Kevin Hale and Adora Cheung - Startup School 2019 by the Numbers

Hello Berlin! We've made it to week ten. This is the last stop on the startup school tour. We have something kind of special today. So Adora and I are doing a presentation together, and it's all based on numbers from the last nine weeks of startup school. She's going to sort of kick it off, and what we're going to do is probably talk, like hopefully twenty minutes, and then we'll open it up to Q&A, taking it to all of your questions. There’s anything else you want to talk about, and then we'll have like more mingling and eating and such until though we have to kick you out and go have you work on your company or apply to YC, that's what you still have left to do at 8:00 p.m. Fedora, 5:00 a.m. or Linn time, 5:00 a.m. Berlin times. So funny!

All right, thank you, Kevin, thanks for being here. So Kevin said my name is Vidura, one of the partners at YC, and we're just gonna go over what we learned during startup school. So it's week ten; we made it! Congratulations! The startup school team has been, in the past few weeks, traveling around the world visiting cities where there’s lots of startup school founders. We’ve done 20 events at this point, over 15 cities, and we’ve interacted with over 2,000 founders in person.

So similar to how we started, we wanted to end startup school by looking at the numbers. Most of you don't know this, but startup school actually grew a lot during the course itself. When we held orientation back in July, at the end of July, we were around 20,000 founders and 21,000 startups with over 20,000 people auditing the course just for the content. And as we let more people in during the course, as of today, the number swells by over 40%. So you'll see we have now over 40,000 founders in startup school and over 30,000 startups. There are over 77,000 people actually subscribed to the content.

Oh yeah, what's with the number of people increasing between 30 and 50 percent? Really cool that female founders increased the most here. So, one of the things that we wanted to look at was whether startup school had made a difference. So we looked at the last two batches to see how last year's participants compared to the rest of the population in regards to their ability to get funding from YC. So in the end, from the past two batches, 9.7 percent of startup school applicants that applied got an interview versus 9.1 percent of non-startup school applicants got interviewed. That means 6.6 percent— you did startup school; 6.6 percent you had a higher chance of getting interviewed, and we think this year it will be much, much higher because we focused the course a lot on helping you build that application and just feel start up on the right things to get those questions done well.

All right, so if you look deeper into the numbers, as of today, 106 startup school companies have been accepted into YC. For a group of people who are more likely to get interviewed, you'd think they look amazing on paper. They got until YC, but you would be wrong; startup school founders accepted into YC are surprisingly different in every way. In every city we visited, founders are constantly asking us to see if YC still wants to accept companies that are not perfect, that are like them, but not perfect. The big takeaway is that YC is not afraid of warts, and we can assure you no startup we've ever accepted has been perfect.

So, dig a little bit deeper. In 206 companies from startup school that have gone to YC already, 40 of them had no traction when they got in. So that means no revenue, no users yet. 57 also had no revenue, and 50 had no funding whatsoever, meaning zero, zero dollars cash in the bank. Some of them didn't even incorporate yet; quite a few of them were still working at jobs or might have been in college and still working part-time. 40 of them applied previously. So this is one of the things we encourage—you—even if you don't get in, to keep trying again. A lot of people apply three to four times; I’ve been on that for that—it's try—they get in. Some of the companies we didn’t really like the idea they applied with, and they actually pivoted during the batch, and so that counts for a dozen of them. And a lot of them applied late, so by the time you watch this video, you're applying late, and we still encourage you to apply.

And finally, this is the most asked question in our tour, which is, if I'm a solo founder, should I still—can I still do YC? And the question is absolutely yes! Ten single founders got in from startup school. We average around 10 percent of the batches around solo founders now, if not a little bit higher. We highly encourage every team to have a technical co-founder, but if you don't, it does not disqualify you. In fact, three of the startup school companies that have gone into YC are not technical.

All right, so how can you replicate their success? It's very simple, and I’m about to repeat a lot of things we’ve already said during startup school, but the best thing you can do is when you talk about your startup, there’s a question on the applications: What are you making? And that is to be very clear on what it is. And if you’ve watched Kevin’s lecture from a couple weeks ago, the next few slides will seem very familiar.

So one is be clear on what you’re working on. Know jargon; not only be clear, but don’t use all the words in the dictionary. Be very concise in how you get your point across. So to be even more specific on that question on the application of what are you working on, startups that got accepted to YC use, on average, 78 words to describe what they do. So 78 words might seem a little or a lot to you, but in fact what it does is it looks like this: it’s two paragraphs and gets your point across well. So if you have something that looks longer than this, then you should really go through your description and start cutting words and finding sentences that you really don’t need to have them there.

All right, so if this is what a description should look like, what should a description not look like? Well, here’s an example of this—this is literally one of my favorite ones because as you’ll see next of a description that did not get accepted or get an interview. So we couldn’t fit the whole thing in here, but if you read like this guy—or I forgot who the founder is, but he basically knows he needs to write something short but decides to write a wall of text. In fact, that wall of text was the longest description we got of all applications in the past two batches; it was over 2,000 words! Obviously, please don’t do this; it doesn’t help us understand your startup any better; it actually just confuses more and frustrates us. It basically tells us that you don’t know how to be efficient when it comes to getting people to understand what your startup does.

So if you think about it, if this is how long it takes you to describe what you do, if you can imagine, when you come in for a 10-minute interview, the interview will be done before you even finish your description. Furthermore, what culminates at the end of YC itself is a two-minute demo day pitch. So again, that’s even less time to get your point across, and so you really need to start practicing now on how you deliver your—how you explain your company.

The good news is, in startup school, when you apply or when you've registered, you have to fill out your company description, and we actually forced you to write it in less than 200 characters. Now, this is a histogram of a number of characters of words that startup school companies use, and you can see that a lot of you got through the 200th character and probably wanted to write more but had to stop because we forced you to stop it. And the result is this description actually averages around 40 words or less, so by force we’ve made you do it in less than an average of 78 words. So that’s good news! So when you transfer your descriptions to your YC application, we suggest you to not expand much further than the number of words you’ve already used.

All right, so one thing we noticed and we’re worried about was the lack of editing on your hand when you’ve entered in that startup school description. So over the course of the 10-week course, we’ve been trying to help you make that description a little bit better; in fact, a lot of the group sessions were designed to help you with that. But as you've been practicing pitching every week, many of you didn’t actually go back into your profile and edit your company profiles with a better description. It’s actually the one thing we do at YC to help startups more than anything else: really get their one or two-liner descriptions.

So the next article we’re going to probably focus more on helping you do that and forcing you to document your changes over time. The median number of edits is only one, so we think that most likely the description you have right now can be improved on drastically. So if you need more tips on how to do this well, Kevin's last lecture, how to evaluate ideas part two, that’s the name on the YouTube channel, please please go watch, as you see a lot of the slides were lifted from there.

So all right, I’m going to let Kevin take over for the next bit. You guys submitted a lot of data to us. Every week, we ask companies to let us know about their progress. If you've got an email from the Kevin inspirational bot, then you know that basically we’re trying to get you to just spend 10 to 15 minutes at the most, so just tell us like what's been going on with your company. And in the last nine weeks, that’s been over 124,000 pieces of data, and so we’ve got a lot of interesting insights from that.

The first thing is concerning what startup school companies considered their biggest obstacle. So, not surprising to us, but it is kind of worrisome. Number one was funding! We were worried about lack of resources and you could move forward as a result of not having money. And the other one that’s interesting on here is like meeting a co-founder at number six. The reason I would kind of point these out is that if you look at what people said actually improved their KPI, the metric that they considered mattered the most when we did the text analysis, it looks very different. So you think your biggest problem is one thing, but the thing that would actually help you the most are very different.

So, number one, of course, is launching, getting out the door. Once you launch, you can move that KPI, which is wonderful. Other stuff on here, a lot of it concerning product. The thing is dealing with the users. If you're having problems where the difference between what you think your biggest obstacle is and that makes the biggest difference to KPI are two different lists like this, watch Adora's lecture. She talks about how to prioritize your time. It’s actually doing really, really well on YouTube; it’s the fastest spreading one. A lot of people have found her tips in there on how to think about your to-do list in a much more KPI-driven way—very, very helpful.

We asked you to talk to your users and tell us what did you learn from them, and these are the top things that people got. Not surprisingly, they're focused on product, and basically the best thing you can do is make something people want. The top two insights because they're focused on conversion and modernization—I've got two lectures that we did back-to-back on improving landing page conversion and pricing, so those are good places to start if you want to improve things for your users.

Now actually, most of the participants, unfortunately, in startup school did not launch. And when you were considered not launched, the KPI that you tracked was weeks to launch. And so these are the numbers and how they came out: people on average went from WTF weeks to launch of seven weeks down to five weeks. For some reason, that should have been zero! Right?

Now, what's actually super interesting here is that the median weeks to launch is about four. So that indicates that on average it’s being actually very much impacted by people having very large weeks to launch—that's really bad. If you haven't already, you're gonna find a theme here for this presentation: watch Michael's lecture on how to plan an MVP. He talks about how you should only be making the first version that takes just a few weeks to launch.

And so even that median is double the number that we recommend for our companies. In that talk, he also coined a new term that we've been using: it's Herb's going out YC—the heavy MPP. If you're a hard tech or biotech company, you might need to build something that's a little more substantial. And in fact, you guys actually believe you do need to; you're almost double the number of weeks to launch that you estimate.

And unfortunately, if you look at the graph by comparison—so this is the average startup school weeks to launch that’s going slightly down over time. This is the biotech or tech people; you guys actually go up over time. So now add features over the course of the week, and that’s not good, because actually from our experience funding over 200 biotech or tech companies, they should be following different advice or rules than other companies, which is why we recommend, if you haven’t watched it, watching Jared's lecture on advice for biotech in hard-tech companies. He talks about seven wonderful examples of YC companies that built a much lighter weight heavy MVP and creative ways of getting their company off the ground even though they don’t have millions of dollars and lots of time to build the first version.

For the companies that did launch, you guys did great! So this number represents the percentage of the weekly updates that were submitted that came from companies that launched. So at the beginning of the course, a third of the weekly updates came from launch companies, and by the end, half of them were from launched companies. That’s super awesome! In fact, 2,253 startups—we can actually track switched from weeks to launch to a different KPI during startup school. To put this into perspective for us, it took over 10 years for Y Combinator to launch that many companies through our program, and in short, while we did it in 10 weeks.

Adora talked about group sessions. All right, as you know, every week during startup school, we matched you with other startups to help you practice talking about your own startups. And we rated everybody and would rate each other on three metrics: Did you understand this idea? Were you excited about the idea? And would you want to work with the founders? For the first question, over time you guys got better at explaining. As you can see, the differential from each week to week was over 0.1%. To put that in perspective, a 0.1% increase represents almost 100 startups because of the number of startups in startup school, and by the end there was a 7.1% increase overall.

In terms of improving on selling yourselves, you guys also got better over time. There was a 9.5% improvement from the beginning. The biggest improvement was actually getting other people excited about your idea, and so there was a 17% improvement since the beginning. We know that there’s a lot more work that needs to be done here. We want to, in the future, create a lecture to help you get better at getting other people emotionally excited about your ideas. Being able to do that helps you with things like recruiting and getting customers on board and also investors on board in the future. If you want more help in structuring your investor pitch, Kevin’s lecture for a second lecture—it was called how to evaluate startup via ideas part one—I would go watch that!

All right, so while the quantitative ratings made it easier to get founders involved, startup school founders loved it when you actually gave each other qualitative feedback comments after the group session. We were amazed to see how supportive everyone was with each other in terms of giving this feedback. In fact, over 93,000 pieces of qualitative feedback were created during the course. So many of you have offered to help with introductions to customers, often whispered words of encouragement that we know made a huge difference to someone struggling to make the magic happen. One of the main reasons for doing these group sessions together is sort of like emotional support in group therapy because we’re all going through the same things.

And so I think that achieved the results we were looking for in that sense. The generosity we saw in the comments is what makes doing a startup in a batch incredibly powerful. So thank you for everyone for taking the time to do this; it really means a lot to everyone. On the flip side, not all of you are great; a bunch of you no-showed. So 19% of startups who requested a participating group session did not show up—boo! Yes! So this graph shows updates, requests for group sessions, and attendance and no-shows over the last nine weeks. Oddly enough, 23% of startups that submitted a weekly update didn’t even request a group session.

We’re looking for the next course of how to make this easier and more convenient for startups because we think that this has been actually helpful not just in improving your pitch but also for moral reasons. Interviewed interestingly enough, this graph actually shows where we messed up in the first week. So that big gap there between the number of updates submitted and requested shows that you guys were very confused about how to request group sessions at the very start of startup school, and that’s why we gave it to you one for free on that first one. And that sort of stabilizes as we get down.

You're never just like how you guys are probably using metrics to help improve your startup—we are, too, as well! Okay, so we also asked you guys to submit how you’re feeling—what's your morale on working on your startup, and overall out of 124,000 pieces of weekly feedback. And you can see the trend is you got a little happier over time. But we actually want to just segment this out to see what actually impacted, what made a big difference for morale for companies, and so it led to some of the most interesting and unique insights from the program for us at Y Combinator.

First one: ethical startups, no surprise, are happier. You’re just able to get more done. One thing to notice is you’ll think that the difference here and changes are small, but because of the amount of data that we’re collecting across each one of them, that these are pretty statistically significant. And so if you're looking to be quantitatively happier, these are ways to do it: If you launch, you’re a little bit happier than the unlaunched.

One thing that helped a lot of companies to rethink how they thought about launching was watching Cat’s check on how to launch again and again. It just really changes that bar from being a launch that requires press to being one that is just about shipping and deploying on a regular basis. If you're full-time on your startup, you feel a lot happier, obviously, probably because you can also make more progress. If you're making money, you’re happier. The whole algorithm, “Money can’t buy you happiness”—not true. Adora’s got a good lecture on this on focusing on revenue and how to set KPIs. It’s her first lecture from the startup school, and I actually think it’s one of the unappreciated ones. It’s the lecture that really talks about why 99% of you should be focused on revenue, and be honest with yourself about whether you’re actually doing that or not.

Here’s one that’s really surprising: people who did not flake on group sessions were way happier than people who did. Basically, feeling like you're doing a startup and having other people to talk to you and being part of the community and batch meant that you're happier working in your startup than those who did not participate at all. The happiest cohort was actually people who were in teams, and this is actually one of the main reasons we always say that we want companies that try to have co-founders. It is so hard to do a startup on your own, and this is usually morale is the thing that kills a lot of companies and founders.

So again, when we talk about we want you to try to have a co-founder, it is to help with the longevity of your startup, not because we just don’t want to select those that don’t have teams. And then this was the most surprising thing that we learned from startup school: that founders in teams that talked to users were way happier over time than those who didn’t. And what's interesting about these two cohorts is that the people who did not talk to any users at all—they were the only cohort that actually had morale get worse over the course of school, in the nine weeks that those who didn’t.

So the lesson, the moral of this morale: are you feeling sad? Talk to users! It doesn’t take very much to do it. The median number of users that startup school teams talked to was just three users a week; it actually provides a ton of insight for you, but apparently, it’s better than Prozac! If you want more information about how to talk to users and a better way of handling that in a sort of structured way that helps you improve your product, watch Eric’s lecture from the first week of startup school.

Let’s bring it home, Adora! In summary from this presentation, startup school helps you get interviewed. It's okay not to be perfect—like you can get accepted even if you have a wart. Try to write your description in 78 words or less. Prioritize launching; you’re probably taking too long to launch. Practice makes you better at pitching, talking about getting people to understand you, get excited, and want to work with you. And lastly, talking to users makes you happy, which is why we’re going to switch over to talk to you guys. Thank you so much! [Applause]