Ran across a timely academic research on the first internet boom/bust - aka the “dot com” era. Goldfarb, Kirsch, and Miller has published a huge amount of material on the era. Furthermore, they are widely considered the leading business historian on the time between 1998-2004 (really between Netscape IPO and the rise of web 2.0). In fact, they are major contributors and curators to the Business Plan Archive. The 50 page report is titled Was There Too Little Entry During the Dot Com Era?.

Seems like deja vu lately with a very public discourse on the simultaneous rise of web 2.0. bubble and bust (so which is it?) between Arrington and Fred Wilson. With commentary from NYT and Zoli.

Perhaps its because the paper is 50 pages of equations, regression, and bayesian analysis; no one picked up on its publication. But I do find that it provides a eye openning analysis of the era and how we have learned (perhaps not individually but as an institution) and applied it to web 2.0.

Before I jump in, I’ll have to say that I’m biased. I had a Flatiron partner as my board member during the dot-com era. And I myself was roasted by the Fucked Company mob (sometime deservingly but other times not). I dont feel so bad now, as the Pud’s company, Adbrite, is now run by a CEO who was also tarred and feather by Pud’s creation I believe studying failure is important, but even more importantly, it needs to be constructive (like Goldfarb, Kirsch, and Miller). The Techcrunch Deadpool serves an important purpose but it has the potential to degenerate into useless personal attacks and rumor mongering. Either way, Arrington is like a hedge fund manager, he likes volatility and profits from it . . . the Deadpool is his hedge or even an instrument to profit from the web 2.0 cycle. (dont get me wrong, there is nothing wrong with it, I would be doing the same if I’m a jounalist . . . no he is not a blogger, not anymore)

The paper tries to prove a few points

1. “Get Big Fast” was the predominate strategy for venture startups and it created a cascading effect commonly called house of cards or greater fools theory in which assumptions build on top of other assumptions driving irrational behavior.

2. Ironically, the GBF strategy created a huge concentration of investments in a few large companies - belief that first mover + funding amount is a barrier to entry which prevented more startup in any space from getting funded or even started.

3. As a result, eventhough the AMOUNT of venture investment might have been irrational, the number of startups was actually lower than ideal.

A few surprising numbers crunched by Goldfarb, etal support this

Exit rates of dot com firms are comparable with or perhaps lower than exit rates of entrants in other industries in their formative years. Five year survival rates of Dot Com firms approach 50%.

Survival is unrelated to the receipt or the amount of private equity financing. VC-
financed and other privately financed firms were neither more nor less likely to survive. There is no evidence that return on private equity investment was positive or that, conditional on survival, internet traffic ratings was higher for private equity-backed firms.

In more detail,

Annual exit rates for autos during 19001909 averaged 15%, 21% during the 19101911 shakeout and 18% during the period from 19101919. The annual exit rate from the tire industry during 19051920 averaged 10%; it was 30% during the shakeout in 1921 and 19% during the period from 19221931. The exit rate from the television (production) industry was 15% during the period 19501952. Finally, the exit rate from the penicillin industry was 5.6% during the period 19431954 and 6.1% during the period 19551978. These numbers suggest that the exit rate for Dot Com rms is in line with other emerging industries.

In sum, the survival analysis shows that in our data private equity investment is not related to firm survival and this result is robust across many specications. Moreover, we also found unremarkable IRR and no relationship between web traffic rankings and the receipt of VC funding. We interpret these results as consistent with the hypothesis that pursuing a GBF strategy was, on average, a poor strategy for most internet businesses during the late 1990s.

Well, you know where I’m going with this . . . we are smarter than we thought! Look at the current web 2.0 modus operandi . . .

- small investment is not neccesarily bad (as large funding round does not equate success)
- VCs are not the pre-dominant source of equity capital
- GBF is not the preferred strategy
- double or triple the # of competitors in any space compared to dot-com
- re-cycling or old 1.0 ideas (cause ideas failed for reasons other than the concept itself)

In short, failures are the REQUIRED by product of needed experimentation to exploit an opportunity as big and important as the Internet. Bubble or not, deadpool or not - the opportunity justifies irrationality as the rational thing to do.

To be clear, we do not posit that there was insucient investment in internet ventures. Rather in the absence of a belief cascade, more entrants might have received smaller amounts of funding. To envision how these events might have unfolded, consider the case of Webvan, a $1 billion internet grocery venture that entered many major cities in 1999. Webvan turned out to be a spectacular failure. Absent beliefs about the necessity of GBF, we might have observed many smaller-scale startups all experimenting with different models perhaps in different cities delivering grocery products to the consumer. Instead, we observed a single very large bet on one particular delivery model. In general, mistaken belief in GBF concentrated too many resources in too few ventures. In this sense, we argue, there was too little entry.