Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
YC W13 Will Be Smaller (ycombinator.com)
311 points by jkopelman on Dec 2, 2012 | hide | past | favorite | 161 comments


Why was 66 ok and 84 not?

Maybe it had less to do with magnitude than direction.

I look forward to reading about yc start-ups, but increasingly find myself shaking my head, wondering how some of them will ever amount to sustainable businesses. I've always attributed this to the fact that yc must know a whole lot more than me.

But pg's recent disclosure that so many yc start-ups have co-founder issues really got me wondering. I find it unimaginable that any team can work itself into such a good position and then blow it away over seemingly petty issues. Is it possible that some trend other than scale is at work here?


One thing to remember is that the goal in many cases is not to build a sustainable business, but prove that there is a market for this thing (of any size). It's not unrealistic for some of the more oddball single utility "companies" to find themselves aquired and rolled up with other strategic acquisitions, and the lot of them presented as some sort of B2B enterprise package worth far more when summed up.

In other words, with VC backed companies, the thing these companies are making isn't really the product in many cases is it? It's the companies themselves. Just like a gear isn't really much, but add some springs, and a few other odds and ends and you have a watch that people will buy. But it takes a bigger picture visionary to figure out that the company that makes gears and the company that makes springs should be both acquired to make that watch company. Till then, they have to try and find their way by either spending their investment dollars till they can find that visionary, or find a market supplying micro-miniature gears to other watch making companies.

The founders and investors are rewarded for taking the high risk of finding out that there is a market for micro-miniature gears after the acquisition. But until then everybody wonders why one would want to get into gear making in the first place?

YC is basically in the meta-business business.


The real goal of YC is likely to create a great productive and interesting life for the organizers. If the previous number was too hectic and they're scaling back, congratulations to them for choosing quality over volume.

Paul Graham's recent essays on growth and startup ideas reflect a much deeper and more nuanced insight into the startup process than you give him credit.


There have always been lots of startups working on crazy sounding ideas, since the first batch. I don't think the proportion of those has changed.

(If anyone is wondering why we don't just reject all the startups with crazy sounding ideas, Airbnb was one of them.)


I don't understand why Airbnb was considered as a crazy idea and I hear literally everyone refer to it as an example of crazy idea. I always thought it was a genius idea from the first time I heard about it. But the way I see things, it doesn't matter what the idea really is as long as there is a market (not too big nor too small) and the team's ability for solid execution of the idea (kick ass product, partnerships and vision).


It used to be "sleep on the floor during conferences on air mattresses, with the host providing airbeds". I personally would have had a hard time extrapolating from "surge capacity for huge numbers of people during events" to "routine spare-room or whole-house rental", which seems a lot more like homaway/vrbo/etc. to me., a good business. The former is more like couchsurfing combined with the Red Cross, and IMO not good as a business.


You probably heard about it post-yc. It was initially a thing for renting out airbeds (airbedandbreakfast.com)


I think even that is a great idea...albeit with a much smaller market capability than what airbnb is now.


That's pretty much what http://www.couchsurfing.org do, but for free. Clearly there is both supply and demand for such a marketplace. Were airbnbs or YC aware of couchsurfing?


It wasn't a crazy idea at the time. couchsurfing.com had been a thing for years. "couchsurfing with a better UI" doesn't sound like a billion dollar idea, but it doesn't sound crazy either.


What everyone else said, PLUS it was a marketplace startup in a crowded space. Marketplace startups are ridiculously hard to curate, Airbnb did some really great hacks (with Craigslist especially) to finally make it work.

But there were many people who thought this kind of thing could be huge, and plus the non-profit Couch Surfing was pretty big at that point if I'm not mistaken.


Market is semi-legal. How do you gain traction? Network effects (that whole traction thing (re: gaming/tuning Craiglist)), liability, etc.

Retrospective analysis biases "initial" perceptions post-facto.


Ok, so what were the predictors of failure?

IMO, if you can talk about these features/predictors, they are true predictors. Otherwise, if people can game the system if they knew these features, I'd think they're more artifacts about your current process than attributes of failing startups (and if it turns out that these features are actually causative of failure, rather than just predictive, everyone can be a better startup too).


The fact that one can conceal a flaw doesn't mean it's not a real flaw.


In your defense, after the first few rounds of YC I remember VCs and bloggers squawking about all YC companies being features and not products, and that you were deliberately mass producing acqui-hires.

To me, it's clear that the quality of the average YC company has gone up.


"Things are always breaking at YC, because our strategy is to find bottlenecks by hitting them."

Can you share the details of what broke and how?

Even if you won't share your strategies for detecting & fixing the bottlenecks, I'd love some background on what they are.


The craziest ideas are usually "frighteningly ambitious" ones, I guess. So, it's easy to understand that these are the ones you, guys, like most.


> so many yc start-ups have co-founder issues

My guess would be that it's the number of "companies in simultaneous crisis" that grew beyond control. I used to warn new managers that somewhere just above 8 employees is the point where you always have one that is going through a serious illness in the family, divorce, addiction to drugs/video games, or something else that will take up all of your attention (if you let it).

Extrapolating to companies, I'd bet that the senior YC staff can handle, say, 5 companies in crisis (e.g., "co-founder issues") but growing to 10 was just too much and things started falling through the cracks. Or became unfun, since when you're handling crises you're ignoring the stars, which is a very common new manager mistake.

And unfortunately you can't scale the VCs to handle more crises; you either need more time from senior YC-crew or you have to find the equivalent of a human resources department to handle those issues and just stop trying to fix them yourself.


This is an interesting consideration with expedition/adventure team size. There is a tradeoff with redundancy (someone will have a spare x) that keeps the group moving, and the increasing liklihood of somone haveing a problem at any given time. You need to minimize the latter, b/c not just its slower/more comlex, but it eliminates the redundancy (your 1 spare is now used gone, but larger group still to worry about, etc). This is often why an ideal team size is 2, not 8-10.


I look forward to reading about yc start-ups, but increasingly find myself shaking my head, wondering how some of them will ever amount to sustainable businesses. I've always attributed this to the fact that yc must know a whole lot more than me.

I was at the first startup school (in Cambridge) and I remember thinking that about reddit whereas the one I thought most promising (click fraud detection) seems to have flopped.

Other than making a bet on the people, I wonder if YC also has no idea about which startups will succeed.


ClickFacts still operates, although it doesn't have such astonishing story as Reddit does. More info: http://thenextweb.com/insider/2012/08/05/y-combinators-first...


Reddit is still not profitable.

It's an amazing edifice of the Internet, but it is not profitable.


I wouldn't be surprised if that quickly changes, since the ad placeholders (center right of page) have recently become revenue-generating ads.


> sustainable businesses

Given that a lot of YC seems to come from PG's own experience, which involved selling Viaweb to Yahoo, I'm not sure "long term sustainable business" is necessarily what these guys are going to shoot for. And I think that's ok too: some people are just cut out for taking a business from 0 to 80, but not managing it long term.


You think that the experience of selling 1 company (viaweb) trumps over the experience of funding over 300 companies across 7 years, 2 of which are worth at least 1 billion dollars?

Thats like saying, given that Apple experience comes from building the first computer they must not know anything about mobile.

YC experience comes from doing YC itself.


Apple did not build the first computer.


pendatic. His point was that YC's expertise with advising startups comes from doing YC, not from selling Viaweb to Yahoo (it was, but only at first)--just as Apple's expertise doing the iPhone came from doing the iPhone, not from making computers (it was, but only at first). Whether Apple made the first computer or not is irrelevant to the point he was trying to make.


How is that a minor detail? The first electronic computer was built in 1946. Apple hasn't even existed for half of the time electronic computers have existed on Earth.

Seriously folks, this is some dumb, dumb stuff. Even the senator who gave the infamous "tubes" speech probably knew that the Apple didn't invent computers.


> You think that the experience of selling 1 company (viaweb) trumps over the experience of funding over 300 companies across 7 years, 2 of which are worth at least 1 billion dollars?

No, but I really got the impression as YC grew over the years, from what was reported of it, that PG modeled it, at least initially, after his own experience. Doubtless it has changed over time as they've learned.

My point was merely that there is not necessarily a "build a company to last years" in the group's "DNA".


"A piece of software to build eCommerce websites that takes a portion of revenue." doesn't sound like a long-term sustainable business?

The Viaweb crew took an exit - a good exit, in my opinion - but their business model was sound, profitable, and strong.


> The Viaweb crew took an exit - a good exit, in my opinion - but their business model was sound, profitable, and strong.

Seems that way to me too. But they opted for an exit, rather than deciding to work at it long term.

I think the world is better off with YCombinator, rather than some evolved form of Viaweb, myself, but the point was merely that for them, not doing something "long term" is not a failure.


Teams don't usually blow it away over petty issues. What they do is manage to ignore/avoid/be blind to serious issues for stretches, especially when things (on paper) seem to be going well.


That does happen. But having been in my fair share of partner disputes, banal and stupid issues blow up teams just as often.

When you're on the outside looking at a team, it must seem as if the members share a purposeful determination. And they often do. But just as often, the startup effort simply becomes a new norm, inhabited by human personalities, inclined towards squabbling just like dorm roommates.

Membership in a startup team can amplify drama as much as suppress it. It fuels personal squabbles with a sense of import. The best way to resolve the overwhelming majority of conflicts is to walk away and let them cool down naturally. But in a startup, it's easy to tell yourself "we have to resolve this problem right now; the fate of the company is at stake!"

It's also probably the case that a lot of long-term simmering "important" issues actually started out as manageable problems that could have been dealt with, but for poor conflict handling early on; either the issue is broached and turns into a scalding fight, or fear of that fight keeps the team from addressing it early enough.

You would think this would mean founders with stable marriages would have an advantage. :)


tptacek, whenever I read your stuff, I think "he gets it".

I bet stable founders have more stable relationships with their SO's.

Dave Ramsey writes about how his company uses dinners with spouses as a part of their interview process. They won't hire anyone "married to crazy". They also fire anyone who cheats on their spouse. I found his book quite misogynistic and offensive (not to mention having the worst title ever), but I wonder if he's onto something.

The idea of YC (or any company) using personal relationships as part of their filtering process makes me extremely uncomfortable, but it's interesting to think about.


Dave Ramsey probably either doesn't really do this or has never asked a lawyer about it, and is in any case offering incredibly bad advice. It is unlawful to discriminate against someone on the basis of marital status. When you invite a candidate out to dinner with their spouse and then later decline them, you've communicated two things:

* That you were serious enough about the candidate to take them and their spouse to dinner; in other words, that you were probably inclined to hire them.

* That it is possible that the dinner, which only could have been an issue for candidates with spouses, had something to do with the decision not to hire.

Investment decisions are very different from employment decisions. You can probably discriminate against investment opportunities for any reason whatsoever, including race or religion. But dinners with spouses as part of an interview process seems like a great way to get sued repeatedly.


As far as I can tell it has nothing to do with marital status, ie married or single. It has to do if the person is crazy. You can hire or not hire people based on personality and fit with the company, so I'm not sure why the spouses personality and their relationship fit for the company would be off limits.


It obviously does discriminate against people by marital status, because if you're not married, you're not liable to be rejected based on the spouse-dinner filter.

It's a little like suggesting that you're not filtering based on religion if you ban yarmulkes.


I'd say its closer to finding out the single guy is in a gang and since he spends time with crazy people you don't want him. But I guess your point is at least a bit valid. That being said, it seems only a few states have a problem with using married status in hiring.


IANAL but I don't believe marital status falls in the protected category. http://en.wikipedia.org/wiki/Protected_class


Not federally, but yes, for purposes of employment, in many states including California, New York, Illinois, and Washington.


As tptacek says, many states do ban discrimination on the basis of marital status, and they have a good reason to - historically most businesses refused to employ married women at all on the rather sexist basis that they should be at home raising kids.


In my previous job I asked my dev lead from Mumbai to help me with interviewing candidates to join us on a project here in California. On a conference call with a female interviewee, he asked if she was married. He hadn't asked the previous two candidates (males) that we'd interviewed.

After the call I told him that was illegal here and he can't ask that question in an interview. He was shocked. He said Indian women are expected to devote time to their families once they are married. Single women have much more time to devote to the project and would be much better for our team.

I've known that women have been discriminated in this way for a long time, but I had never encountered it personally until then.


I suspect the more likely reason is that they didn't want to lose employees who decided to quit to go home and raise kids.


That was my first thought. "Isn't this illegal????"


Federal laws that prevent discrimination as an employer don't apply to investors.


I meant in the case of Dave Ramsey. That guy is crazy.


I was just thinking about this today. To interview a potential co-founder I'd have a meal with the person and their SO. Then I'd observe the dynamic between them. Does the potential co-founder treat the SO poorly? Or is it reversed? Ideally they treat each other well. But there's more subtle hints to observe. Do they interrupt each other? If so what's the fallout?

The point is to assess the person's "people skills." I think "people skills" or in general "emotional maturity" is by far the most important factor in someone you need to work with on an intense level. Be that co-founder, SO etc.


The whole idea some people have of meeting a cofounder a few times (especially in an artificial environment like a hackathon) and then making an all-of-nothing decision is flawed, IMO. The best process is to spend a LOT of time with the person, and work with him on progressively larger projects. Yes, how they interact with people is an important factor, but I'd try to spend many months (or years) on the process, not a few meals.


Or how they treat the staff at a dinner. Many other routes vs making some people uncomfortable that you may "require" a dinner with a SO. If they don't have a SO, what, invite a friend to an interview?


Good point. I wouldn't formally require the SO. I'd just phrase the invite to make it clear that my SO and I would like to have dinner. If they have an SO but don't bring them that's another potential signal (could be good or bad).


To sort of support this, I know a VC that won't work with any divorced person until two years after the divorce. He says it takes that long to get back on course. And he has been through one, so he knows.


I understand that investing in a person is very different from employing them, but this still creeps me out. It would be disturbing to see this become another valley meme (it is probably illegal to discriminate against employees based on marital status, for what it's worth).


For a place that exists because of an engineer field, the Valley seem pretty prone to cargo cult thinking, fads, and superstition.


I get your point, but there a lot of non-engineers in this business. I'm also not sure that engineers have that much of an edge in avoiding cargo cult behavior over other professions with a similar overall level of education and compensation. I'm talking about engineers as a class, not an idealized Engineer.


Was not in the valley; actually in the midwest.

And this was entirely in the context of a startup.


Wow. Getting divorced seems rough enough already; sucks to be judged unemployable on top of that.


Was only in the context of a startup.


I wouldn't take on a cofounder/exec hire/major investor without spending a little time with them and their partner. Not because I think I'll detect some red flag, but because my wife is great at reading people and might observe something important.


I might view it in reverse: having been in a start-up is an advantage/practice for stable marriage.


s/stable marriage/early mission to Mars


There's a metaphor that's going to get me hit with a couch pillow in here somewhere.


"shaking my head, wondering how some of them will ever amount to sustainable businesses."

pg wrote about this -- the best startups seem like bad ideas but are good ideas. They seem lame. I don't think there is any way for you to judge a startup by "reading about" it. You haven't met the founders, etc.


For the best companies, the MVP will be like the tip of an iceberg. A small, innocous thing that is visible. But will large mass/depth to the idea. Like IBM negotiaing will Bill Gates on DOS. So, yes there is something to this...most articles don't capture the founders deeper insights (if it exists), and often thats intentional (strategic disclosure).


"I find it unimaginable that any team can work itself into such a good position and then blow it away over seemingly petty issues."

Why are you calling them petty?

Do you find divorce similarly unimaginable? Most co-founders know each other less well than most newlyweds and spend way more time together.


And I find it unimaginable that people divorce over petty issues as well.


IMO, having been involved in a co-founder dispute, in hindsight the issues were not at all petty and reflected major differences in attitudes and long-term goals of the co-founders. They SEEMED petty at the time, because people, specifically young people, tend to do a poor job of identifying core causes for interpersonal disputes. One risk of investing in young founders is that this sort of thing will happen at a higher rate than if they were older, but for a number of reasons that PG has discussed at length, the pros of generally investing in younger founders trump the cons relating to emotional immaturity.


Goals change over time. Maybe one person thinks their best exit is via the publicity of running the operation, while another person thinks their best exit is via selling the company, while another thinks the best exit is to extract profits from the company.

And companies don't found themselves to be easy to break up, for the same reasons people don't set up prenups on marriages.


But pg's recent disclosure that so many yc start-ups have co-founder issues really got me wondering. I find it unimaginable that any team can work itself into such a good position and then blow it away over seemingly petty issues.

People going into YC with obvious dysfunctions aren't going to get a "yes". The remaining groups are those that:

a) are good at managing stress and disagreement - so don't show any

b) haven't yet had a major stresses that will cause problems - so don't show any

Since dealing well with stress and disagreement is a learned skill I would imagine that YC's focus on younger founders may make the (a) group fairly small.

Getting accepted into YC immediately puts companies into a stressful situation (moving, spending 24/7 on startup, the "this is our chance to make it" vibe, etc.) - so folk in (b) who haven't had to figure out how to deal with disagreement well are going to have problems.


"The reason we accepted fewer applications was that in summer 2012 we grew too fast." I am curious; from a ROI perspective, are there any limits how big each YC batch could go or put another way why is "fast" fast? Can YC grow as a partners-based organization as much as demand exist or does PG's attention draw a limit, as would occur in a star-shaped organization structure.


I think it is scale. The success of YC is built on the success of their startups. Therefore, I don't see YC as a not a long term (I'm talking 20+ years) sustainable and scalable business model. Is their process reproducible? Yes. Are there deliverables reproducible? No.

If YC wants to continue to grow (financially), it could possibly:

1. Diversify - Perhaps start a pure VC wing

2. Segregate - Provide specific expertise in verticals (travel, hospitality, sports, etc), horizontals (like consumer web, mobile, etc), or regions (emerging markets, global markets, etc)

That being said, I can't help but think that YC employees probably struggle to understand it's ultimate financial growth strategy while at the same time realizing that in it's current state it's already an extremely valuable business and provides those involved with loads of continuous personal satisfaction both financially and psychologically. So while it may sound like I'm discrediting the organization, I'm dually jealous of the "good problems" they have.


What are some indicators as to whether a startup can and will survive co-founder drama? Co-founder drama is more common than not. YC has such a profound reputation and emphasis on co-founders, maybe the application should just directly ask for examples of past experience with co-founder drama and what people would do differently for the future.


I'd be curious to know what broke. Companies in the last batch having a harder time getting funding?


The short version is that we learned YC in its then form was O(n^2). As the batch size grew, it was hard for every partner to know exactly what each startup was doing. Especially since things often change fast at startups.

To grow bigger we have to (a) make it easier for each partner to know what each startup is doing, or (b) make it unnecessary. The answer will probably be a bit of both.

The symptom was not startups having trouble raising money (that seems to be working about like it usually does) but us partners always feeling confused. We were not ahead of the aircraft in the way we'd been in the past.


As the batch size grew, it was hard for every partner to know exactly what each startup was doing. Especially since things often change fast at startups.

You could be running up against Dunbar's number (http://en.wikipedia.org/wiki/Dunbars_number) for the more social (as opposed to purely technical) aspects of what is going on at each startup. 66 startups with an average of a little more than 2 people per startup is hitting the upper limit (150) of what the human brain can handle socially. 84 startups with the same average exceeds it.

In my own experience, once a company or social grouping exceeds 50 to 150 people (the more tightly bound the social group, the lower the count), things change noticeably, so I'll be watching with interest what YC does to get around this.


That is certainly one of the hypotheses. The reason I'm skeptical is that it seems such a fashionable wall to hit that it's almost tacky. Which I admit is not a valid reason to be skeptical.


Sounds like you've gathered a fair amount of your own data similar to what the “Startup Genome” project is aiming at.

http://techcrunch.com/2012/11/20/startup-genome-ranks-the-wo...

This would definitely make for interesting reading if/when you choose to share details.

Also, it sounds like at least part of the Y Combinator size issue is a version of Dunbar's Number, applied to business rather than social relationships.

http://en.wikipedia.org/wiki/Dunbar%27s_number


We wouldn't want to share everything on our list of predictors of failure, because some would stop working if applicants knew about them. Some though are things applicants could only fix by actually making their startups better, so I should probably write something about any of those I haven't already talked about.


A brief bullet-point list of the ones you can talk about (both those you have and haven't written up already) which have been statistically validated at YC sounds overwhelmingly valuable.

Including the Ns would be even more overwhelmingly valuable. I.e. I can calculate a Bayesian likelihood ratio and get some idea of the confidence from e.g.:

* Founder has green hair: Successes 10/80 Failures 40/122

However I would much rather have the brief bullet-point list than not have the bullet point list with lovely numbers attached.


Somehow I am afraid that this sets the stage for an endless discussion about the "hidden" ycombinator acceptance criteria. I am already looking forward to the "how I uncovered x of pg's criteria" posts.

Implying that there are secret rules to "a game" make it a highly attractive subject to try to discover these.


Yeah I was just going to point out Dunbar's number. At that point you have to introduce more layers of management, rather than everyone just knowing each other.


I was part of the YC batch at 66 and felt all the partners knew a lot about our start-up and could speak to us really easily, asked for updates, etc. Felt like a small group even though it was large.


Paul, of course, it's far easier to write out a suggestion than to actually do it, and free advice on the Internet is seldom worth what you pay for it. Anyhow, my idea might help...

I might be misunderstanding your "ahead of the aircraft" statement, but personally, I think it's unreasonable for you to expect YC partners to be "ahead" of the startups in a batch. In fact, if YC partners are not behind in understanding a startup, then that startup is in trouble. The startups may not have the benefit of experience possessed by the YC partners, the startup founders should always know the state of their own business and market better than the YC partners do.

I believe one of the reasons why you regularly advocate being concise is due to your own personal overload problems as well as the stated problem of all YC partners not keeping up with all of all of the issues in all of the startups in a given batch. It's the classic many-to-many communication problem where the ability of partners to assist and advise startups is limited by their ability to stay informed.

Even if you're never able to isolate every single metric a YC partner should know, identifying as many of these metrics as possible and having a reporting mechanism would still reduce the load on YC partners. Though its not publicly known, I'd bet you do already do something along these lines, either formally or informally. Your "growth.html" essay seems to hint in this direction. None the less, making improvements to your identification, collection, and reporting of useful metrics should still be a worthwhile investment.

I sincerely doubt you would want my help implementing a startup metric collection and communication system, but I'll offer to help anyhow. It's a fascinating problem and it seems like a really fun challenge.


66 still seems a very high count to even remember all the names, leave alone keeping track of where they are.


I think they cheat initially by remembering the company name, e.g. calling an entire team "the Airbnbs" or something, until they get to know the people individually. So it's at least 66 groups vs. 150 people. The company names or nicknames are usually also way more distinctive than the ~10 guys named John or Jack or Mike in every group.


the O(n^2) thing sounds reasonable until i start to think about it. but then i get confused. that's some kind of cost, but my naive interpretation (number of partners scales with applicants, so total interactions is n^2) spreads that cost across O(n) partners, so the cost per partner is O(n).

what am i missing? am i just being too literal, or have you identified some real process that is O(n^2) per person, or is the total cost more important than i think? or am i just being dumb?

or maybe even handling O(n) per partner is too much. i guess that may be all it is. my initial reading was that O(n^2) explained why 66 and 84 were so different (when linearly, they're pretty similar).


The way I read it, he's referring to communication channels needed for everyone to be on the same page (see Metcalfe's law, for example: http://en.wikipedia.org/wiki/Metcalfes_law). Traditionally, with a homogeneous group of people, the count of communication channels will be O(n^2), as you can see in the images at the top of the wikipedia article, because everyone needs to talk to everyone else.

In this case, it's not homogeneous; there are n startups and m partners, but if every partner needs to know what's going on with every startup, you still have O(mn) communication channels. If, for example, you assigned n/m startups to each partner, the total communication channels needed would reduce to O(m+n). I believe pg is referring to something along those lines when he says he'd like to make it unnecessary for each partner to know what each startup is doing.


You're not being dumb. You just sound similar to, say, a graduate student doing theory who hasn't been outside in a while. :)


(a) can be solved using technology to some extent. What you need is a tool to record what a startup is doing and why certain decisions were made.


So it's not so much the batch size that matters. It's the quality of the investor.

From the about page: > All venture investors supply some combination of money and help. In our case the money is by far the smaller component. In fact, many of the startups we fund don't need the money.

This is intriguing, because investor as I imagine would like to spread there bets as much as possible.

From an investors point of view, isn't it counter intuitive to solidify around a smaller number of start ups.

Although, I agree that they should invest more time, as a path to success. Showing care for your investments seems all around wise.

Just food for thought, I'm not sure if I have a point most of us don't already realize.


Virtually every other investor makes an order of magnitude fewer investments every year.


+1

Pg, can you elaborate on this a bit?


You mentioned looking for predictors of failure. What have you learned since writing this essay on mistakes that kill startups? http://paulgraham.com/startupmistakes.html


I wish there were a way you could publish the predictors of failure in a way which wouldn't let people "game" the system unreasonably.

Actually, I think you have always published many of the predictors of failure. http://www.paulgraham.com/startupmistakes.html seems like a good start, but there are probably more specific indicators during an interview or during 3 months before Demo Day.

The ones which seem most relevant are all variants "not making something people want" -- either not making something effectively at all, or that which you make is a bad idea/market, or making something which is a good idea but crap implementation. Obviously several potential causes of each.


Is there any chance that there's a small "valley gold rush" effect going on, sort of what happened in the late 90's and early 00's in SV when people saw the valley not as an opportunity to build something awesome but rather ONLY to get funding and get rich?

From what PG seems to be saying, it's more of an organizational problem, and I'll take him at his word. However, I just feel like I've seen a lot of (admittedly impressive) people applying to YC because they see it as the path to easy VC funding and lots of press only. And maybe a resume booster. I can see the effects of that being that it becomes harder to identify who is earnest and who is not, especially with such impressive people throwing themselves into the pool.

Either way, I think it's nothing but a positive to see YC continuing to innovate and tweak.


I think that there's an element of that. People who are just joining yc for the name are there for the wrong reason.


Considering that you guys had the same funding formula (7k for 7%, give or take) in the funding dark ages of 2004-2005 as you have now when it's much easier to raise money, I think everybody who takes a deal from you, as opposed to better terms from someone else, is there at least partially for the name.


Certainly the name has value, and smart founders will use that, but if that's all they are after, then they're missing the most valuable bits (and are therefore probably not very effective founders, and a bad investment).


That's totally understandable, I can't even look down on doing it for the "wrong" reason if I'm being honest. Sad for me if it diminishes the likelihood that something that will actually change my life (like Codecademy did) will come around.


Our company interviewed for W13 and was not accepted. Some thoughts:

If you want to make important changes that could affect a startup's interest in accepting YC (financial terms) or make it much harder to get in (even at the interview stage) than it has been in the past you have a duty to do so before you hook hundreds of entrepreneurs into committing to sit with you for 10 minutes.

It was a good experience to prepare for and attend YC's interviews. However, it took valuable time away from the most important people in our company, both from a preparation standpoint and a distraction from execution.

Our company is based on the west coast, and the $800 in travel reimbursement did not cover our costs to attend (we went ~$200 over and were very lean. You upped the criteria, lowered the reward and didn't completely cover the cost to try out. Small dollars and a few days matter a lot to a company at our stage.

It is one thing to tweak your program and make it harder to get in. It is another to do so in secret or after you've gotten entrepreneurs' hopes, dollars and time committed to trying.

I hung out at the YC office before and after our interview, and the environment is not like you project: http://static.picwing.com/18075/e_106219_t540.jpg It is actually pretty ruthless, and I picked up a sense of realized power and eliteness.

As an example, YC put a really nice selection of organic-looking food right next to entrepreneur registration table. This food was not for the entrepreneurs, yet it was on full display. I approached this food by accident and was actually told by a YC staff member (in a tone that suggested a silly mistake) that there were "bits and scraps" available on the other table for us.

That is kind of what it felt like in this process. A bounty of food that is kept within sight, but actually quietly kept quite far out of reach. It ended up feeling not that much different from pitching a traditional VC.


I never pitched a VC. I actually never met such mythical creature, maybe you were about to touch his unicorn salad. Anyways, explanations for this:

- they are actually giving a heads up on how most of people will treat you or;

- there are just too many interviews and no time to make anyone fell comfortable.

Edit ====

Anyways, I remember in 2006 when I barely could speak english and was doing a master degree in the US. I needed to find a part time job to help me pay for unexpected expenses.

There was an interview at an investment firm: I felt like I was worth less than nothing. I almost cried after leaving the building. Anyways, 20 minutes after I left they called me and offered a position.

When we feel vulnerable anything people say to us we tend take it the wrong way or overreact.

Good luck with your venture.


I echo your thoughts on the overall process for interviewing at YC. We interviewed on Wednesday and they were running late on interviews, but didn't inform us. We had to ask several times, as opposed to getting an update from their team.

Morning food situation was also awkward as you mentioned.


Potential ways to get past the 66 bottleneck: 1) Try it in the winter. I theorize (without as much data, obviously) that "summer projects" are more likely to be drama filled than something Jan-Mar. 2) Consider 2.5 or 3 classes. You'd still end up with 150+ companies per year. I know it would sort of screw with the current schedule, but there could be some overlap, although at the cost of partner happiness. Maybe have the .5 session be split over a longer period of time, with interviews happening slightly offset from S and demo day slightly offset from W. The longer period would be useful for certain kinds of startups. 3) Invest in a human cloning startup. Once there are 80 clones of each partner, scaling becomes easier. This may take tens or hundreds of years to be effective.


But would the defective PG clones in the early years fund too many bad startups?


I'm curious: what did you find were the predictors of failure? Was it a matter of focusing on problems (with the assumption that the teams with the least problems were, on balance at least, more likely to be successful)? Or did you find that there were certain counterintuitive factors that made a team likely to fail, even if they seemed solid?


I'd be interested in this too. I'd like to know if it's in a format where you could judge yourself/your own idea and see what score comes out.


Maybe it is the large batches themselves. What if, instead of a bi-annual intake of two large batches YC moved to an always open intake with classes starting every month. That is, to stream many small batches continuously. Effectively running 12 academic years concurrently rather than 2.

Everything from lean manufacturing to project management has come out in favour of the small batch size, maybe the thing to do is to work out how to split the existing large batches into many smaller batches.

This might also help increase your learning as you would have a much tighter feedback loop.


50 is still a huge number. Airbnb's class, just 4 years back, was only 16.


I expect significantly more people applied for this batch than Airbnb's batch.


Might this have something to do with Dunbar's Number of ~150?

http://en.m.wikipedia.org/wiki/Dunbar%27s_number


An idea I had to scale beyond this number might be to use a spaced repetition system in conjunction with an email client https://en.wikipedia.org/wiki/Spaced_repetition


> We don't plan to stay at 50, or whatever exact number of startups this batch ends up having. We've never had a predetermined batch size; that's just the number we ended up with when we tightened our filters as much as we could.

Congratulations to those of you that were accepted into W13. It looks like W13 was the most difficult to get into yet.


Did you interview? Presumably you spoke to previous interviewees. Any difference in questions that you noticed?


Yes, I did and unfortunately I didn't get in. The interview was nothing like I expected, i.e. it was just PG and I talking about the idea. I intend to write a few blog posts covering the different stages of the application process as long as PG doesn't mind.


That would be awesome.


It would be very interesting to hear more about what the key predictors of failure you looked for were?


I'm curious how you fix the post-lunch bias without overcompensating? Eat small snacks throughout the day instead of lunch?


Is this from "Thinking, Fast and Slow"? If so, how do you know it's not actually the break/rest that helps, rather than the food? I don't see any controls in that experiment take a break of equal length but not eating, and others who ate without a break.


Anecdote: I discovered when quitting nicotine that first thing in the morning and after lunch were the times when my willpower was weakest. Your mileage may vary.


Didn't "Thinking, Fast and Slow" mention that by injecting glucose again you got back to the initial state?


It might have, though I don't remember anything like that. I did look, because I disputed the conclusion, but I might have missed it anyway.


Eating isn't the problem. The problem is making decisions throughout the day. It takes willpower to say no and ask tough questions, and your willpower gets drained through the day.


He's describing the opposite effect – people are more likely to say yes in the morning and right after lunch, times when they would arguably have the most willpower.


Glucose replenishment restores willpower significantly.


I think if I were invited to an interview I'd bring chocolate.


Novelty chocolate stamped with your startup's name, to make it a little less transparent. :-)


If you know how strong the bias is, couldn't you just weight the post-lunch scores downward by the same amount?


What does it mean that the interview process is "decentralized"? Seems pretty centralized to me: only YC partners or friends of partners do the interviews, and they even require traveling to Mountain View in general


Three parallel interview tracks.


How is it determined who is on which of the three interviewer groups? If it's static, is there a group with a higher acceptance rate than others?


They're shuffled, and then afterwards we swap a few to tracks where the people have domain expertise. E.g. most of the hardware startups got moved to our track, because we had Trevor.


I think 50 is even a large number. YC's real value to investors is selectivity. As long YC maintains that brand, investors will continue to have a YC-biased perception making it easier for these startups to raise funding. The same selectivity makes getting into YC enticing as it validates them and their ideas. I'm sure the reason behind the cutback has been scalability ( I made that comment before and it made PG defensive a bit) however, the move ensures the longevity of YC's mystique.


Their number sounds like it was based on the partners' personal feelings of being able to be of service and value to their startups, their portfolio. It seems that the aspect of YC the partners are trying to preserve is their ability to be personally involved to some degree rather than just straight selectivity.


Did YC bring in less interviewees this batch than summer 2012?


Slightly fewer, because we used the last 3 interview slots each day to have another track re-interview startups we were on the fence about. That was one of our new techniques for avoiding mistakes. IIRC there were 20 slots both times, so we therefore invited 17/20 as many startups as last time.


I assume by decreasing the amount of startups interviewed in order to re-interview a few of the startups, you guys believe that you make more mistakes in the interview process than in the application review process. That is kind of surprising.


Oops, there were 19 slots last time, so we invited 17/19 as many this time.


There were a lot of interviews this week. Far more people were interviewed than were accepted.


I think the math behind YC and incubators can sometimes falsely justify an increase in the quantity of companies accepted.

If you look at each class and say, "Well, last year we had 5 breakouts out of 50. Why not increase that denominator to change the numerator?"

Unfortunately when you do this too quickly, the numerator doesn't change. It stays the same or sometimes even decreases (in my example, holds at 5).

Staying focused while growing is a very difficult thing to do.


At 86 companies YC is more likely to hit up against the limits of Dubar's number than at 66 companies (or 50) - assuming, of course, that Dunbar's number is meaningful.

As an aside, funding to the point of creating financial carcasses in the previous cycle could have effectively lowered the Dunbar number for that batch.

http://en.wikipedia.org/wiki/Dunbars_number


I think that the real story is that alternative investments (e.g. venture capital) have really dried up. Venture capital firms across the valley are having a very hard time raising cash. Groupon, Zynga, Pandora, and the disappointment with Facebook's IPO have pushed away a lot of institutional investors. I don't see capital returning anytime soon until we see some big, successful Silicon Valley IPO stories.

(insider)


Actually the thing that struck us most this last cycle was the huge number of (often unpromising) startups that had already raised several hundred k. If anything it seems like there is a glut of early-stage money, not a dearth.


It would be an interesting control if you re-interviewed YC companies 6, 12, 24 months out from YC, to calibrate the interviews. I guess you can just watch the video of the interviews of the successful or unsuccessful startups.


Obviously in the case of an unpromising startup, you don't care, but how does that 200k+ change your perception of a potentially promising startup? You've blogged a bit about this before, but have you ever asked yourself (or the applicants) "Why are these guys even sitting in front of us?"


It's interesting that Josh Kopelman submitted this.


I'd guess there is some correlation with time spent on HN, times of access, comment rate, language usage, etc. HN may be one of the largest data sets on a candidate that YC has, so as big brotherly as it sounds, I can't imagine YC could resist analyzing it.

It might sound trivial, but it is probably a good indicator of impulse control, attitude, and cleverness.


What were the predictors of failure ?


I think it was the right decision, as said, the result of going mainstream, has started to show, if instinct of more than 1-2 partners say that its better to slow down a bit >> rethink >> adapt then its best to do so. Slowing down won't hurt you in any way.

Perhaps the answer is quality, due to a overshoot of no of companies funded, the quality went down, and then it hit you guys, that 'hey! the last of many haven't been any AirBnB or Reddit' so its better to slow down a bit, and let the levels go up again on it.

"Excessive fishing in a lake can be best balanced by under fishing" that is what this looks like to me.


Give yourselves more time by adding a month to each session. Reduce the number of sessions per year and increase the duration of each.


Dunbar's number is the amount of social relationships an average human can have: 150.

Maybe at 84 startups the total number of founders went above 150?


The whole program along with the partners probably pushed it above 150. As is the case with many groups going past 150, you'll have to start implementing other social structures besides relying on personal relationships to keep it going at the same pace.

Going to be interesting to see what happens, especially due to Y-Combinator being rather a unique case in this regard.


I am happy with the decision.

Having a lot of startups in every batch reduces the focus of YC partners In these situations, even startups with a very good potential (like Airbnb & Dropbox) won't be able to achieve much.

YC can increase the number of partners, advisors, mentors or cycles per year (instead of diluting their time over a lot of startups).


YC has gone mainstream , highly competitive, but so are the egos of the cofounders, not everybody is willing to work straight for 12-24 months for a potential exit on the same idea


12-24?

If you're successful you'll be working for far longer than that. Even if you're not successful good teams will continue to iterate or change ideas, which will often take far longer than 24 months.


What are the indicators of failure that you focused on?


Just curious what kind of failure it was? They can't say 'nobody of the accepted startups succeeded' because it's too early to tell that.


What's considered a failure? Startups that broke up, shut down or didn't raise? Isn't it too early to tell for most companies?


Curious: what methods are you using for your analysis? Feeding data in some software, or just pen+paper+brain?


I think smaller is always better, it would be fine to take on 10 companies and just hot house them. I think that would be awesome.


Were the choices announced?


manage growth --- always the hardest thing for startups and those around them


Must be due to the Series A crunch eh?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: