The implication that this can indicate which number that gives you the highest chance of success is completely bogus. Without taking account failures (or distribution of attempted companies by founder count) you have no data whatsoever for discussing what is most likely to succeed.

Makes me ask, what is this doing on HN?

1. Collecting karma for the submitter. ("Why X is foo" followed by "why X is NOT foo" followed by "Foo is irrelevant, it's actually all about bar").

2. Promoting the guy on the other side of the link. Unfortunately Vibhu Norby is one droplet in the indistinguishable white mist that makes up the ever-expanding Svbtle Borg, so it probably isn't working all that well.

(Yes, I totally do 2. Every chance I get to link to myself, I take.)

Problems:

1. How did Deloitte define their data set? (edit: I found out, it's in the report)

2. Does fast growth == success?

3. How is a founder defined?

4. What about companies that failed?

Whenever controversial something hits the front page, there's always a spate of quick and easy responses that turn up in the next few days.

That being said, I think my list of problems still remains.

You say that you have a breakdown of how many founders different successful companies have.

Well, you are already ahead of Ryan Carson and Paul Graham in that you've risen above an exchange of anecdotes at 20 paces.

But I think you oversold it in the title.

"How many founders do successful tech companies have?" is not supported by the data.

First, it's based on a list of "fastest growing". Lots of companies grow fast and then abruptly stop growing. You just substituted one metric in for "success" and didn't say that's what you did.

Second, the list is based on self-submitted reports (scroll down to "Selection and Qualifying Criteria") -- "Deloitte has not audited the ranking ... Some companies that may be eligible to appear in the ranking are not included because they did not submit the required information or otherwise declined to participate".

Third, it doesn't deal with failures, which is more interesting because that is the default condition.

Fourth, it's restricted to North American companies only.

Fifth, how did you define "founder"? That seems to me to still be a slippery wildcard in this whole discussion.

Take Ryan Carson's case. He owns most of the equity. Is he the sole founder? If "founder" is based on legal control, yes. What about the day-to-day? He says he has an employee who is a de-facto co-leader. Are they a founder? Maybe not. What if they'd been there right from the start? And so on.

Good on you for doing the tedious slog of looking up a hundred companies, but ... well, I think you've oversold on the title.

And that's what my gut, knee-jerk, ex-Slashdot cynicism reacted to.

A founder is someone listed as a co-founder either on the management page, Wikipedia, or in press mentions. It's not really slippery. It took so much time because I wanted to make sure I was doing it right.

> Like I said in the post, there is no reason to believe that biases how many founders they have.

Different cultures have different ideas about individual risk and teamwork.

> tons of revenue ... 10-15 years old.

Gosh, you mean young companies have more scope for rapid growth and therefore wind up on a self-selected list of rapidly growing companies?

In no way could this sample be biased against failed young companies.

Also unanswered: are there fewer single founders because there are fewer single founders to begin with, or is it because they fail more often?

I don't see how that conclusion can be drawn from this kind of data.

EDIT: It looks like the post has been edited, so this criticism is no longer relevant.

I don't even need to look at the data to say that in the US most successful companies have at least one white, male founder. Does that mean that if you start a tech company and are non-white or non-male you should make sure to partner up with one to ensure success? Is it just a bias due to demographics? It it a real factor related to an intrinsic advantage of white males in making successful startups (even if it's just an external cultural factor)?

Or, is it a much more complex issue that can't and shouldn't be so easily infographed and turned into a convenient piece of folk wisdom?

Taking this data, which incidentally is extremely incomplete, and using it as a rubric on how to start your company is little more than cargo cultism.

I feel This statement assigns specific attributes to the data-points in question where as the original data is more generalized (and suited to the argument). Doesn't the assignment of random attributes skew the point being made and just try to forcibly fit a separate tangential argument?

Let's look at this from a different perspective. Assume that the data conclusively shows that 2-founder startups have a higher chance of success than 1-founder startups, by a huge margin (say 10x). Does that mean that you should seek a co-founder before launching your startup? Even then, it does not. Consider two different startup ideas, one is founded on a very solid plan with a good chance of success, the other is founded on a bad plan with a lot more risk. The good plan is going to have a much easier time of finding a co-founder, because the potential co-founder will look at the idea and see that it has a lot of merit and potential and they'll be enthusiastic about joining in, whereas the bad idea will have a much harder time finding co-founders. However, imagine if someone with a bad startup idea pushes much, much harder to get a co-founder because they have learned that startups with co-founders are more likely to be successful. Will that help the company's prospects any or will it just warp the equation and result in a bad company failing even with a co-founder?

I think it's more important to go about a startup with your head screwed on straight than to attempt to imitate the forms of previous success. If you find someone you trust who seems to be a valuable addition to your startup as a co-founder, then by all means make that partnership. But if you can't find a co-founder for whatever reason but you are still quite convinced of your business plan then I don't think you should wait. You should certainly be careful that you aren't deluding yourself and be sensitive to any criticism of your startup idea nonetheless, but a good plan is a good plan even if you go it alone.

Also, the amount of founders could be neutral to the success, in such case neither is countering their own success, and again just a matter of samples.

Correlation is not...

What is really necessary is a statistically significant sampling of startups from their inception and tracking the percentage of success and failure depending on various factors. There is simply not enough data in the study at hand to say that 1-person startups are less likely to succeed than 2-person startups, in fact it's quite possible that they are more likely to succeed, we just don't know.

Your next two sentences explain why this isn't true.

I did some digging on the US census site and found these numbers (for 2008):

- 22,614,000 returns were filed for non farm sole proprietorships. Of these, 16,434,000 indicated any type of income. Combined, these 16.4 million sole proprietorships reported a net income (less loss) of $265 billion.

- 3,146,000 returns were filed for partnerships. These returns showed a total of 19,300,000 partners. Of these, 1,609,000 returns indicated any type of income. Combined, the 1,609,000 partnerships reported a net income (less loss) of $458 billion.

Some numbers:

- 72.67% of sole proprietorships reported income.

- 51.14% of partnerships reported income.

- 22.614 million sole proprietorships generated $256b in net income (less loss). This is an average of $11,320.42 per proprietorship started.

- 3.146 million partnerships (covering 19.3 million people) generated $458 billion in net income (less loss). This is an average of $145,581.69 per partnership started. Or, this is a net income (less loss) of $23,730.57 per partner.

Using US census information on non incorporated businesses (in 2008) provides some interesting space for thought. A higher percentage of sole proprietorships report income, but on average partners in a non incorporated business report more income. This seems to indicate that it might be easier to get started if you go alone, but adding founders increases the magnitude of success (when it works out). Logically, this makes some sense, though I hesitate to make any conclusions based off of this data.

While this data is interesting, it is ultimately quite useless at answering the question. First off, this data is on all non incorporated businesses (other than farms). Therefore, this data may or may not be applicable to technology. The Deloitte report (and the writer's research) indicates that the top 100 technology companies have an average of 2.12 founders. The US census data gives an average of 1.62 founders per company. That difference could indicate differences in the two populations (ie - tech companies versus all companies) or sampling differences (Vibhu looks at 100 companies, I'm looking at data on 41 million).

The problem with a single individual managing responsibilities is that there are still plenty of tasks that need to be done so it becomes a question of what you can automate and what you can outsource for as cheap as possible. Basically, good delegation and management skills.

This mischaracterizes the dialog.

The issue is about a specific type of company, startups, and a specific definition of "startup," that used by venture capital.

PG's comments were as narrowly focused as his essays and YC itself. His data set was not the Tech Sector, but a particular type of business.

Not really ... it only indications that more single founders fail to grow fast enough to make Deloitte's list.

I'm happy that there are a bunch of single founder companies. I'm planning to start one, as a single founder.