Going Through Y Combinator (S13): Nine Lessons Learned

Last year, I shared our story of applying, our application, and some advice on the YC application. Now that the Winter batch is in full swing and the application is open for the summer, I thought I would share our story of going through YC and nine lessons I learned in the process.

I have had a lot of trouble writing a post about going through YC. It was one of the best and most productive experiences in my life; it was also one of the most difficult and stressful experiences in my life. As I have attempted to write, I've faced some difficulty in avoiding both what could be a boring recitation of facts that are widely known and a sensationalized portrayal of the experience.

YC left me more scared and yet with more optimism and excitement than any other three-month period I have lived through. YC is hard. You're stressed out. You're underslept. You think your company is on the verge of collapse (and the companies of some of the people around you are). The partners' advice is brutal — brutally honest and unvarnished in its delivery. YC is relentless. Difficult. Stressful. Thrilling. Unforgettable. Worthwhile.

The best metaphor for YC is a slow-cooker. Every week YC cooks a meal for fifty companies (now seventy) in a bunch of slow-cookers. That's all it takes: time, heat, and a pressing deadline for delivery. Sometimes you throw all the ingredients in and get a good dinner or a good company. Sometimes it's crap. No one seems quite sure beforehand which it is going to be.

If I ever found another company I would do YC again. Well, if they’d have me.

Lesson 1: Do YC if you can. It's worth it, but it isn't easy. (I promise the other lessons are more surprising!)

But let’s go back to the start.

If you get into YC, they call you the same night as the interview. The phone call is brief: do you want to accept the offer to join YC, and can you do an introductory talk on Wednesday in Mountain View? We said, "Yes".


The orientation involved two parts: one where Paul Graham (PG) talks and one where Kirsty talks. This may change now that Sam Altman (Sama) is YC's president, but the gist will likely be the same.

PG condenses the most important lessons of YC into about two hours. It's a dizzying talk, even though most of what he says are in his essays. The most important piece of advice is to "start now". The day you get into YC you can start going to office hours, and you're told when Demo Day is. In effect, YC is on, even if the dinners don't start for a few weeks. Get working. There are a fixed number of days between acceptance day and Demo Day. Anything you do before Demo Day you can use to impress investors on Demo Day. Anything you do after Demo Day still matters to investors, but it likely won't matter until much later.

Kirsty Nathoo spends the second half of the day describing the basics of corporate setup. She goes over all of the associated paperwork, the YC investment, the YCVC notes, and advice on what to spend money on. It was a fast, expert primer on corporate law, corporate finance, startup financing, benefits, payroll, and more.

Part of being a startup founder is learning tons of stuff which you might not be comfortable starting with. That's what's great and what's challenging. One thing we learned early on is to embrace all the learning: being a startup founder is about constantly doing new things outside your comfort zone.

Lesson 2: Be comfortable doing anything. Learning anything. Working on anything. Most parts of founding a startup are unglamorous. Like talking to your lawyer about how to get the best strike price on your options for your employees. Or making sure that payroll taxes get paid. But that's what a great company is: a thousand carefully, but quickly, made decisions.

Moving to Mountain View

YC recommends that you live close to YC for the duration of the batch. They say that this is because they’ve found a high correlation between startups that attend YC events and those that succeed.

It’s also the case that there is a lot less to do in the Valley than in San Francisco if you’re young and childless. You will work more. Part of PG’s advice is to do little other than code, talk to users, sleep, eat, and exercise. That advice is much simpler to follow when you’re living in Mountain View; there is only so much you can do on Castro Street.

Lesson 3: Avoid distractions when starting a new company. Find a way to build and do little else. Go to the wilderness — by which I mean Mountain View — if you have to.

Dan and I lived with Kai and Claire from True Link Financial in a small house that has passed from YC company to YC company for a few years. Claire and Kai have become some of our closest friends, more broadly. They’re both amazing people, dear friends, and quite well-rounded in their interests. They also became reviewers and supporters of our work, with a much deeper understanding of working in fintech than most of our other friends.

Living with another team with whom we grew so close was critical for us. Otherwise, I think we would have torn ourselves apart. I'm not sure it’s necessary for everyone, but for two guys from New Jersey anything else would have led to a steel-cage death match.

Dan and I spent a huge amount of time together over the summer, but one rule we had is that we spent Saturday apart. We needed that time.

Lesson 4: Don't give up all human contact — just most. Do what it takes not to murder your co-founders in their sleep.

Idea iterating

We spent most of the first few weeks trying to figure out what we were going to build. Whether this is what YC meant by "start now" could be up for debate.

We were trying to figure out — in the abstract — what our business model would be. We spent most of this time just talking to businesses about what they found annoying in dealing with their banks.

In order to think through this, Dan and I walked around the same block in Mountain View several hundred times.

Over the approximately six weeks between orientation day and Prototype Day, we iterated through six ideas in sequence while talking to nearly a hundred potential customers: APIs for commercial banking (what we applied with), WealthFront for businesses, near real-time cross-bank settlement, financial reconciliation support, building a bank from scratch, and API gateway for commercial banking sold to banks. Given his experience and passion, Brent was most interested in the platform ideas.

The details of those business ideas, except the last, aren't so important to the story, but the process was helpful. We were not building anything — which I don't recommend — but we were making use of lots of potential customers to figure out what they would pay for and what their pain points were in the interface between their businesses and their banks.

Almost everyone we spoke with said: (1) we don’t like our bank, (2) we want a technology-forward bank, and (3) we want working with our bank to be as simple as working with new payments companies like Stripe, WePay, Balanced Payments, etc.

And then we just got lucky.

Our product

Starting in October 2012, about six months before we incorporated and eight months before we started YC, Dan and I had been talking to banks about an API for ACH. Our plan was to connect a bunch of banks and then have a cross-bank same-day settlement system. We were talking to J.P. Morgan, Capital One, Wells Fargo, Silicon Valley Bank, City National Bank, Citi, et al.

Then, while we were endlessly flailing around in circles in Mountain View, several of those banks came back to us and said that they wanted an API gateway. Would we sell the experience we were describing to them as a white-labeled or co-branded experience?

We were in the bank software business. Our first product was an API platform-as-a-service for commercial banking.

Lesson 5: Always be willing to iterate and always, always, listen to someone willing to pay you for your product. Paying customers are hard to find and they're often right.

Prototype Day

Prototype Day is one of the best days of YC. It's like a mini Demo Day. The order of the companies is randomly chosen and every company has a few minutes to present. It’s early on and you still don’t know most of your batchmates all that well. Prototype Day lets you learn about what everyone is doing and a little bit about every person in your batch.

The partners tell you not to prepare. However, we found preparation to be focusing, and we chose to put together a presentation. We used that as an exercise to figure out how we would pitch our idea and company to a large group of people.

After each presentation, PG tells the company's founders some of the things they did wrong. It's useful, although it's best not to be randomly chosen to go first or second. It's your first introduction to how investors might react to your pitch.

Prototype Day ends in a vote. Every founder gets to vote for two companies, then the vote is tallied, and the top ten companies are announced. I can't find any of the published results of any past Prototype Day, so I don't want to give any results, but we did fine.


YC hosts a dinner with a prominent startup founder (or Ron Conway) every Tuesday evening. The talks are interesting — more inspiring than informative. A frequent theme is how to not get screwed by your venture capitalists. YC dinners are off-the-record, so I won't share any of the details of the talks we experienced, but they're enjoyable.

The most significant value of the Tuesday dinners is that it's the one time each week when YC turns into a co-working space. Companies arrive hours before dinner. You have many of your office hours for the week, you chat with Kirsty or the Levys (YC's two lawyers) about a random question you have, you catch up with everyone in the batch, and you get as much user feedback as you can get in the hours before dinner starts.

Lesson 6: Set deadlines for yourself. Always have something to fight towards, even if it's not concrete. Measure your progress even if the measurements are artificial.

The dinners are weekly milestones. Most people work their butts off all week culminating in Tuesday. Report back. Rinse. Repeat.

The Batch

One of the most important aspects of YC is that they fund companies in batches. And not just so YC can have dinners with good speakers. Batchmates make fast friends who stick with you long after YC is over. Dan, Brent, and I each invested in several companies in the batch!

You go through an adventure together. Batchmates are always willing to offer help with a product question or support you when you're down (and there will be times when you're down). It's hard to come out of Tuesday dinners without feeling energized because you've gotten that emotional lift (and competitive spirit) from your batchmates.

Office Hours

There are now seven different types of office hours. The regular and group hours are the only ones that happen frequently, so I'll describe them.

In regular office hours you meet with a partner for ten to twenty minutes. You talk about the progress you have made in the last week or so, discuss the particular problems you're facing, and ask any questions you have. In our batch, each company selected a partner or two as its primary partner(s), but now I think they assign one of the full-time partners to each company. We chose Geoff Ralston, who was perfect for our personalities: direct, to the point, and no-BS.

Group office hours happen every two weeks. Our group partners were Paul Buchheit, Kevin Hale, and Kirsty. Group office hours are like regular office hours, except they're public and shared. Each startup takes a turn giving an update and interacting with the partners. This is useful because it allows you to see the problems that others are facing (and also how they solved them!).

Selling Software to Banks: Slow Weekly Progress

Selling big, complex software to big, conservative institutions is hard. It is slow. During our YC batch we almost always thought we were just a couple of weeks away from having a deal: next week we will sign a deal with bank X, we said. We said that every week. We told investors that. We told YC partners that. We told friends that. We believed it. We were wrong.

Now that we have signed proof-of-concept deals that we are only now negotiating into definitive agreements, I can say with confidence that we were naive. We didn't know how enterprise sales worked. Experience has educated us. Harsh experience.

However, I think this made us look bad to YC. In retrospect, we were like boys who cried wolf. PG has said that the best way to tell which startups are going to make it is to look at their weekly progress. We did a lot every week: tons of meetings with banks, product and security reviews, etc. — but I am not sure any of it looked like progress.

Lesson 7: Every YC company (every company!) is a shitshow in one way or another, even if you often think it's just you. There's power in realizing that that's OK. It took many of our batchmates too long.

In December, I voiced my concerns about crying wolf to PG in an email when he asked about our progress (I think he emailed the entire batch). He said:

Don't worry too much. Deals of this sort are always slow.

If I were in your shoes, I'd focus on sales of whatever type I could get. As soon as you reach breakeven, everything changes. Slow customers are no longer fatal once you're profitable; they can decrease your growth rate, but they can't kill you, and if the market is big you'll eventually get big.

Deals of this type are slow. All enterprise sales are slow. YC was amazing, and I tell all my friends to apply, but it is also difficult for companies like ours - companies where a small number of large deals define them. Those types of companies may be on the path to success, but they aren't like consumer companies that can show that desired week-over-week growth.

It is an expectations-versus-reality disconnect. What is possible for other companies was never possible for ours. Yet YC holds up a mold for all companies to fit into. As YC grows and includes more enterprise companies, I'm curious to see how they manage that challenge.

Fundraising Too Early Before Demo Day

YC tells you to avoid doing it in the first two months. We did it too early. It was a mistake. In fact, we screwed up our entire fundraising process. Three friends told us that normally people who mess it up so badly can't recover; we somehow still raised a healthy seed round. I'd say we fucked the whole thing up, except we did end up with millions of dollars. If I ever find the time, I'll write another post on what not to do, but YC is right: don't raise money until you're ready.

I think that the ideal time is likely about two weeks before Demo Day; at that point you may or may not feel ready, but you're going to have to be so it's time to start acting like you are. People ask on Demo Day how much money you've raised, and the answer shouldn't be zero. But if the answer is that you closed the round (or that you got too many "No"s, too early), then that might be just as bad.

Lesson 8: Take advice. Sometimes your friends and advisors are right even if you think they must be wrong. Don't go with the crowd but try to do the right thing for your company at the right time. We didn't have the confidence to refuse introductions and meetings with big VCs until we were ready.

Rehearsal Day

Roughly a week before Demo Day you have a Rehearsal Day where every company goes through their initial pitch. Every company presents, there is a vote at the end, and you get in-stream advice. In fact, this time PG will just interrupt you with advice. This is the beginning of a week long sprint that culminates with Demo Day.

You have to practice this presentation so much that, despite its being scripted to the word, down to the second even, you sound perfectly natural. This proved not to be a strong point of mine. I can give good, if not great, off-the-cuff presentations — in fact, I relieved my stress by giving everyone else's presentations (sometimes in a joking tone) and helping batchmates come up with new phrasing, new ideas, new ordering, etc. But getting down the exact wording that I agreed to with PG? That took me all week.

Demo Day

The night before Demo Day is Alumni Demo Day in the same venue; I did not do well. I felt like I spent every single waking moment between that and our presentation on Demo Day practicing. I walked to the Computer History Museum from our Mountain View house giving the presentation to myself over and over and over again.

I think the first time I gave the presentation perfectly on stage was on Demo Day itself. It was a good time to peak.

When not presenting at Demo Day, you are surrounded by investors. It's ten solid hours of answering questions, discussing valuation, and trying to close some deals.

It was exhausting. It was exhilarating. But that's YC for ya.

Lesson 9: Enjoy yourself. Building a company is exhilarating, even if it's exhausting. You are building something through grit and persistence, so it's best to have some fun while you're doing it.

Thank you

Thank you to those who helped edit and advise me on this post — my co-founders, Brent Goldman and Dan Kimerling; my batchmates Aaron Feuer, John Gedmark, Glen (Professor) Moriarty, Claire McDonnell, Jake Heller, Nathan Wenzel, Patrik Outericky, and Maran Nelson; and the best-damn-proofreading-friend there is, Kate Brockwehl. I am always in debt for her countless suggestions of great value: and for all the present semicolons.