6 ms·

They're also not independent trials. If you just failed 9 times in a row, you're probably not the best at startups.

GuiA 9y ago
Or maybe you learned tons and it makes your 10th attempt more likely to succeed? Ultimately these ponderations are kind of meaningless in the absence of great data to frame them with.
- jabgrabdthrow 9y ago
  At which number would you no longer make the point you just made? 100 failed startups? 50,000? You’ve got a lot of faith in people. My number is maybe 5.
  - always_good 9y ago
    Dunno, depends on the kinds of startups you're trying to create. For example, I don't think you're "bad at startups" just because you create more ambitious ones with a higher chance of failure. Some people are at a place where they can take more risks, it doesn't make them an abject failure.
  - gargarplex 9y ago
    I've been seriously, actively interested in building a tech biz or startup since 2003. In 2003, I started selling chat bots on the Internet. In 2005, I cofounded an affiliate marketing website that cracked the Alexa top 10k In 2007, I built an app that scaled to 100k MAUs and was acquired for $20K In 2009, I started a SaaS but couldn't break $~600/month in revenue. In 2010, I built a business that ran for over 5 years and supported me entirely for two of them, but probably grossed less than $100K total revenue. But it couldn't grow, as I struggled to find a repeatable and predictable customer acquisition channel. In 2013, I got involved in Bitcoin. Finally one home run after a portfolio of mediocrity. [I am choosing to omit various projects from this list, but between 2003-present, there have been a lot.] In 2013, I also started doing stock market research with ruby. In 2015, my work had attracted enough momentum to bring on an outside angel investor and even piloted the strategy in the markets with $1.5MM. Didn't perform that well; investors withdrew their capital. In 2017, I started a SaaS that is currently doing ~$2800/month in revenue. I also did consulting. I grossed about $190K this year from my business efforts - note, that's nearly 10 times faster (in terms of growing revenue) than I was 7 years ago. I also fixed a major hole in my previous business skillset, i.e., predictably acquiring customers - something I could never do for the biz in 2010. Furthermore, I have a lot more emotional maturity than I did 7 years ago - though Sam Altman was precociously mature, with the "emotional maturity of a 40 year old" (Paul Graham, on meeting Stanford undergraduate Sam Altman). So I can tell you definitively, that: - My skills are expanding due to systematically resolving the deficiencies in previous attempts. - These improvements are a result of deliberate practice. - Some people may have the ancillary skills, resources or characteristics that position them to perform better from early on - YC seems to be adept at identifying these people, but also credit them for making 1,500+ bets. - My best success to-date has been as a result of luck. I'm 30 years old now.
    - JamesBarney 9y ago
      Would you mind going into a little more detail about the hole that you plugged and how you plugged it?
      - gargarplex 9y ago
        For me it was learning all about sales. I actually got a job selling industrial equipment specifically for this purpose. I learned that a good sales person is an agent who advances the best interests of the buyer as well as the seller. I also had the epiphany that my ads and website are - essentially - salespeople, selling to exactly one visitor at a time. So if I can't sell without the website, I can't sell with the website. But once I can sell, and then once my website can sell in my stead, it's about getting the website in front of as many people as possible - just like a productive sales rep will make many phone calls per day. Just as a salesperson will tailor her approach based on the market or personality of the buyer, a good website should do so as well. Therefore, I have segmented mailing lists, different landing pages for different buyer types, and so forth. You may also find this comment I wrote about a year ago useful as well, https://news.ycombinator.com/item?id=13283234 https://news.ycombinator.com/item?id=13283234
        user5994461 9y ago
        +1 Sales and marketing is as important as the software. They can't exist without each other.
    - ChristianGeek 9y ago
      Bitcoin is not a win until you sell it at a profit (which you may have done...I’m just saying.)
    - sihui_io 9y ago
      Thanks a lot for sharing. I'm impressed by your perseverance and so glad to see things have been working out for you lately. Best of Luck!
- gwern 9y ago
  If I recall correctly, one of the Compass VC analyses covered serial entrepreneurs and found that there was no increase in success rates and each attempt looked independent. There's just that much randomness in it, and I suppose the various selection effects cancel out.
  - raoulj 9y ago
    I googled around for this document to no success. Do anyone have a link to this?
    - gwern 9y ago
      Looking for it, I think I am remembering either http://10years.firstround.com/ http://10years.firstround.com/ (which says, specifically, that experienced founder returns aren't better because their increased success rate is already priced in, which is not quite what I claimed) or https://www.econstor.eu/bitstream/10419/90815/1/777058081.pdf https://www.econstor.eu/bitstream/10419/90815/1/777058081.pd...
- antisthenes 9y ago
  If you had the social and monetary capital to attempt 10 startups and fail, one wonders if that money and effort would not have been better invested elsewhere (if we assume that the goal of the startup is to achieve a successful exit), perhaps in easier forms to generate passive income or what have you.
- matte_black 9y ago
  Or you might think you’ve learned tons but you’ve actually learned all the wrong lessons and are spiraling hopelessly confident in the wrong direction. Failure doesn’t owe you anything.
sarreph 9y ago
My thoughts exactly — thanks for writing this. Helps to explain how there are so many 'anomalous' serial-entrepreneurs :)
fsloth 9y ago
What I recall of chances of success, earlier failures do not correlate with success - either way. So basically one is throwing fair dice every time. I might be recall wrong...
- TheCoelacanth 9y ago
  That is with dice. Startups are not independent random events because some people are better at making startups than other people are.
  - computerphage 9y ago
    I think your parent is saying that when you actually measure whether your claim is true, it isn't actually correlated in the failure case. I remember reading something that corroborated this [1]. That source claimed that previous founders of successful companies had a much greater chance at additional success, but previous failures didn't increase your chances (ie. you don't, in fact, learn more from failure than success -- at least when it comes to founding failed startups). Edit adding source: [1] https://papers.ssrn.com/sol3/papers.cfm?abstract_id=933932 https://papers.ssrn.com/sol3/papers.cfm?abstract_id=933932
    - JamesBarney 9y ago
      Ycombinator has made a lot of money based on the idea that some people are better at creating start ups than others. They don't just bet on ideas they also bet on people.
      - tikhonj 9y ago
        On the flipside, it could be that Y Combinator has made a lot of money by creating an environment that helps startups succeed. Apart from money, they provide founders advice, structure and a set of motivated peers. More cynically, they also provide prestige, validation and connections. I have no doubt Y Combinator would still be successful even if their selection method was not very good. (I tend to think about top universities the same way—the quality of Harvard's admissions process is not the core reason they graduate so many successful students.)

GuiA 9y ago

Or maybe you learned tons and it makes your 10th attempt more likely to succeed? Ultimately these ponderations are kind of meaningless in the absence of great data to frame them with.

jabgrabdthrow 9y ago
At which number would you no longer make the point you just made? 100 failed startups? 50,000? You’ve got a lot of faith in people. My number is maybe 5.
- always_good 9y ago
  Dunno, depends on the kinds of startups you're trying to create. For example, I don't think you're "bad at startups" just because you create more ambitious ones with a higher chance of failure. Some people are at a place where they can take more risks, it doesn't make them an abject failure.
- gargarplex 9y ago
  I've been seriously, actively interested in building a tech biz or startup since 2003. In 2003, I started selling chat bots on the Internet. In 2005, I cofounded an affiliate marketing website that cracked the Alexa top 10k In 2007, I built an app that scaled to 100k MAUs and was acquired for $20K In 2009, I started a SaaS but couldn't break $~600/month in revenue. In 2010, I built a business that ran for over 5 years and supported me entirely for two of them, but probably grossed less than $100K total revenue. But it couldn't grow, as I struggled to find a repeatable and predictable customer acquisition channel. In 2013, I got involved in Bitcoin. Finally one home run after a portfolio of mediocrity. [I am choosing to omit various projects from this list, but between 2003-present, there have been a lot.] In 2013, I also started doing stock market research with ruby. In 2015, my work had attracted enough momentum to bring on an outside angel investor and even piloted the strategy in the markets with $1.5MM. Didn't perform that well; investors withdrew their capital. In 2017, I started a SaaS that is currently doing ~$2800/month in revenue. I also did consulting. I grossed about $190K this year from my business efforts - note, that's nearly 10 times faster (in terms of growing revenue) than I was 7 years ago. I also fixed a major hole in my previous business skillset, i.e., predictably acquiring customers - something I could never do for the biz in 2010. Furthermore, I have a lot more emotional maturity than I did 7 years ago - though Sam Altman was precociously mature, with the "emotional maturity of a 40 year old" (Paul Graham, on meeting Stanford undergraduate Sam Altman). So I can tell you definitively, that: - My skills are expanding due to systematically resolving the deficiencies in previous attempts. - These improvements are a result of deliberate practice. - Some people may have the ancillary skills, resources or characteristics that position them to perform better from early on - YC seems to be adept at identifying these people, but also credit them for making 1,500+ bets. - My best success to-date has been as a result of luck. I'm 30 years old now.
  - JamesBarney 9y ago
    Would you mind going into a little more detail about the hole that you plugged and how you plugged it?
    - gargarplex 9y ago
      For me it was learning all about sales. I actually got a job selling industrial equipment specifically for this purpose. I learned that a good sales person is an agent who advances the best interests of the buyer as well as the seller. I also had the epiphany that my ads and website are - essentially - salespeople, selling to exactly one visitor at a time. So if I can't sell without the website, I can't sell with the website. But once I can sell, and then once my website can sell in my stead, it's about getting the website in front of as many people as possible - just like a productive sales rep will make many phone calls per day. Just as a salesperson will tailor her approach based on the market or personality of the buyer, a good website should do so as well. Therefore, I have segmented mailing lists, different landing pages for different buyer types, and so forth. You may also find this comment I wrote about a year ago useful as well, https://news.ycombinator.com/item?id=13283234 https://news.ycombinator.com/item?id=13283234
      - user5994461 9y ago
        +1 Sales and marketing is as important as the software. They can't exist without each other.
  - ChristianGeek 9y ago
    Bitcoin is not a win until you sell it at a profit (which you may have done...I’m just saying.)
  - sihui_io 9y ago
    Thanks a lot for sharing. I'm impressed by your perseverance and so glad to see things have been working out for you lately. Best of Luck!
gwern 9y ago
If I recall correctly, one of the Compass VC analyses covered serial entrepreneurs and found that there was no increase in success rates and each attempt looked independent. There's just that much randomness in it, and I suppose the various selection effects cancel out.
- raoulj 9y ago
  I googled around for this document to no success. Do anyone have a link to this?
  - gwern 9y ago
    Looking for it, I think I am remembering either http://10years.firstround.com/ http://10years.firstround.com/ (which says, specifically, that experienced founder returns aren't better because their increased success rate is already priced in, which is not quite what I claimed) or https://www.econstor.eu/bitstream/10419/90815/1/777058081.pdf https://www.econstor.eu/bitstream/10419/90815/1/777058081.pd...
antisthenes 9y ago
If you had the social and monetary capital to attempt 10 startups and fail, one wonders if that money and effort would not have been better invested elsewhere (if we assume that the goal of the startup is to achieve a successful exit), perhaps in easier forms to generate passive income or what have you.
matte_black 9y ago
Or you might think you’ve learned tons but you’ve actually learned all the wrong lessons and are spiraling hopelessly confident in the wrong direction. Failure doesn’t owe you anything.

sarreph 9y ago

My thoughts exactly — thanks for writing this. Helps to explain how there are so many 'anomalous' serial-entrepreneurs :)

fsloth 9y ago

What I recall of chances of success, earlier failures do not correlate with success - either way. So basically one is throwing fair dice every time. I might be recall wrong...

TheCoelacanth 9y ago
That is with dice. Startups are not independent random events because some people are better at making startups than other people are.
- computerphage 9y ago
  I think your parent is saying that when you actually measure whether your claim is true, it isn't actually correlated in the failure case. I remember reading something that corroborated this [1]. That source claimed that previous founders of successful companies had a much greater chance at additional success, but previous failures didn't increase your chances (ie. you don't, in fact, learn more from failure than success -- at least when it comes to founding failed startups). Edit adding source: [1] https://papers.ssrn.com/sol3/papers.cfm?abstract_id=933932 https://papers.ssrn.com/sol3/papers.cfm?abstract_id=933932
  - JamesBarney 9y ago
    Ycombinator has made a lot of money based on the idea that some people are better at creating start ups than others. They don't just bet on ideas they also bet on people.
    - tikhonj 9y ago
      On the flipside, it could be that Y Combinator has made a lot of money by creating an environment that helps startups succeed. Apart from money, they provide founders advice, structure and a set of motivated peers. More cynically, they also provide prestige, validation and connections. I have no doubt Y Combinator would still be successful even if their selection method was not very good. (I tend to think about top universities the same way—the quality of Harvard's admissions process is not the core reason they graduate so many successful students.)