good example. Companies like,,, Digital Entertainment Networks and Pseudo Networks all may have had good products, but they were much too early for the broadband marketplace.

4) Many startups were fundamentally uncreative and “un-Internet”. Many failed Internet startups began with ideas that involved little more than shoveling an existing business model onto a web site – or copying another company that did it. Just as “shovelware” in the content world involved transfer of magazine or other traditional media formats directly to the Internet, so too did much e-tailing simply export catalogs to the web. Online retailing of “stuff” is perhaps the most obvious and uncreative use of the Internet, and like shovelware, it largely fails to take advantage of the interactive features that give the Internet its power. The more creative – and sometimes successful – e-commerce startups leveraged Internet tools to produce such innovations as pricing “bots,” collaborative purchasing, person-to-person trading, e-procurement systems and name-your-price bidding systems for perishable inventory.

5) All we, like sheep, will go astray (with enough pressure). Amid speculative bubbles that last as long as the dot com one we have recently witnessed, even the most disciplined investors can conclude that the rules really may be different this time and eventually give in to the wicked ways of the herd. Ironically, it is many of those most righteous hold-outs that inherit the iniquity of us all – those that capitulated and invested just as the bubble was about to burst lost both their shirts and their integrity. By contrast, the prodigals that jumped on fads with the alacrity of 13-year-olds had already cashed out handsomely. Few of us are immune from speculative frenzies.

6) Free is folly. The junkyards of many innovation cycles are piled high with business plans built around the idea of giving something away free and “making it up on XYZ.” A decade ago while I was working at Ziff-Davis we received a business plan that called for giving away free fax machines and “making it up” on faxed advertising. That’s only one step sillier than giving free exercise machines to health clubs in order to sell advertising blasts to sweaty boomers. The numbers simply don’t work out for most free models. The Internet’s low incremental distribution costs nourished a large crop of freebie wannabes – and now the “F” section of our shutdowns list is very long indeed with names like; Freerealtimeworld; Freeride;;; – you get the picture. Next time around – we’ll focus on our value proposition a bit more closely.

7) We used narrowcast to broadcast. A surprising number of entrepreneurs, presumably in the search for the big play, decided to use the Internet, the ultimate narrowcasting medium, to reach the widest and most undifferentiated consumer markets imaginable. In using the WaterPik of the Internet to water the broad consumer garden, entrepreneurs bypassed the many rich demographic lodes that the Internet enabled them to mine for a fraction of the cost of the big play. Many of the big and broad consumer startups ranging from Value America to Webvan went aground on the inherent low margins and the massive marketing and infrastructure costs of such ventures.

8) The $50 million rule can kill. Many dot com casualties fell victim to the temptation to gin up business plans to meet the size criteria of the typical venture capitalist. A typical VC firm, in order to justify the time it spends on an investment, needs to dispense fire-hose amounts of cash, implying that the recipient business must be fairly big, able, say, to generate revenues of $50 million in three years (hence the $50 million rule). The resulting dynamic creates a sort of theme park of co-dependency – VCs dangle big carrots to encourage bigger thinking on the part of entrepreneurs whose DNA already is programmed for grandiosity. The sad result is that many of these inflated business plans were overfunded. They were never destined for the fifty-mil world, but would have made nice $10 million to $20 million businesses had they been more appropriately financed. In retrospect, angel investing, with its ability to funnel smaller jets of funding, would have been more appropriate for many of shoulda-been-niche plays.

9) It’s hugely difficult to build chicken and egg simultaneously. Many of this past year’s disasters stemmed from business models that required the startup to build both a critical mass of buyers and a critical mass of sellers – and do it at the same time. Many B2B marketplaces fell into that category, as did collaborative purchasing models, rewards programs and many others. It requires huge amounts of money to create either half of the equation in a many-to-many model. Investors that want to create the next eBay had better plan to spend a lot of time and even more money to do it.

10) Prediction tools must improve. As we observed above, the biggest mistakes of the dot com bubble were mistakes of timing – of misjudging the speed and direction of development and adoption and placing investment bets accordingly. In order to avoid those mistakes in the future, we need better predictive tools to plot the speed at which new technologies will spread. Spreadsheet gymnastics by 20-something b-school graduates should not dictate our investment decisions. We can produce better predictions. We have the data – from decades of technology innovation. We have the ability to analyze the data – after all, Everett Rogers wrote “Diffusion of Innovations” 40 years ago. We have the history. The dot com bust suggests we should begin to learn from it.


Marco Derksen
Partner bij Upstream

Oprichter/partner Upstream, Marketingfacts, Arnhem Direct, SportNext, TravelNext, RvT VPRO, Bestuur Luxor Live, social business, onderwijs, fotografie en vader!


Marketingfacts. Elke dag vers. Mis niks!