147 comments

[ 0.26 ms ] story [ 251 ms ] thread
I would be very interested in seeing a cohort analysis of YC company valuations by batch :-)
Nice to see more detailed stats on this stuff. One figure I'd really like to see is the number of companies worth more than $10M & $50M.
how good or bad are those numbers? only 8/970 companies are worth > 1b and 40 are worth over 100M.

I assume YC is doing great for themselves, so I'm wondering, what is the rule of thumb to say an investment was successful? (for YC, knowing they invest cheap and early.)

Is it once a company reaches 10M? 1M? Where there is an exit?

It would be interesting to know how many companies YC says the investment was a success (either made money or will if no exit yet).

These numbers are great from aggregation/press perspective, but they are not enough to tell if YC is doing good or bad. For instance, last 3 YC batches had ~250 (or maybe 300) companies funded. That is almost 1/3 of the total companies funded by YC. It will take another couple of years to determine what the success rate for these companies is.
how good or bad are those numbers? only 8/970 companies are worth > 1b and 40 are worth over 100M.

The billion dollar number is freaking incredible. There are 115 venture-backed private companies globally that have unicorn valuations[1]. For 8 of them to have been through the same accelerator is amazing.

[1] http://graphics.wsj.com/billion-dollar-club/

Valuations are currently so high and IPOs are so difficult that you shouldn't pay much attention to the valuation stats.

If all of those companies wanted to exit at the price of their most recent raise, they wouldn't be able to. That means that at least some of the valuations are unrealistic.

>8/970 companies are worth > 1b and 40 are worth over 100M.

Sorry to be negative, but worth according to whom?

If none of these companies are public, the valuation comes from a very narrow group of private investors. In what seems to be a bubble (and an echo chamber) in S.V., these valuations may be vastly overstated.

That said, kudos to YC for being transparent.

To all the down-voters:

Do you not think that solid data such as revenue, revenue growth and profit/loss (or at least burn rate) would be a much better measure of success? Some non-public companies (like Uber) publish some of these numbers.

Excellent point, most of the companies couldn't exit at whatever valuation they raised money last.
(comment deleted)

  Number of companies we offered to fund yesterday for the first YC Fellowship: 32
That's a figure I was looking for (took the interview yesterday, didn't get in).
Can you write a blog post on your experience? I'd be interested to see how closely it resembles YC.
> Number of companies in the last batch: 107

Wow.

Just wow.

Interesting way to bury the lede that 32 companies received a Fellowship offer. The estimate was 20. (http://techcrunch.com/2015/07/20/y-combinator-just-introduce...)
Honestly if we were prepared to take 320 that would have been good. The quality was outstanding.
I really underestimated the amount of applications that you guys received. I was thinking maybe a few hundred, like 800 or 900, at the most, but 6500? I was way off.

I was wondering though, if the Fellowship program showed great results and you guys decide to continue it, how much equity would you ask for? Have you thought that far ahead yet?

I have no inside info, but I think YC Fellowship could be a startup lead generation system: when you get a Fellowship grant and your company begins to work out, you will probably apply to YC and get accepted. Then YC receives equity (and only the equity of the huge winners matter financially).
You will see many coming back stronger for YC W16 application: 2015 is big time in a lot of fields. The YC ecosystem (alumni and network) is so diverse and lively nowadays that it could even generate self-sustaining innovation, more than universities and corporations.
What does it mean by saying "got their start with us"? I believe many companies have been around for awhile before YC. Many already with revenue, traction and funds raised (although this might be the exception to the rule?).
This is a relatively new phenomenon which is also why the YCF program exists now. Time was, you had a team and a demo.
The great majority of companies we fund, even in recent batches, have effectively zero revenue when we fund them. But certainly more have a product and users than they used to, which is why we're doing the fellowship.
YCF is an awesome move guys. Stays true to the funding of a good team/idea philosophy YC began with.
I would also just add that you should still feel welcome to apply to YC's normal program if you're still in the idea or prototype stage.

Like kevin said elsewhere in the thread, it's about finding the amazing founders in the margins -- we want to see everything, and YC can still be a good fit for everyone.

Since applications are opening tomorrow, it would be interesting to have a breakdown of interviews/acceptances based on various criteria: sole founder, has revenue, has user, etc...

It would allow prospective applicants to think about whether to apply, and hopefully keep the pile a manageable size for the people reading it.

Might be interesting, but the odds are the wrong way to think about this. Our wins come from the margins and we're looking for people to beat the odds. Most of the time we are surprised what ends up becoming big.

The numbers reveal that it's hard, but startups were ALWAYS hard. It's still the hardest thing I've ever done and most founders say the same. As far as YC is concerned, we don't want people to try to make this easier for us. One of the unique properties of YC is that we're trying to fund startups at scale and that means we don't want founders thinking about trying to make it manageable for us by bowing out. That's how I miss out on someone on the margins. It's one of the reasons why I love working on the YC software team to try and solve these challenges. So please, don't take away my work. :)

My favorite founders are filled with grit and perseverance. They don't calculate whether they can change the world before starting, they just start. So if you're ready to do something hard and feel like it's within you to change the world, please do apply...the numbers be damned.

Cool. You may have bought yourself 3 more applications then. I know a couple of other groups (aside from mine) who would benefit from this programme.
I have admired YC from afar, and wish to be part of it. After much reading on PG essays, Sam A interviews and YC demographic (young male founders), the odds are against us. Last week's NY Times article on Amazon culture, reminded me of YC. There are few similarities in concepts (work life balance, female discriminations). YCF was targeted for young founders (Sam's interview mentioned low burn rate). I understand that startup life is hard, but thinking that older people will find it harder to cope is very naive. There are lots of new innovations that can be developed with the experiences and knowledges from older founders. True that there have been some young extraordinary unicorns. But I am sure that experiences and knowledges can help to build a better long lasting businesses.
Yes the odds are wrong. But I think there are now too many good applications submitted with high growth or revenue that you guys will just skip through next Zuckerberg before his startup takes off. Your application method currently does not in anyway measure how much the users love the product, how big it can be. My thoughts are that YC at the moment is currently pretty blind to really great startups that haven't got traction yet.

Thats why fellowship experiment is going on, but I think you guys would have more time to read applications and "discover" great startups if you didn't encourage every steve-jobs-wannabe to apply.

just my 2 cents

Agree with Kevin, this might encourage the wrong way of thinking when applying for YC.

Average people, that do average things, generally think in terms of the odds "what's my chance of this, or that.." when deciding about things.

Extraordinary people don't care about their chances, they do things because they are determined or obsessed about whatever it is they are doing, and stop at nothing to make it happen. These are the types of people that go on to build great companies and are the types of people YC generally looks for in the applications.

So if you're the type of person that doesn't care about the stats: you should definitely apply!

These stats are great, or at least they sound great to an aspiring entrepreneur. I've seen stuff like this for a long time on HN, and a feeling took hold in me that I've only started to really shake recently, a feeling that the best and most determined founders will always get into YC, and that failing to get into (or apply to) YC is an indication of weakness.

But just remember that, even though these numbers look amazing, you can create a business via other means and still possibly achieve whatever sort of success you're after -- you don't have to join an accelerator, or get prestigious seed or venture funding.

Or at least, I think that you shouldn't have to.

Sam also just tweeted that 300 YC companies are no longer around. Amazing how much transparent they are and how great YC is at picking and training. So many really crazy ideas that most would think are insane and YC is able to find the ones that make sense and help them. Seems like that's a bit less than 1/3 of all YC companies

Tweet is here: https://twitter.com/sama/status/636586179970752512

YC filters applicants for dedicated founders who will not give up easily. It is amazing what can be accomplished by manic perseverance.
Last week I saw this from Sam.

> If we don’t invest in these companies, they don’t happen… The thing about Y Combinator that’s cool is that most companies won’t happen if we don’t fund them.[1]

I thought YC was always looking to fund those that would exist with or without YC. I have seen less and less "crazy bets" every batch, but I have to say there's more diversity of founders and topics (biotech, energy, farming). Companies generate revenue before DemoDay and many are startups that work for other startups. It's not a bad thing but would love to see more Airbnbs or Reddits around. Those are the ones that can generate a true impact on society and can only exist thanks to YC giving them a chance. The fellowship can be a great starting point for those.

[1] http://venturebeat.com/2015/08/23/sam-altman-and-jessica-liv...

I think Sam is saying that without YC, there are some companies that may not end up succeeding or being able to raise and continue on (without YC's network, advice, speed, focus etc. to help them)

The distinction is that they want to fund companies who would continue working on it even if they didn't get into YC. They are so passionate about the problem that they will try to make it succeed no matter what.

YC has said that Airbnb would have failed without their help. How could they not update their thinking? Being determined to succeed is necessary but insufficient. Every successful startup received a big helping hand early in its life. YC's business is to provide that opportunity to founders that otherwise wouldn't have it, such as Airbnb, which no other investor would back.
What's more intriguing is how was YC able to see through what the other investors saw as an issue?

Even to this day I'm astonished that AirBnB exists and yet YC was able to see something in them years ago.

YC also despised the idea, but invested because they liked the founders. But they also presumably liked the 600+ YC founders that failed, so you can see how much investors can really predict.
You mean at the series A? Their original idea was really different.
(comment deleted)
My understanding of the story was that YC was going to pass on Airbnb but then one of the cofounders mentioned the cereal box story and that changed PG's mind
>YC filters applicants for dedicated founders who will not give up easily

Well, they try to filter by that quality, as would any other incubator or investor. It would make no sense to accept apathetic founders who will surrender at the first obstacle. Personally, I'm skeptical that they having any advantage in identifying such abstract traits. How would you identify it from a 15 minute interview?

That's actually mostly in line from what I've heard heard from other VCs. After 5-7 years, you expect half of your portfolio to be either out of business completely or "walking dead" (cash flow negative without enough traction to get positive before the end of the runway).

Though you do have a point; I would expect a higher failure rate from an accelerator given that they fund riskier ventures.

The failure rate is probably skewed low by YCs growth rate. For example, 1/9 of the companies were in their most recent batch, and it's unlikely that any of them have gone out of business yet.

Still impressive though. Any success rate over say 10-20% seems really impressive for an incubator.

The breakout-success rate thus far is also skewed low by YC's growth curve. Things have still just begun.
Yeah. A failure rate by cohort (a "churn rate" of sorts) would be handy.
This is also rather misleading - because 300 have been shut down, 904 have been funded, which includes 107 in the last batch, which can be discounted entirely from these stats. This gives us 300/797 have shut down, or a percentage of about 38%.

What I'd like to know is the value, at exit, of the companies that have exited or gone public. Because I can guarantee you that the "valuation" of $65B+ is a totally meaningless number. This includes every single company that had a huge, unsustainable up round which will almost certainly be devalued based on future financings or exits.

> Because I can guarantee you that the "valuation" of $65B+ is a totally meaningless number. This includes every single company that had a huge, unsustainable up round which will almost certainly be devalued based on future financings or exits.

"Meaningless" is hyperbolic. How much would you pay for a share of Airbnb? More than nothing, I assume. I think one can conclude something from consummated, informed responses to this question.

No - I stand by that as a literal use.

Does 65B represent the actual money value that people will pay for the shares of 100% of the ~600 currently existing YC companies?

Does it tell us the average value? The mean? Standard deviation? Quintile distributions? P/E? Is that number just based on valuations from funding rounds? Projections?

It is literally meaningless. It is not verifiable. The standards that are used to calculate it are not explained. There is no explanation to how it relates to the companies in the portfolio.

I'm a bear by nature. I think that the current batch of SaaS unicorns have an unsustainable valuation, and I think this unverified number feeds into that.

Who else wants PC and Lawstudent2 to continue this debate? :) I'll bring popcorn.
Valuation means the price that investors or acquirers were willing to pay. That someone was willing to pay that is a verifiable fact.

What the returns will be is an unknown, of course. But to say that is "meaningless" is either naive or a misuse of the word. Stock prices and market caps are decided the same way.

Public stock prices are decided by supply-demand balance. In no way does this valuation resemble that mechanism.
Public stock prices are the last price someone was willing to buy that stock at.

Private stock prices are the last price someone was willing to buy that stock at.

It is slightly less simple, because you may give a discount for advice, but private stock prices are definitely reliant upon supply and demand. If one investor wants to set your valuation at $1B and another at $100m, you go with the one who offered a $1B valuation. There is occasionally a discount for advice/connections, but the mechanism that the highest bidder wins is generally the same. Private market stocks are by no means exempt from economics or supply and demand.

I don't know if you are making a disingenuous argument or you just don't know about liquidity, or indeed the effects of sample size on estimating a changing value, but what you've said here has no relationship with reality.
You can calculate the mean yourself by taking total valuation divided by number of companies. I'll do it for you:

$65 billion / (940 companies) = $69 million / company

Happy to buy any shares you have in any of the YC unicorns.
The problem is using the words "valuation" and "worth" interchangeably. Just because someone pays $50 million for 5% of startup doesn't make it magically worth $1 billion. No one would actually pay anything close to $1 billion for the whole thing, hence it is not worth anything close to $1 billion. That's just the implied valuation.
AirBnB: $24 Billion. DropBox: $10 Billion. Zenefits: $4.5 Billion Stripe: $3.5 Billion. Instacart: $2 Billion Twitch: $1 Billion

Total: $45 Billion.

Source: http://graphics.wsj.com/billion-dollar-club/

So they value the rest of the portfolio at $20 billion. Does that seem reasonable?

I believe that they just added up acquisition amounts and the last funding rounds of each company. I don't believe that number represents what "they value" the portfolio at beyond that.
>> Although these are very imperfect indicators of success, here they are.

There you got your response at the beginning of the post. It seems that you are attacking for the sake of attacking? Sam is providing these data because people asked for them.

I think the point you're wanting/trying to make is that private market valuations are ridiculously high. But that's not YC's fault. It should enjoy this while it lasts, and hope that as many of its investments as possible get to liquidity before the music stops.

A more interesting observation is that two companies account for over half of the $65 billion. Airbnb's most recent valuation was reportedly $25 billion, and Dropbox's was reportedly $10 billion. These companies are not likely to go away overnight a la Homejoy, but their valuations are going to be hard to sustain. Dropbox in particular would be a very tough sell to the public markets at its private market valuation given its comp[1].

Making analysis even more difficult is that today's big money, late-stage deals have lots of strings attached, so these valuations are hard to assess without knowing all of the details.

Final comment: despite the constant suggestions to the contrary, YC's portfolio appears to live in the very same power law reality as most early-stage investment portfolios in Silicon Valley.

[1] https://www.cbinsights.com/blog/dropbox-valuation-bubble/

another stat from sama on twitter:

> Oh, another important stat: about 300 of the companies we have funded have shut down.

Interesting twitter update by sama

> Oh, another important stat: about 300 of the companies we have funded have shut down.

This is a hard gig. Even after you get into the top accelerator in silicon valley.

Top accelerator in the world. Even so I find sama's comment further up about not encouraging people to start an accelerator quite funny.
"Market cap" is a poorly-chosen descriptor because the vast majority (all?) of the companies in YC's portfolio are private.

The use of the term "market cap" is especially ironic given Sam's recent post on "financial misstatements"[1]. If founders are expected to use financial terms accurately and appropriately, shouldn't investors be expected to do the same?

[1] http://blog.samaltman.com/financial-misstatements

It looks like it says "valuation" now. However, he's not trying to pitch you on investing in anything, so he can use whatever language he wants. The point of the blog post that I took was that it is illegal to mislead potential investors about your company.
So in your mind the only reason one should use accurate language would be to avoid the risk of civil or criminal litigation? Fascinating.
why is market cap not accurate language? Even private companies have market caps. These market caps are set in the private markets. Market cap = # of shares outstanding * price of each share. Applies to both private and public companies
Please talk to somebody who works in finance. You will virtually never hear the term "market cap" used to describe the value of private companies. If you use this term in this context, it will be assumed that you don't know what you're talking about.

There's a reason Sam updated the language in his post...

I work in finance :) true, the term is not often used for private companies. But that doesn't make it any less real.
(comment deleted)
Considering there is probably a vintage effect, and currently 5% of YC companies are worth over $100 million, that means that most likely 5-10% of founders that join YC will become millionaires over time. That's pretty impressive.
Keep in mind that there really haven't been many large exits, which lock in the valuations. I think YC is doing extraordinarily valuable things, but let's not get too blinded by big numbers. On-paper millionaires are one thing; actually being able to lock that money down is another thing entirely.

Edit: Forgot a word.

That's not a super high bar. Most of these founders could get $200k jobs at top tech companies, invest $100k/year, and be a millionaire in ~10 years. Much better than 10% odds!
Founders also get paid well at startups once they've raised funds. There's a large difference between having a couple million dollars amassed as retirement savings and having millions amassed long before retirement. The lifestyles and opportunity spectrums are wildly different between the two.
Investing 100k/year seems a bit much. With ~40% tax, the take home is closer to 120k. Add in Bay Area rent, leisure, random fixed costs(car insurance, significant others expenses, pets, etc) and its probably closer to 80k at best.

Though, I do wonder, how feasible would it be to end up with a 2+mill nest egg(or however much you need to live off the gains) by ~35 and retire right then and there.

(comment deleted)
I keep reading that seed investment firms consider it "good" if 4 out of every 10 startups they fund turn out to be something of value. Over time this is showing to be true.
Valuations are numbers coming from thin air,, I wonder what will be the total profits of YC company. When i ask profits ,its actual net profit ,, not some accounting gimmick !
Number of companies funded by YC that have shut down: ~300

Does this include aqui-hires?

Also, would you know how many companies rejected by YC are worth more than $1 billion? Zero or non-zero? :)

Why would it include acqui-hires? Companies rarely broadcast the fact that they're acqui-hiring, so it's not always a black-and-white issue.

Also, saying a company was acqui-hired makes a VC look less successful, so they have an incentive not to be honest about it.

As a clarification, by 'shutdown' do you mean failed, or does that include people who sold or were aquihired?

Good to see the 90% of startups fail 'statistic' blown out the water with some facts!

> Good to see the 90% of startups fail 'statistic' blown out the water with some facts!

If YC companies were representative of all startups there wouldn't be any point in going through YC.

(comment deleted)
(comment deleted)
~11% of their investments were in the last batch so I wouldn't read the ~30% dead rate as conclusive, a lot will need more time to die.
It would be great to get stats about exits as well. How many exits? Total valuation at exit? Median/average age at exit? etc.
Ok, so I did some quick numbers on performance. It looks like the first batch was in 2005, meaning ycombinator has been at this for 10 years. Have they always funded with 120k? Assuming that (and not present valuing the older money):

Total Investment: ~$131,600,000 Total Companies Value: >$65,000,000,000 YCombinator's 7% Value: $4,550,000,000 Total Return: 3357% Annualized Return: 42.5%

Obviously, it costs more than the initial investment, but these are really nice numbers if compared to a mutual fund, etc. Did I make any mistaken assumptions?

(comment deleted)
You're not factoring dilution in at all, for one. Also, the 120k is more recent.
Need to take into account dilution in subsequent fundings. So much less than 7%. #s are still out of this world though
The $120K is only since 2014:

https://blog.ycombinator.com/the-new-deal

From 2012-2014, YC companies received an automatic $85K follow-on investment from a syndicate of investors, though this was not funded by YC. From 2011-2012, they received an automatic $150K from Yuri Milner. From 2005-2011, it was $6K/founder (I think very early on, it was a flat $20K). So YC's capital expended has been a lot less over time than the current deal would reflect.

Need to take dilution into account here. If we assume a (optimistic?) 1% stake, that's $650M.

YC recently (1-2 years ago?) started giving 120k. Before that, it was 10-20k. So let's say half of what you said is the total investment = 65M.

That's a 10X return. I actually was expecting higher. Am I missing something in my numbers?

that's probably roughly right. But why did you expect higher? they are generating somewhere around 20-30% IRR. That seems massive in the current low yield low return environment
20-30% irr is not very high given the risk involved. a lower risk portfolio that is leveraged can give you this return quite easily (in a mechanical sense). in that case you have or take a low-risk low return investment and add risk and return by adding say 10x leverage to a 3% return.

In this case, the more relevant number is $$ in and $$ out. and also the optionality to continue the business going forward. That is what seperates the two cases out, and arguably make the one quite distinct from the latter.

A leveraged portfolio is a bad comparison because you could also leverage this portfolio. Let's compare apples to apples. There is no unlevered investment right now that can give you such high returns, that I can think of.
No, its not a bad comparison. Any decent investment vehicle is not putting money out the door unless they are clearing 20% in their base case returns. The point is you expressly RAROC the two portfolios. A retail investor is not going to look at any of this as relevant, so avoiding leverage is pointless. Anyone benchmarking this asset class has access to leverage (and in this market, at low yields).
right. But since I can lever anything, leverage as a factor is irrelevant. I could also lever this portfolio and get 100%. The point is you need to compare apples to apples, not apples to oranges.
(comment deleted)
This is incredibly off in many ways. Someday we will just release our numbers.
Sorry about all the FUD around YC's success. I'm curious where all this negative sentiment is coming from. Reminds me of this gem: https://news.ycombinator.com/item?id=35079
Because we expect more from YC. No one doubts that YC has been a hugely positive impact in the startup world. It seems below YC's standard to share selective metrics that might not tell the whole story.
It would be great to see how many jobs YC has created too.
Without stating where / how, can you give us an indication of dilution in a "YC Never invests again, company does series A, B & C" scenario?

It is complicated, I am sure, but YC participation in subsequent rounds, but just as a "what to expect as an investor", what dilution would / does one expect?

(comment deleted)
I think there's a big elephant in the room here. Companies usually take time to fail, especially with VC funding. The companies from the newer batches are going to be skewing those fail statistics. If you only look at companies from 4 years ago and older, you might get a more realistic impression.

It still won't be 90% though, but YC companies are widely known to not be representative. It's the highest profile accelerator, so it attracts the best talent. Just like people who graduate from Ivy League schools earn more on average, but mostly that's because they attract people who are above average to begin with.

But if they choose to provide stats each year thatll work out over time right? I mean you have to start somewhere
Is there a calculation for REAL ROI?

Not based on current valuation but based on actual exits and money returned to YC?

Doesn't look like. But 5% of 65 billion is 3.25 billion. Nice ballpark stats.