April 05, 2017
I went through Y Combinator with 80,000 Hours, a non-profit, in summer 2015. It had a huge impact on the organization, and as a result we grew our key metric 12x the following year. I subsequently moved to focus on my other non-profit, the Centre for Effective Altruism. Y Combinator had been such a good experience that I applied to take the Centre for Effective Altruism through, too. We got in (I believe I’m the only non-profit founder to have gone through twice), and are currently in the Winter 2017 batch. It’s just as good as the first time.
The fact that YC accepts non-profits isn’t yet well-known, and even fewer people seem to know what a good opportunity it is, or how they should think about it. So in this post I’m going to explain the program a little more. I’m going to assume you know the basics of YC; if not, start here and also read this. And bear in mind this is just one person’s view.
tl;dr - Think of YC like taking your non-profit through a top university like MIT. You get direct instruction, a change in attitude by being surrounded by smart, ambitious and hard-working people, and a credential that opens doors. The disanalogy, for non-profits, is that YC pays you to attend.
For-profit startups that get into YC give 7% of their equity for $120,000. As a non-profit you instead get a $100,000 grant. (I like to say that’s $100k for 0% of equity, which makes your valuation infinitely high.)
You are expected to move to Mountain View or nearby for the three-month period, and it’s advisable to stay for at least a month afterwards if you’re fundraising. Depending on where you’re coming from, this is expensive (because of travel, rent and office supplies): for us, coming from Oxford, it cost us something like an additional $10,000 per founder (though some founders had significant others which increased the relocation cost).
There are usually 3 or 4 non-profits in the batch, out of about 120 startups in total.
You’re generally treated just like all the other startups – you have group office hours with other startups, and are assigned two YC partners to mentor you. But there is also a part-time partner, Robby Walker, dedicated to non-profits. He’ll look out for you and advise you, in addition to your regular advisors. Occasionally there are non-profit only events. This means that, on average, non-profits probably get more attention from partners than for-profits.
First, you don’t give away any of your company, you just get a grant.
Second, even though applying takes a reasonable amount of time, especially if you get an interview, the process of applying is a valuable process in itself.
For both 80,000 Hours and the Centre for Effective Altruism, the application took about 20 person-hours of quality work (including getting data on things we were lacking data on; getting feedback and revising) and 12 person-hours (across 4 founders) of fun work where we wrote and created the video. Preparing for the interview took about three full days of the whole founding team working together. Nerves for the interview, travel and celebration added another day per founder to that, and in each case the CEO spent a bit more time on preparation in addition to the founders. For four founders, that comes to about 170 person-hours.
This seems like a lot. But even if we hadn’t gotten in I think it would have been a good use of time. Much of the benefit of YC is in taking a set of general heuristics and making them part of your intuitive reflex knowledge; writing the application and preparing for the interview is a major first step on that journey. You are forced to state the mission of your organization in a simple, ultra-clear and compelling way. This helps you refine the mission of the company, and helps you learn how to sell that mission. With a non-profit just as much as with a for-profit, a significant part of a founder’s life is selling the vision and mission of the company. The application process also helps you get on top of your numbers. The questions they ask are carefully chosen to get the clearest insight into your organization. If you can’t answer them well, backed up by data, you probably don’t have good insight into your own organization.
A particularly useful point of learning for non-profits is being able to tell at least some plausible story about how your organization could grow to 1000x its current size: how are you building a recursively self-improving engine, where growth fuels greater growth?fn-1
Similarly, preparing for the interview is a pretty good use of time even if you don’t get in. For a start, it’s a fun unifying experience for the founders. Having agreed-upon answers to fifty or so questions helps your founding team sync up on core aspects of the organization and your strategy. Practicing responding to questions in under 15 seconds (through mock interviews and the iPaulGraham app) and getting feedback on the quality of those answers gets you far better at presenting your core ideas with absolute clarity. It helps you craft founding stories that will be invaluable for setting organizational culture when the team is too big for everyone to know each other.
Moving to Mountain View (or nearby) and settling in does use up a bunch of time, probably about 4 days for each person who moves. But by that point the benefits of going through YC should clearly outweigh these costs. And if you’re an ambitious non-profit startup, you might well want to consider moving the Bay anyway; if so then YC gives you a chance to see whether it’s a good fit for you and your team.
Through group office hours and one-on-ones, the YC partners will help you figure out what metrics to choose, how to get growth on those metrics, and help you solve problems you’re facing along the way.
Both for-profit and non-profits benefit from this advice. But there are many ways in which I think it’s even more useful for non-profits (and a couple of ways in which it’s less useful that I’ll cover later). As a non-profit, it’s hard to get advice from people who’ve help build large fast-growing organisations, because the ecosystem you live in is generally very different from that of for-profit startups. And as a non-profit, you don’t have the carrot and stick of profit and loss to guide your decisions.
The mentorship from the partners might be a little different than you expect. You should imagine incredibly smart, quick, and experienced people who initially know almost nothing about your company. They are armed with a set of heuristics for startup success, and they use these heuristics like hammers to hit your idea to see how robust it is.fn-2
The experience can be disorienting. They will start attacking your ideas way before you think you’ve had a chance to explain them: in part that’s how they learn about the organization; in part it’s an incentive to get really good at conveying your core message as succinctly as possible.
Examples of heuristics include: “talk to your users to figure out how to build your product”; “launch early”; “pick a single metric to optimize for“; “use revenue as a metric if you can”; “aim for 10% weekly growth in that metric”; “do whatever will get you weekly growth; don’t think too hard about the long-term vision”; and “avoid hiring”.
You might have never thought in terms of these heuristics before. The heuristics aren’t always accurate; they are often most applicable to for-profit consumer tech companies, so the further away you are from that the less the standard YC mantras apply to you; and even then, these heuristics are only going to be accurate 70% of the time.
Even though the heuristics aren’t always accurate, they are almost always provocative and interesting. You should push back when the heuristic doesn’t make any sense in the particular case. They know they’re hitting you with heuristics and will adapt their advice, picking a different hammer.
You can get these heuristics from the how to start a startup lectures, the startup playbook, and so on. But learning how to grow a startup is much more like learning how to play Go, or how to write well, or how to play music. There’s only so much you can learn from a textbook; the much more important thing is absorbing a general mindset and attitude, and for that you need a tutor.
For example, you are probably not being ambitious or action-oriented enough. The best founders have basically no gap between thought and action; for them, waiting for a plan to become action is like listening to Muzak on the phone while waiting for customer service — it’s unbearable. The correct attitude is to think, “What’s the goal? And what’s the absolute quickest hackiest way I can get there?”. The advice “be ambitious and action-oriented” is not new; the reason you aren’t currently being ambitious enough is not because you haven’t had the thought to be more ambitious. It’s exactly the sort of thing that can’t be learned by reading it in a book. You need to have someone scary in front of you scolding you for not being action-oriented enough, and extolling you when you try something really big, hairy, audacious and hacky.
Because the skill of growing startups is practical rather than theoretical, you learn as much or more from your peers and from the general atmosphere of YC as you do from the partners. If you’re surrounded by people acting in particular way, you’re going to start acting in the same way.
I think that this is a particularly important reason for non-profits to go through YC.
Currently, startups attract many of the smartest and most ambitious people in the world. Non-profits don’t. Many of the most ambitious people in the world think that non-profits are pretty crappy, and they’re pretty justified in thinking that because most non-profits are pretty crappy. This means that the peer group you get as a non-profit is not the most motivating.
80,000 Hours and the Centre for Effective Altruism are doing well by non-profit standards. If we surrounded ourselves only with non-profit people, there would be a temptation to coast. But at group office hours, I’d often feel like one of the least competent people in the room; and the organization I was helping run was doing merely fine by YC startup standards. This pushed me to work even harder and optimize my time even further.
Because the third sector doesn’t currently have a great reputation for competence, it can be hard for non-profits to recruit the most ambitious and talented people, who want to be around other ambitious and talented people.
By having the Y Combinator stamp of approval, suddenly you’ve moved yourself into a different reference class. There’s a clear signal that you’re an exciting organization with big aims, and are therefore much more interesting for other ambitious people. For example, Jesse Avshamolov was Director of Growth at Teespring (YC W13), found out about 80,000 Hours through the YC network, and then left to work for 80,000 Hours.
80,000 Hours found that the biggest credential-related impact was via hiring. But it wouldn’t surprise me if the brand helped other non-profits a lot in other ways, too.
I believe that YC is a great fit for many non-profits. But the non-profit track at YC is still small and embryonic. So it’s worth bearing in mind the ways in which non-profits can have a harder time at YC than for-profits.
YC’s program is geared towards Demo Day, a two-day event where all the startups have just two minutes and 30 seconds to pitch their company to investors and then set up one-on-one longer meetings. For the for-profits, it’s a goldmine.
As a non-profit, you might find yourself feeling bad about yourself on Demo Day. Many for-profits will raise several million dollars of investment in just a few days. This almost certainly won’t happen as a non-profit. Some YC non-profits have managed to fundraise in the millions of dollars. But, when this happens, it typically takes a long time, such as several months of full-time work from the CEO. Most non-profits currently raise very little directly from Demo Day.
Many non-profits choose not to fundraise at all at Demo Day, instead pitching some other aspect of their product (for example, New Story pitched funding the construction of homes, rather than funding New Story itself). 80,000 Hours raised a little bit of funding, but didn’t prioritise it after Demo Day because the time investment to raise a dollar of funding from YC introductions seemed much greater than from other funding avenues we had.
At the moment, the YC funding infrastructure for non-profits still isn’t that well developed and so you shouldn’t think of YC as a gateway to a ton of funding. Some people at Demo Day are there only to look at non-profits. But it’s certainly not a replacement for a sustainable funding model.
An alternative potential source of funding is the effective altruism ecosystem (including GiveWell and Open Philanthropy and our Effective Altruism Funds). The effective altruism community is full of people who are really serious about donating, have comparatively clear standards about what sorts of projects they are interested in, and are acutely aware of the opportunity cost of time that non-profits spend fundraising. For example, once I introduced Founders Pledge to the effective altruism community (because it seemed to me like a very promising organization), they quickly raised $1.5 million, using what I’d estimate was no more than a few days of their time.
There is starting to be significant overlap between YC non-profits and the effective altruism community: as well as 80,000 Hours and the Centre for Effective Altruism, YC non-profits include New Incentives and No Lean Season, both of which are ultimately gunning for a GiveWell recommendation. This synergy is exciting and, if this trend continues, the YC non-profit funding problem might be solved.
My model of YC advice is as follows: the founding ideas of the YC worldview were built with for-profit consumer tech companies in mind. YC has broadened since then (e.g. to hardware, enterprise companies, non-profits), and they do adapt their advice, but the advice comes with the legacy of early YC, and many of the partners built their careers in consumer tech. So the further away you are from a for-profit consumer tech company, the more you might feel out of place, and the more you’ll need to scrutinise and interpret the advice for your own situation.
This means that an organization can be great but not a great fit for YC. If you’re creating a policy think-tank or a basic science research organization, you can potentially do a huge amount of good but you probably won’t benefit from going through YC (though you never know). And it means that some of the YC advice just won’t fit you as a non-profit.
For example, the partners give a lot of advice about how to present your company to investors. (The YC partners’ ability to hear a few garbled sentences about a company, then immediately restate a one-sentence description of the company in an utterly clear and compelling way borders on magical.) But the advice doesn’t straightforwardly transpose to non-profit fundraising. I’m not sure how to give better advice: the problem is that the reasons people donate to a charity are so varied compared to the reasons major investors invest in companies. For effective altruist donors you (generally) have to have compelling numbers or arguments and demonstrate high standards of honesty and transparency; for foundations you have to show how you’re a really good fit for their grant-making aims; for some donors you just have to pique their interest in some way or another; for many donors you have to build up an ongoing relationship with them. And, if you’ve got a sustainable funding model, you’re much more likely to be doing the equivalent of bootstrapping — primarily raising funds from people who use your product in some way.
A second way to be careful with the YC advice is with choice of metric. YC is well set-up for helping startups optimize towards a metric. For for-profit startups, they’ve also got good answers to the “what metric?” question: use revenue unless it’s really not possible; in which case use whatever metric is the closest proxy for revenue.
For non-profits, the choice of metric is much harder. Ideally, the metric for all non-profits would be the wellbeing-adjusted life-year (approximately: how many people you benefit multiplied by how much you benefit them, over the long-term, and incorporating some organization-specific value judgments about, say, animal welfare and population ethics, or perhaps non-welfare values); that’s the equivalent of revenue. But, unlike revenue, this metric can be very hard to track, especially once you consider long-run effects of your actions. And the partners just don’t have expertise on the issue of how different proxies (such as bednets or deworming tablets distributed) translate into wellbeing-adjusted life-years. In the non-profit world, it can therefore be easy to optimize for a metric (say, number of textbooks provided, or microloans provided) that doesn’t correlate at all with improvements in people’s lives.
Again, this is an area where the potential synergy between effective altruism and the YC non-profit program is exciting. Effective altruism is heavily concerned with figuring out what activities convert best into increases in wellbeing; the combination of the effective altruist approach to choosing a target, combined with YC’s power at helping organisations optimize towards that target, is potentially very powerful.
(And, having said that, YC did have a big impact on 80,000 Hours regarding choice of metric: the heuristic “use whatever metric is closest to what you actually care about” was an important one for us. 80,000 Hours moved from active newsletter subs that graduated from target universities (a leading metric at best) to measuring impact-adjusted significant plan changes, and as a result was able to make much better-informed decisions about what to focus on.)
If you’re going to apply, you should really learn what they’re looking for. Easiest is to look at examples: Here’s 80,000 Hours’ application and here’s the Centre for Effective Altruism’s application. You should also read all the relevant advice on applying, starting here. And you should note their request for startups, which include some areas specifically noting non-profits.
I’ll add to that a few things I think are particularly important.
First, prioritize clarity and concision. People often associate startups with jargon: “We’re a platform for leveraging disruptive market penetration in the early adopter space”. Non-profits have their own set of contentless jargon, like “engage with stakeholders”. Founders suffer badly from the curse of knowledge: they’re so invested in their own company that it’s difficult to imagine the mental state of someone who knows nothing about their product. You need to cut through this: write the most concise, no-bullshit description of what you do that you possibly can. If it can’t be understood by a twelve year old, you probably need to revise it.
(A caveat: of course, a lot of jargon is actually useful. “Users”, “churn rate” and “freemium” are all useful bits of technical terminology (even if “freemium” is a linguistic abomination); the same is true for lots of technical software terms. If you’re really confident that the jargon is helpful you can keep it in, but otherwise avoid.)
Relatedly, the most important thing to convey is what you do in concrete mechanical terms; you should clearly state this in response to the first question. Do not “start with why”. Consider telling a simple story of how a user (for non-profits, this might be a beneficiary, a donor, or both) interacts with your product or organization. This is a special risk for non-profits: for a lot of fundraising, it’s sufficient to explain the problem (malaria!) and just tell people you’re offering a solution (bednets!) and people will trust that you’re pursuing a sensible solution. That doesn’t cut it here.
Writing in this way might seem unnatural. My advice is to first adopt the YC style and manner of thinking even without understanding the reasons behind it; then having internalized it you’ll later understand why that style and manner of thinking is good; then once you’ve really learned to see through and understand the YC lens, you’ll get a feel for the situation when that lens isn’t the right one for your problem. In jazz, there’s a motto: “emulate, assimilate, innovate”. It’s true here as well.
Second, if you’ve already launched, describe your growth. Similarly, when you describe anything about the organization that can be measured, explain through numbers.
Third, as mentioned before, try to show how you have, or might have, a sustainable funding model. The way that YC became so successful is that they had an early big success (Reddit) that drew attention to them leading to a massive success (Dropbox), that drew even more attention, leading to a gargantuan success (Airbnb). They’re looking to try to fund a similar mega-success in the non-profit world, in order to make non-profit startups a more attractive path for the next generation of founders. And for that you either need a sustainable funding model or some explanation of how you can massively scale.
Finally, put effort into the video. Many partners use the video as a first screen. The on-line advice might suggest that you should just press record and say what comes into your mind; I didn’t do that, and I wouldn’t recommend anyone else should either. For both 80,000 Hours and the Centre for Effective Altruism, we figured out what the most important things we wanted to convey were (which will always include describing what you do), wrote a script, practiced it a ton of times (modifying it a lot as we practiced to sound natural and compelling); and then shot it dozens of times. You want to avoid seeming scripted; but the way to do that is to do learn your lines and do twenty takes.
One question you’ll have to address is who you list as founders. In the YC-world, ‘cofounder’ is a technical term that doesn’t imply the person literally founded the company; in the Centre for Effective Altruism’s case, only I am literally a cofounder; Tara, Kerry and Sam joined later. In the for-profit case, ‘founder’ means “person who’s running the organisation at the time of YC, who owns at least 10% of equity, who’s recognized as a founder by YC.” The 10% of equity requirement obviously doesn’t apply to non-profits, which makes the issue of who you list as cofounders trickier.
You should definitely have more than one founder (if there is one), and you shouldn’t go above 4 founders. You definitely shouldn’t include anyone unless they’re already spending all their time at the organization, you’ve known them a long time and you can imagine them running a large part of the organization when you’re much larger. But within those bounds, my advice would be to err on the side of including founders. In both the Centre for Effective Altruism and 80,000 Hours’ case, we applied with four founders. Because the benefits of the advice and mentorship are such a big part of what makes YC valuable for non-profits, having more founders means more total surface area between your organization and YC. But do bear in mind that, if you have employees, YC will create a greater divide between the people listed as founders and the non-founder employees, because most YC events are founder-only. It’s also true that having two very impressive founders and two merely okay founders probably makes your application look worse.
There are a couple of bad reasons for not applying I should mention. First is that you’re non-technical. Now, it’s true that having a technical cofounder is going to help your application, as is using tech as part of your mission. In part, that’s because YC is only looking to take on organisations that they think they can add value to, and they know they can add value to organisations that have a technical element to their product. And I think that very many non-profits should consider having a technical cofounder, simply because it so often allows you far greater productivity and scale. But having a technical cofounder isn’t an absolute necessity: The Muse got into YC without a technical cofounder, as did the non-profit InnerSpace.
The second bad reason is being too late-stage. Because you’re not giving up equity, you can apply later stage as a non-profit than you can as a for-profit. There are examples of later stage companies going through YC — Quora went through when it was valued at $900 million — but they’re very rare, and the deal for for-profits is almost always $120k for 7% of equity. But YC welcomes later-stage non-profits: depending on how you measure it, the Centre for Effective Altruism is 7 years old.
Apply. Do it.
For-profits build something people want, therefore they make money (or could make money if they wanted), which they can then reinvest in building and selling their product. The feedback from success to growth is very close, and where it isn’t close (you’ve got a product that will take a while before it generates revenue) there is an ecosystem of investors to close the gap. For non-profits, nothing like this exists, so it’s super important to have at least a tentative vision about how you could turn success at your mission into increased funding, which you can then turn into greater success at your mission.