> "Imagine that you have to place a bet between two new companies as to which one is going to experience several years of very high growth, to one day become a billion dollar company. Now imagine that you know nothing – zero – about either company except that they were founded in the same year, and that one has raised $10 million in venture funding, and the other has raised $0."

Well, it's a great thing we're never, ever, ever in that situation!

> "After all, there are a lot of attributes that predict success in a startup. Highly talented founders, large market opportunity, a great idea, experienced advisors, etc. etc. We can debate this list as much as we want; the fact is none of these can be looked up in Crunchbase."

Well shit guys, it looks like Crunchbase is the only reasonable way for reporters (or even the general public) to find out about a startup, its team, its advisors, its board, and its market. Let's pack it up and go home.

This whole thing sounds like "VC dollars raised is not a bad metrics because it's easier than actual research".

The metric that 0.2% of all companies represent a whopping 16% of winners is not a relevant statistic. It's based on the premise that there are 500,000 new startups in the U.S. every year. I'm guessing this number is counting every incorporation which means many companies would be double or triple counted. 500,000 separate companies would mean that 1 out of every 7 people are starting a new company every year. Regardless, this does not represent the total field of companies that a reporter can choose from to write about. The percentage of relevant companies in the 1000 that are VC backed is going to be much higher than the percentage of relevant companies in all 500,000 incorporated.

I also don't agree that a reporters job is to predict winners. If you're beat is startups, then in my opinion, it should be enough to simply cover companies that your userbase would find interesting or useful. As long as the story succeeds in delighting your readers there's no shame in reporting on a company that ultimately fails.

The graph here doesn't prove causality (or even really a correlation).

On the one hand, PG argues that funding is not a good indicator of success, and on the other hand, the author here is saying look how successful all of these funded companies are! It's a false equivalency.

The author is doing exactly what PG talked about, glossing back over history with 20/20 hindsight. As indicated, even the prestigious YC has ~75% of returns from 2 startups.

The article doesn't talk about how many companies failed, spectacularly, in order for those other organizations to succeed. Cuil flamed out $100M+, and there are countless other examples.

I'm not saying the author is wrong, only that this didn't prove causality to me; ergo funding=success.

I think the author misunderstands the reason entrepreneurs are annoyed by the tech press' preoccupation with VC-funded startups.

Bootstrapping today is much less likely to be a failure-to-raise-VC than a conscious choice. Capital efficiency has skyrocketed but the press clings to the idea that capital needs legitimize companies.

There are plenty of other publicly-accessible metrics to identify bootstrapped "companies that are relevant and interesting." Unfortunately, most reporters remain ignorant of them.

So here's a few:

- Traffic: Quantcast, Alexa Rank, Hitwise - Traffic Value: SEMRush, AHrefs, Spyfu - Authority Metrics (links): SEOMoz, Blekko host rank, PageRank - Social Metrics: SharedCount