Why do they call this a bubble? Almost all startups fail, and this has been true since the beginning of time. If there's suddenly a glut of seed investors willing to give startups a shot at survival, we're only increasing the chances of unearthing another dropbox.
As was said at the time, the portfolio effect requires investing in many startups to diversify risk. Most startups are high-risk investments that will obviously fail. And only extremely qualified investors should invest in startups.
One major issues with all these accelerators is the mentors are increasingly less qualified and experienced to give advice. The other major issue is the startups themselves are not focusing on actually making anything people want, but rather focusing on raising money. Paul Graham said as much at Demo Day for YCS12, where the startups were prohibited from speaking about fundraising efforts: http://ycuniverse.com/ycw13-likely-less-than-50-companies
This isn't about the start-ups in the incubators being a bubble, but rather the incubators themselves. There are too many incubators supporting too many companies, and if incubators don't have exists, which many won't, they aren't making money and therefore will fold. It's a flawed business model for all but the top few.
Yes, people frequently use the term 'bubble' these days when they actually mean 'fad'. Bursting bubbles attract more attention than passing fads, apparently. But accelerators are clearly a fad, not a bubble.
Then you also have companies like Scion using the incubator concept as a marketing campaign where they do provide a small amount of seed funding, and mentoring to young entreprenuers, but dont take any equity in return. So instead of selecting the business most likely to gain a return on their investment, they select companies that best represent their brand and the image they are trying to create.
I think incubators (except for truly valuable ones) are doomed because have these options:
a) raise seed money from an angel
b) develop a mvp, show some traction and raise money from VCs
c) bootstrap a profitable product
or d) go to an accelerator
from my expirience almost all accelerators have will present you with a bad deal (say 10-20k and office space) and claim to provide you with more value by helping you out with their expertise in other areas. In exchange they want a 20-30% equity cut. BUT the value add through their network (for most accelerators) will be negligible that's why it is a bad deal.
Every other option is better:
a) Angel -> can be a vocal defender and loyal partner or early adopter of your product
b) VC -> You get money at a fair valuation
c) Bootstrap -> Keep all the equity and stay hungry
So for what it's worth I think reducing the number of accelerators would not be a bad thing for the startup eco system.
Really? The first accelerator, the most prestigious, exclusive and selective asks for 6-10% and there are others asking for 30%? No wonder Thiel said getting into another accelerator was a negative signal.
YC originally gave something like 10K for roughly 6%, giving a rough valuation in the 180K range. HackFWD is giving, for two people, 110K for 30% which implies a value of roughly 370K. On the numbers alone the deal isn't as bad as it sounds.
The only first-hand data I have is seeing the valuations YC companies raise on their notes, vs. what I consider to be fairly comparable teams (if not companies) from 500 Startups and no incubator. The YC companies are maybe 2x 500 Startups, and 3-4x people from outside the Bay Area. Given that a lot of YC teams weren't bay area insiders when they joined YC, it's a no-brainer. (and I'd consider 500 Startups worthwhile in general as well; I just think YC is better)
I think a top FB/Google tech lead who knows investors who leaves to go do a related startup probably doesn't see as much of a bump in valuation from YC as someone who just dropped out of U of Michigan and is living in Ann Arbor, but there are plenty of other reasons why giving up 2-10% to YC is more than worthwhile.
"I think a top FB/Google tech lead who knows investors who leaves to go do a related startup "
I suspect there's a strong selection bias in YC and other incubators (take dropbox, for which Drew is an MIT grad), which probably makes my request for numbers somewhat foolhardy.
I believe the YC application process is an effective filter, but I don't think it particularly needs to be -- I'd be willing to bet that the marginal admit and marginal reject are close enough in quality that YC and its benefits are a major difference in outcome. (It might be a harder problem than elite college admissions, since while you can pretty readily identify top credentials and signs of failure, successful startups are more frequently hidden behind cosmetic flaws (or personal biases) than successful students are hidden in with C/D students with a bad GPA.)
YC has gotten pretty big now relative to a certain part of the startup ecosystem, but look back in 2005-2008 for when YC was smaller, and the success rate of YC companies vs. non-YC companies of that time. It seems pretty good, which warrants the valuations.
EDIT:
I'm not saying HackFWD is dumb money, with regards to money smartness they are definitely better then most other European accelerators. My original (GP) statement is not about HackFWD but accelerators in general.
You're still assuming that HackFWD is dumb money. From what I can tell, they have expertise in European markets that probably surpasses YC and they give a much longer runway to develop without funding pressures (full year as opposed to a few months). There's no constant pressure to prepare for demo day.
I'm not saying there isn't a value-add to YC, but rather that its bad to assume other programs don't have value.
There's an argument to be made that longer runways don't help.
YC arguably made a mistake for many companies by going from 20k to 20+150k, because with 170k you can waste a lot of time and also there's a lot left to fight over if it fails early. This is mostly why YC went to 20+80k now.
I think the right thing is probably 50-100k plus a low-hassle way to get another 250-300k, and then 1-2mm. So, something like YC plus non-stupid early investors and something like a raise shortly after Demo Day. (If I had a 10-20mm fund, I'd love to fill in the 250-300k gap)
There's a difference between monetary runway and time runway. YC solved the money problem but didn't solve the time problem with throwing more money.
To put it different, you can throw billions of dollars at a company, but they'll still operate in stress mode if you tell them a demo is due in a few months.
What they should have done is extended the program length while giving more money.
I think there are probably two distinct types of company. One where you do a 1 month dev, 1 month launch, and can have traction; another where having a year helps (hardware, or some enterprisey stuff, or some platforms).
There is a strong bias toward "having launched at Demo Day". If I had two things to work on, and got into YC, I'd absolutely pick the one I could get launched before Demo Day.
There are some companies which think they're type 2 but really have elements of type 1, and should do those elements first. (which is kind of what my company did; we wanted to do generalized secure/trustworthy cloud computing, and should have focused on specific applications of that technology, which is what we're doing now)
It's probably a net win for society and for an accelerator if they push some of the "fake type 2" into type 1 even at the cost of excluding some genuine type 2.
An easy hack for YC would be to have companies accepted into YC but let them choose to defer until the following Demo Day. You do need the initial 20-50k ASAP to get corp formation and other things done on a good foundation, but most of the "genuine type 2" companies are founded by people who can float their own living expenses for a year or two (or do very low time commitment consulting to do so), so it doesn't need to be more money.
It doesn't surprise me at all. Just like lower-tier universities have harsher rules for students, or cheap hotels have to ask for a larger % as a deposit, the more exclusive you are, the more you can trust the people you choose.
If you have the pick of the best people, you don't need to take a big cut to show a return.
I saw one accelerator in the UK that takes a 75% cut. (Not joking).
Unrelated note... On first load of the page in Chrome, could not scroll. Reloaded. Several times. Scrolling finally works. Decided tl;dr as the story is of little interest to me. Back button redirects back through a huge array of page elements that you have to go through before returning to the actual previous page. How did this garbage make it out of testing? I'll be avoiding bloomberg news links in the future.
The research behind this article is pretty flakey. Measuring accelerator success by exits only works when you do cohort analysis, not when you compare companies created 6-7 years ago (YC and TechStars) against companies created in the last couple of years (the others accelerators).
YC's scaling back of class sizes wasn't anything to do with the quality of the cohort dropping, but rather due to YC's ability to handle a cohort of that size.
First-time startup founders tend to make the same mistakes as each other, and that's one area in which it's easy for accelerators to add value (some do and some don't).
Clearly some accelerators are better than others, and some will undoubtedly fail, but this article doesn't really provide any compelling argument for it.
I'd never heard of "Unreasonable at Sea" -- this seems like a really interesting concept.
Another thing I've been thinking about is renting some houses in Hawaii, Thailand, etc., where teams could work for ~3-4 weeks on a new project (1983-Apple-Mac-team style), and then bring their friends/families for a week or so at the end.
"There are already signs of paring back. In December, Y Combinator’s Graham wrote about his accelerator’s decision to reduce its class size, from 84 companies in the summer 2012 class to less than 50 in the current session."
This was nothing to do with any overall trend. It was just because YC's then structure couldn't deal with so many startups.
The other thing this guy doesn't seem to grasp is that we didn't decrease the size of something homogenous. This is not like a restaurant cutting back from 84 tables to 46. Essentially we picked the best 46 out of 84. And in any group of 84 startups, the top 46 will have 100% of the success.
There are so many of these things now that I wouldn't be surprised if a lot of them die off. But if that happened I don't think it would affect us.
I guarantee you that pg isn't claiming he knows which the "Top 46" of 84 are. All he can do is pick the best as he see's them, and hope for the best.
I think he's said in many article, that, in fact, the vast majority of the returns each session will come from only one or two startups, and he certainly isn't able to identify those two in advance.
Between batches we spent a lot of time systematically trying to identify the predictors of failure. Previously we only tried to look for predictors of success. And they are not simply complements.
Which argues against this change, right? If you think the next Dropbox could have some red flags along with its positive signs, you'd want to mostly ignore red flags.
Somewhat. Depends how good the predictors of failures are. If they're strong enough, it doesn't matter how many promising signs there are.
But you're right that this is a perilous move. We only did it because we had to, and I'd like to edge back toward our old approach now that we can scale better. We'll probably always have a tighter filter than we used to, though, if only because some of the techniques we learned for predicting failure seem very useful.
I don't have any counter-evidence to show if the incubator trend is a bubble or not so I am not going to argue along those lines.
I have a very different perspective to offer. I live in Pakistan where there is an energy crisis so serious we don't have electric power half the day. It is in these challenging conditions that we are trying to establish ourselves as a tech startup. Perhaps this is not such an issue in SV but incubators can play a very crucial role in places where the conditions are not so ideal .
For example the incubator we have here, apart from providing seed money and mentoring, provides 24/7 power supply, connects us with businesses abroad (huge benefit, because no one wants to do business with a Pakistani startup, right?), is trying to get some foreign investors on board, etc. In effect, an ecosystem is being built and I dont see how it could have been possible without an incubator at its center.
Perhaps in SV these things dont even matter but they can be the difference between survival and failure in lesser developed ecosystems.
41 comments
[ 2.6 ms ] story [ 106 ms ] threadhttp://www.paulgraham.com/swan.html
As was said at the time, the portfolio effect requires investing in many startups to diversify risk. Most startups are high-risk investments that will obviously fail. And only extremely qualified investors should invest in startups.
One major issues with all these accelerators is the mentors are increasingly less qualified and experienced to give advice. The other major issue is the startups themselves are not focusing on actually making anything people want, but rather focusing on raising money. Paul Graham said as much at Demo Day for YCS12, where the startups were prohibited from speaking about fundraising efforts: http://ycuniverse.com/ycw13-likely-less-than-50-companies
a) raise seed money from an angel
b) develop a mvp, show some traction and raise money from VCs
c) bootstrap a profitable product
or d) go to an accelerator
from my expirience almost all accelerators have will present you with a bad deal (say 10-20k and office space) and claim to provide you with more value by helping you out with their expertise in other areas. In exchange they want a 20-30% equity cut. BUT the value add through their network (for most accelerators) will be negligible that's why it is a bad deal. Every other option is better:
a) Angel -> can be a vocal defender and loyal partner or early adopter of your product
b) VC -> You get money at a fair valuation
c) Bootstrap -> Keep all the equity and stay hungry
So for what it's worth I think reducing the number of accelerators would not be a bad thing for the startup eco system.
YC originally gave something like 10K for roughly 6%, giving a rough valuation in the 180K range. HackFWD is giving, for two people, 110K for 30% which implies a value of roughly 370K. On the numbers alone the deal isn't as bad as it sounds.
I think a top FB/Google tech lead who knows investors who leaves to go do a related startup probably doesn't see as much of a bump in valuation from YC as someone who just dropped out of U of Michigan and is living in Ann Arbor, but there are plenty of other reasons why giving up 2-10% to YC is more than worthwhile.
I suspect there's a strong selection bias in YC and other incubators (take dropbox, for which Drew is an MIT grad), which probably makes my request for numbers somewhat foolhardy.
YC has gotten pretty big now relative to a certain part of the startup ecosystem, but look back in 2005-2008 for when YC was smaller, and the success rate of YC companies vs. non-YC companies of that time. It seems pretty good, which warrants the valuations.
- Just entering YC will generate promo for your product, it will put you in-front of the eyes of thousand of tech focused people.
- Free job postings on HN
- Incredible alumni network of actually successful founders
- Demo day will be covered by major news outlets and attended by lots of investors
- 150k of convertible credit [1]
- Only 10 percent don't raise any money after graduating [1]
- 90% raise an average of 700k USD [1]
- "The average valuation of Y Combinator-backed companies, according to co-founder Paul Graham, is $45.2 million." [2]
Looks way more valuable then 170k of dumb money to me.
[1] http://blogs.wsj.com/venturecapital/2011/02/01/y-combinators...
[2] https://en.wikipedia.org/wiki/Y_Combinator_(company)
EDIT: I'm not saying HackFWD is dumb money, with regards to money smartness they are definitely better then most other European accelerators. My original (GP) statement is not about HackFWD but accelerators in general.
I'm not saying there isn't a value-add to YC, but rather that its bad to assume other programs don't have value.
YC arguably made a mistake for many companies by going from 20k to 20+150k, because with 170k you can waste a lot of time and also there's a lot left to fight over if it fails early. This is mostly why YC went to 20+80k now.
I think the right thing is probably 50-100k plus a low-hassle way to get another 250-300k, and then 1-2mm. So, something like YC plus non-stupid early investors and something like a raise shortly after Demo Day. (If I had a 10-20mm fund, I'd love to fill in the 250-300k gap)
To put it different, you can throw billions of dollars at a company, but they'll still operate in stress mode if you tell them a demo is due in a few months.
What they should have done is extended the program length while giving more money.
There is a strong bias toward "having launched at Demo Day". If I had two things to work on, and got into YC, I'd absolutely pick the one I could get launched before Demo Day.
There are some companies which think they're type 2 but really have elements of type 1, and should do those elements first. (which is kind of what my company did; we wanted to do generalized secure/trustworthy cloud computing, and should have focused on specific applications of that technology, which is what we're doing now)
It's probably a net win for society and for an accelerator if they push some of the "fake type 2" into type 1 even at the cost of excluding some genuine type 2.
An easy hack for YC would be to have companies accepted into YC but let them choose to defer until the following Demo Day. You do need the initial 20-50k ASAP to get corp formation and other things done on a good foundation, but most of the "genuine type 2" companies are founded by people who can float their own living expenses for a year or two (or do very low time commitment consulting to do so), so it doesn't need to be more money.
YC is basically 6-7% now, and bumps your valuation by (what appears to be) >100%, so it's a no-brainer for that reason.
If you have the pick of the best people, you don't need to take a big cut to show a return.
I saw one accelerator in the UK that takes a 75% cut. (Not joking).
YC's scaling back of class sizes wasn't anything to do with the quality of the cohort dropping, but rather due to YC's ability to handle a cohort of that size.
First-time startup founders tend to make the same mistakes as each other, and that's one area in which it's easy for accelerators to add value (some do and some don't).
Clearly some accelerators are better than others, and some will undoubtedly fail, but this article doesn't really provide any compelling argument for it.
Mathematical Definition, Mapping, and Detection of (Anti)Fragility Nassim N. Taleb, Raphael Douady
http://arxiv.org/pdf/1208.1189v1.pdf
Another thing I've been thinking about is renting some houses in Hawaii, Thailand, etc., where teams could work for ~3-4 weeks on a new project (1983-Apple-Mac-team style), and then bring their friends/families for a week or so at the end.
This was nothing to do with any overall trend. It was just because YC's then structure couldn't deal with so many startups.
The other thing this guy doesn't seem to grasp is that we didn't decrease the size of something homogenous. This is not like a restaurant cutting back from 84 tables to 46. Essentially we picked the best 46 out of 84. And in any group of 84 startups, the top 46 will have 100% of the success.
There are so many of these things now that I wouldn't be surprised if a lot of them die off. But if that happened I don't think it would affect us.
I think he's said in many article, that, in fact, the vast majority of the returns each session will come from only one or two startups, and he certainly isn't able to identify those two in advance.
(Not doubting, honestly curious.)
But you're right that this is a perilous move. We only did it because we had to, and I'd like to edge back toward our old approach now that we can scale better. We'll probably always have a tighter filter than we used to, though, if only because some of the techniques we learned for predicting failure seem very useful.
I have a very different perspective to offer. I live in Pakistan where there is an energy crisis so serious we don't have electric power half the day. It is in these challenging conditions that we are trying to establish ourselves as a tech startup. Perhaps this is not such an issue in SV but incubators can play a very crucial role in places where the conditions are not so ideal .
For example the incubator we have here, apart from providing seed money and mentoring, provides 24/7 power supply, connects us with businesses abroad (huge benefit, because no one wants to do business with a Pakistani startup, right?), is trying to get some foreign investors on board, etc. In effect, an ecosystem is being built and I dont see how it could have been possible without an incubator at its center.
Perhaps in SV these things dont even matter but they can be the difference between survival and failure in lesser developed ecosystems.