Industry Insight: A Chat With Startup Kingmaker Y Combinator

What do Airbnb, Docker, DoorDash, Dropbox, Reddit, and Stripe all have in common? They all started at Y Combinator (YC). Founded in 2005, YC was born in the midst of a second wave of Silicon Valley innovation built on software and the application economy, helping launch Software-as-a-Service (SaaS) companies from Gusto to PagerDuty. The startup accelerator and venture capital (VC) investment firm has funded more than 1,400 startups, with a combined valuation of more than $86 billion. YC has mapped out a new hybrid model in the process, combining VC seed investment with startup mentoring and incubation, and connecting the companies directly to a network of investors and entrepreneurs.

Y Combinator Partner Tim Brady was the first employee hired by the founders of Yahoo. He spent eight years as the internet pioneer's Chief Product Officer (CPO). He later founded Imagine K12, an education technology (edtech) accelerator that merged with YC in 2016. PCMag sat down with Brady at the Y Combinator office in Mountain View, CA to pick his brain about his years at Yahoo, how the YC process works, and some of the biggest trends the incubator is watching as it chooses applicants for its summer 2017 startup class.

Startup Acceleration 101

PCMag: I want to start with some key pieces of advice you'd give startup founders when they apply and come in to pitch you. If you had to break it down, what are the characteristics you're looking for in a company—whether it's their idea, technology, business model, or a mix of everything?

Tim Brady (TB): It's a combination of all of them for sure, but most importantly is people. That's what investors do; they bet on people. Good people have good ideas. Good people coming in with bad ideas doesn't happen too often. Be passionate about what you're doing, think deeply about it. Those are the type of people we like to work with. Typically, someone will come in—a scientist or an engineer—and be good at building what they say they want to build. But they haven't put the time into the communication aspect of selling investors, customers, or future employees. We often focus on helping them do that better.

PCMag: Since Imagine K12 merged with YC last year, what are some of the most interesting startups with which you've worked? What stood out to you about them?

TB: Well, in the very first batch after I got here, there was a company called ExVivo. What they do is make an allergy patch. So, right now, if you suspect you have an allergy, you'd go to an allergist and do a scratch test. You might get a rash and it's kind of a painful experience. These folks have created this patch you put on your skin that sucks up these dermic fluids and, if you're allergic to cats or grass or whatever, the patch will turn a color depending on what you're allergic to. So, it takes this long, painful, inelegant process ,and make it fast, cheap, and painless.

That's two engineers and two scientists from Canada. That's it. Their intellectual property is the ability to put this micro needle on your skin without it hurting, still get it deep enough to be able to absorb these fluids. Once they have those, the testing is pretty straightforward. They're currently working with allergists and conducting some clinical trials.

PCMag: So, after you get a new batch of applicants and a startup comes in, what's the decision process like among the partners?

TB: There are some things you need to check off. So, we ask ourselves: Is this going to be a big business? We're interested in working with high-impact businesses. Can the founders make what they want to make or do they have to outsource it? There also needs to be good answers to: Why now? Why hasn't this been done before? What change or insight do you have that gives a compelling answer to that question?

That this is the right moment in time to bring your idea to market. After that, it's a lot about the personality of the founders. Startups are hard. You're going to get knocked down. Is this the type of person that gets up, dusts themselves off, and keeps going?

PCMag: Okay, so the startup is accepted. I want to go step by step through the investment and accelerator process that YC helped to pioneer. Twice a year, the firm invests a small amount of money in a large number of startups, which then move to Silicon Valley for three months before a demo day with investors. Then what?

TB: Every startup gets different things out of it. The great thing about the program is that it's not a cookie cutter process where every company goes through the same thing. The advice we give you, even the people you meet with, are different depending on your needs. We have something called office hours, which is kind of a 1-on-1 [when] the startup will talk with a partner or a visiting expert affiliated with YC and get advice on a problem. There are also group hours so, once you're in the program, we get teams together and do kind of a group office hours where people talk about their startups and talk to other companies running into similar issues.

PCMag: That plays into the larger alumni network aspect of the program as well.

TB: Right. So, those are the mechanics and the value of getting to know people and their advice, and then there's the network. Depending on the person, this could be the most valuable part of what we do. We have a "pay-it-forward" attitude. People help each other very actively. There's an etiquette to it, and people are happy to help each other out. That becomes a powerful thing when you have 1,500 companies across pretty much every industry.

PCMag: Then there's the stage where the startup has graduated and starts to look for funding. I want to touch on the sort of crash course these companies get when it comes to dealing with investors, closing a deal, and often getting acquired down the line.

TB: For most of the companies coming through here, one of the necessary stages is to raise funds. It costs money to grow and build a company. Very few are the startups that make money right out the door and it's enough to keep growing fast. So, investors are important. Our program culminates in an Investor Day. The program is roughly three months long, and on the last formal day, founders do an onstage presentation in front of a roomful of investors, and that kicks off their fundraising efforts. We educate them on how to do that, help with their messaging, and up until the end, we help make sure the company is on sound footing for investment.

The goal there is to make it happen quickly. Fundraising can take a lot of time and, when you're fundraising, you're not building your company. Fundraising is a means to an end, not the end. The quicker they can get done and get back to building their company, the better.

PCMag: What are some recent graduates that come to mind? Startups that have now raised successful rounds and had an innovative idea that you helped them shape, polish, and execute?

TB: There's a company from the last [Winter 2017] class called Lively. If you're familiar with HSAs, they're like a 401k for your health spending. Most of those have been done by banks, and there hasn't been a lot of front-facing consumer know-how. They're a terrible experience. So, these guys thought about it from a consumer standpoint: How would I want it?

For instance, having a credit card that automatically deducts from your HSA. Simple things like helping to implement management software for benefits administrators to manage everything better. Easy account management for the user. Helping company benefits departments implement that rather than the FSAs, which tend to be more complicated and cumbersome. It's a desirable product if done well. It's needed. It's one of those things that's been around but hasn't been done well, from both the customer and the company standpoint. They came in and pitched the hell out of it, showed well at demo day, and are now off fundraising.

Yahoo's First Employee

PCMag: I'd like to talk a bit about your background as well. You were Yahoo's first employee in 1994, at the beginning of a much different internet boom. Talk about that first business plan that attracted Yahoo's initial VC. How, in 1995, did you explain what Yahoo and the internet was supposed to be, and how it would make money?

TB: At the time, it was just an advertising play. We had to convince advertisers that it was worth trying. Our pitch was, "Hey, we can track it." Usually you had to call Arbitron or Nielsen and they would estimate how many people saw your TV ad. We can tell you exactly how many people saw your ad, whether they engaged with and clicked on it to go to your site. That aspect of measurability is what we used to convince people to jump onboard. The more eyeballs we got, the more ads we could sell.

PCMag: It's like trying to explain a simple version of Google AdWords before it existed.

TB: Right. Not exactly AdWords but brand advertising. When we started, there wasn't really e-commerce. You couldn't click through and buy. We had five advertisers in our very first set, and I think it was MasterCard and a couple of other brands looking for awareness. Initially, those five advertisers were plastered all over the site. There wasn't an active ad-buying market. We were the first ones selling online ads. We had to go out and tell people what the internet was and convince them to buy ads on it.

PCMag: You spent eight years as Yahoo's CPO, from the late 1990s through the bubble burst in the early 2000s. Looking back, how did the company change and evolve along with the breakneck pace of the early internet?

TB: We got big quick; advertisers came on quickly. We started adding a lot of employees and generating a lot of revenue—more than we anticipated. All these things we said would be cool if they happened down the line happened in spades, all at once. The number of people online, the amount of content that came online and how fast it appeared, the influx of advertisers that came on; it really was a breakneck pace.

My job was to figure out what users wanted and try to make the internet a daily experience for them. At the time, it wasn't. The content on the internet was finite. You would search and most of the searches didn't have anything. Today, if you put in a query and don't get anything back, you'll think you must've searched wrong. So, managing that and trying to do more and more with the content coming online. Real-time stock quotes, news, seeing what people were doing, figuring out what they wanted, and envisioning what was possible, and putting all of that together.

PCMag: Looking back, especially as the company ran into trouble in the early 2000s, what lessons did the experience teach you that translates to the startups you advise now?

TB: There are a lot of non-obvious things that wouldn't occur to you if someone hadn't told you. The order with which to do things is one. When you're short on cash and trying to raise money, there's an order to doing things that will help you grow faster. That's changing, too, with billions of people in the world now online. It's about watching the net evolve and the changes in how people are using it, and thinking about how to test new ideas in that environment. There's a flexible mindset you need to have. You don't have to change your vision but you might have to change the path you take to get there.

YC's Startup Hype Meter

PCMag: Given the number of entrepreneurs that walk in and out of your office on a weekly basis, YC has a unique perspective on what the industry looks like right now. I want to run down a couple of trends and buzzwords pervading the industry. For each, I want to quickly break down where the money is, where there's too much hype, what the biggest challenges are right now for a startup to be successful in it, and an example of a startup that's doing it right.

Virtual Reality (VR): Very early. We're also going through the hype cycle. People keep getting hyped up and saying, "This is the year," but I don't think we've found the killer app yet where it's a must-have. In the next couple of years, someone will figure that out. There are compelling games for hardcore users, but en masse, what is that app that gets my mom to put on a headset? We haven't seen it yet but we're looking. -TB

On-Demand Economy: The gig economy is still going. We don't think about it like "Uber for this or that." Is there a way to decentralize something that's been centralized? Uber and Lyft saw an industry that was centralized around cab companies and thought, "Everyone has their own car that sits idle most of the day. Why don't we use those?" Same thing for Airbnb. We'll still look at ideas…we don't believe that's over by any stretch…but we go in with a cautious eye on what's real and what's not. -TB

Blockchain: We continue to look at it. We think a lot of regulatory issues need to play themselves out, but a little bit like VR, the promise is there. But it's hard to know how far away that is. -TB

E-Commerce and Payments: A lot of the opportunity we're seeing recently is in developing countries. The payment platforms have largely played out in this country. PayPal, Stripe, and the like. It hasn't everywhere else. So, we're seeing a lot of activity on payments in Africa and Asia, both within these countries and cross-border payments. -TB

Augmented Reality (AR) and Mixed Reality: Before Pokemon Go, had you told me anything about AR, I probably would've given you the exact same answer I gave for VR: It's a little far off, there's some promise, but we haven't seen a killer app. Pokemon Go was that to some degree in AR. There might be one or two AR companies in this batch. -TB
Smart Cities: We've got a handful of companies in this batch focused on providing products to municipalities and governments to help cities run better and more efficiently. This is the first time in two years here that I've seen multiple companies taking a shot at this very specific goal. Whether it's helping police forces do a better job, helping officials make better decisions based on data, improving communication flows within government departments, monitoring infrastructure, putting all these technologies we've developed in business in use for cities and citizens. -TB

Artificial Intelligence (AI) (which I want to qualify a bit: I'm talking less in regards to pie-in-the-sky sentience and more about techniques such as machine and deep learning, neural networks, computer vision, natural language processing (NLP), etc., applied to large and complex datasets): We've seen a huge uptick in applicant companies using TensorFlow or IBM Watson, and taking advantage of those tools out there to make their products better. From an AI standpoint rather than the harder science of building this stuff, we're seeing startups using what's already available. They don't even necessarily need expertise, they just need to know what the tool does for you and write for that API [application programming interface]. We just started an AI track. We don't know whether it's repeatable yet but we've seen enough activity to start a dedicated track. -TB

CRM and Marketing Automation: We still look at it; that's a crowded market. We're more picky than maybe we were a few years ago. What we are seeing is companies taking these large Big Data sets and putting them through APIs to create better products. Although it's not sexy, back-end software sells. -TB

IoT and Wearables: At least since I've been here, there's always been a steady flow of apps. We accept a handful in every class. We believe that stuff is getting better and that we'll continue to see more of it, particularly on the sensor side applied to things like agriculture where you can make huge upgrades to your efficiency. -TB

Self-Driving Cars: The infrastructure for building automated driving is exciting because the upside of making it happen is so big. There's so much money in it; a lot of big companies are working on it. But one company from the last batch is called Lvl5, which does computer vision to read street signs, combining video and GPS data for objects that can't be picked up by LIDAR. Where you actually have to read signs, know what they are, and place those on a map. We're going to go through this period where you have a lot of self-driving and non-self-driving cars on the road at the same time. Self-driving cars are going to have to read signs in order for it to work.

PCMag: Finally, what's a trend you guys are flat-out tired of being pitched on, if there is one?

TB: We look at everything. It's more about the team necessarily than the idea. I wasn't around for this so I'm relaying the story but, when the Airbnb guys came in, they said, "We're gonna get these blow-up mattresses, put them in people's living rooms, and sell that." So, we have to think about whether we're being flexible enough in our thinking to recognize that again. Given that lesson, we try not to disregard anything.

Think about what CEO Drew Houston did at Dropbox. It had been tried a handful of times before he did it. We tried doing it at Yahoo in the late 1990s with something called Yahoo Briefcase, where the idea was similar. Store something online, access it from anywhere, and share it. But 1997 was probably a little early for that to happen. It's timing, and it's understanding the little things that matter, and getting those right.

This article originally appeared on PCMag.com.