Death is an ugly subject, whether of people, beloved pets, or a business venture. It is the final end, from whence there is no return and which no heroic effort can overcome.

It is also far more common for startups than many credulous entrepreneurs, investors, and pundits like to admit. Far more pleasant it is to say, sotto voce, "Lewis there was the CEO of MumblePegFlap," without laughing at yet another company name that sounds like it was churned out by someone's satirical Web app.

But young startup failure is a reality, and it's not a bad idea to try to learn from the patterns. CB Insights wanted to do just that so it studied tech companies that died between 2010 and 2013. Here's what they learned: (Tip of the hat to VentureBeat)

Survivor bias exists.

One insight from the report is that entrepreneurs miss an enormous learning opportunity because of survivor bias. Lots of people want to emphasize that they are serial entrepreneurs; few will admit to being serial failures. And you can probably count aqui-hires as examples of failure, because the acquiring company was interested only in the technical talent, not the strategy or product concepts.

In survivor bias, people focus their attention on the winners: Facebook, Twitter, Square, Instagram, and the like. The focus acts like blinders, making people look past the flops that far and away represent the bulk of startups. The result is that everyone is learning lessons only from the winners, which is dangerous, for two reasons. One is that the combination of factors that makes one startup a hit might not be repeatable for any number of reasons: market timing, the particular combination of talents, connections, or what have you.

There are more potential missteps than successful paths, and you need to learn where not to go. But how can you hear about what no one discusses?

It's easy to hide the dead bodies.

Related to survivor bias is the difficulty in knowing when companies have died. As CB Insights puts it, "Many startups are essentially dead but limp along for years in zombie-like fashion." That is, assuming anyone ever heard of them in the first place. And VCs also like to push their losses out of the way in as quiet a fashion as possible. So, identifying failures in an attempt to extract valuable insights becomes even more difficult.

Because of this hidden tech zombie apocalypse, it's dangerously easy to go to a conference or networking event and listen to someone pontificate grandly about how to make a tech business work when the best advice they could probably give is to shout, "Don't do what I did!"

Money is way more important than you'd like to admit.

Startups need capital. However, with cloud computing and services making it apparently less expensive to mount the operations of a new company, you might have the thought that starting a business is easier and cheaper than ever. It isn't.

According to the study, 55 percent of tech businesses that failed had raised at most $1 million; 70 percent of companies that were goners raised less than $5 million. Relying on your credit cards and a few bucks from Aunt Bertha statistically isn't going to do it, particularly when 71 percent of dead companies are moribund within 20 months of their last funding round. You need to have enough capital to start and then quickly find a way to make money and see a healthy cash flow.

Fads may be fatal.

What do the following tech subsectors have in common: social; marketplace; advertising, sales, and marketing; music; and video? They've been some of the hottest sectors for startups, and they've also have the most dead companies in their ranks.

One logical reason is that a hot market attracts more attempts and, therefore, a greater number of ventures that will flop. Another, though, is that in these areas, profitable business models are often the missing element. It's not as though you build a management analytics package and go out and sell it. The way you make money is fairly clear, even if far from easy. But when you hear--from all these people involved in companies that will fail--that you need to get people and then figure out how to make money, consider that you might be getting some bad advice from the losers. (And, yes, VCs who back a lot of companies that flop also fall into that camp.)