Geoff Ralston: Okay.
The AC's on. We're good. Good morning everyone. Speaker 2:
Geoff Ralston: I am very happy to have my friend, Elad Gil, here. You saw, some of you, or all of you if you looked online, a talk yesterday about how to get meetings with investors and raise money.
If you would be so lucky as to get a meeting with this investor, and even more lucky to raise money from him, that would be a really good thing for your company. Elad's sort of an epic investor. He's had an epic career investing, but he's much more than that. He's not only a fantastic investor who's invested in a lot of YC companies, it seems like all the successful ones. Someday you have to tell me how you figured out how to do that. Maybe we can talk about that today. He's also a startup founder. He's worked at Google and Twitter, and he is increasingly known as an author. Claims it's not his career, but he is an author of High Growth Handbook, which is a very cool book.
I recommend you all read it, but not maybe right while you're starting your startup. Maybe as your startup gets going, because a lot of the topics in there will be relevant for you hopefully in the future. Maybe you can start off the conversation by telling us a little bit about how you got to where you are today. What's your journey to epic investor, and startup founder, and, I guess, company whisperer?
Elad Gil: Sure. First of all, thanks for having me here.
I'm really excited to be here for Startup School.
I moved out here right after graduate school.
I moved out.
I didn't know anybody. Many of my friends decided to stay on the East Coast, and so I was sort of starting fresh when I first came out here.
I had terrible market timing in that I moved out here right as the entire internet bubble was collapsing.
I went into an environment where, a year after I got here, most people were getting laid off, and most people were sort of desperately looking for jobs. So, somebody who was a vice president of product management was suddenly trying to get any product manager job they could get at any level at any company.
It was a very odd environment, and it was a very tough one to plug into.
This is before things like YC existed.
There was no online content around starting companies.
There was no content about basically doing anything, and so what you had to do is really sort of hand-to-hand combat in terms of networking, meeting people.
I basically talked my way into a job by offering to work for free at a startup.
That was my entrée.
It was basically an unpaid laborer on behalf of a company as a way to get my foot in the door and get my foot into technology.
That was basically my starting point.
Geoff Ralston: That's sort of a good lesson for how, even in today's world where there's many more resources, for how you have to ... You have to really pound the pavement.
It's elbow grease to get started lots of times, right?
Elad Gil: Yeah, absolutely.
I think what happened is after ...
I got a PhD in biology.
It's completely irrelevant to software and technology, but I wanted to go into software and technology.
Geoff Ralston: Well, completely irrelevant at the time to software and technology.
Elad Gil: Completely irrelevant at the time.
Geoff Ralston: But it changed later, right?
Elad Gil: It's changed.
Geoff Ralston: We can talk [crosstalk]
Elad Gil: Basically, my first software startup job allowed me then to be a software person, and that was the thing that allowed me to do a transition.
I actually moved out here to join a telecom equipment startup as a product manager, so I worked on hardware.
I had to leave that company, and then I talked my way into working for free at a software company.
Then, I was suddenly a software person, and that's what sort of led to me then eventually ending up at Google and other places.
Geoff Ralston: So, you ended up at Google, but let's talk a little bit about ... Because you ended up working at Google and Twitter. But you started companies in between, and in the end, sort of transitioned into becoming an investor. How did that all happen?
Elad Gil: Yeah.
I joined Google in 2004 and effectively started the mobile team there. So, I helped iAndroid and pulled together early teams for Google Mobile Maps and Mobile Gmail, each of which have hundreds of millions of users, but I had nothing to do with that since I left before that massive growth.
I left Google to start a very early data infrastructure company that was called Mixer Labs, was the company name. GeoAPI was a product name.
That was acquired by Twitter, and Twitter bought us when Twitter was about 90 people.
Then, my job quickly became one of helping the company scale from 90 people to 1,500 people over two and a half years.
The prior sort of super-growth journey I had before that was I actually joined Google at around 1,500 to 2,000 people, and it hit 15,000 people over three and a half years. So, very rapid growth when something's actually working.
Then, I left Twitter to start another company called Color Genomics, which is a big-data-meets-genomics company that's raised about $150 million in venture capital and is up and running still here in Burlingame.
Geoff Ralston: And finally made use of your PhD in biology, thankfully?
Elad Gil: Yeah.
Geoff Ralston: You're an interesting mix of big-company guy and small-company guy. How do you think of yourself? That's sort of unusual, isn't it?
Elad Gil: Yeah.
I guess I think of myself as somebody who is good at operating teams or driving efforts.
It could be a five-person team, or it could be a multi-hundred person team.
That's my self-actualization of what I do.
I actually think there's a lot of people who drop in and out of big companies and startups in terms of they start a company. Maybe it gets acquired.
They run a division.
They maybe go and start another company, or they become an investor.
If you look at Silicon Valley, it's sort of littered with people who've done this sort of back and forth. Ben Horowitz would be a good example of joined a company, which was Netscape.
It went public. He then started a company.
It got acquired by HP. Ran a big division at HP and then co-founded at Andreessen Horowitz. But that was a related career path in terms of somebody who bounced between the two.
Geoff Ralston: It kind of puts the lie to the trope that you have to fire the entrepreneur ... You're either an entrepreneur, or you're a big company person. But it's not quite so simple, is it?
Elad Gil: Yeah.
I think it's a good point because I think there was a real shift when Mark Zuckerberg hired Sheryl Sandberg as COO and to people saying that founders can keep going. Because what happened in the '90s was founders were quickly replaced by professional CEOs if they raised any venture capital. Because often the venture capitalists would take over the board pretty quickly, often at the series A. What that meant is that founders were often demoted, and they'd bring in sort of that old, gray-haired person to run the thing.
I think that really shifted in the era of Zuckerberg where suddenly people said, "Wow, you can build massive companies with the founder CEO still driving it." If you actually looked at the history of the biggest technology companies, that was always true.
Intel was always driven by its founders. Microsoft was always driven by its founders, at least early on, first 10, 20 years. Dell was driven by its founder. So, anecdotally, actually, founder CEOs work really well, especially if they're able to maintain control and not raise external capital, or if they raise it on very good terms where they didn't give up control. But the second they give up control in the '90s, they'd be out.
Geoff Ralston: Like Steve Jobs.
Elad Gil: Steve Jobs.
The founders of Yahoo didn't run the company.
The founders of Ebay didn't run the company. You can go through all the major companies of the '90s.
The founders of Google had Eric Schmidt come in. Up until the Zuck moment, I think, it wasn't a very common practice.
Geoff Ralston: Most of the folks here in Startup School are in very, very early-stage companies. Let's talk about what it's like in the early stage, if you can think back to Mixer Labs and even Color. How did you think ... You were the CEO of both those companies. How did you think about what should have most of your attention during that time?
Elad Gil: I think the singular thing that really matters is product market fit, so building something the customers then want.
That's really honestly the only thing that truly, truly matters. Now, to get there it means you have to hire people, unless you can build it all yourself and sell it all yourself.
It means that you have to raise money if you can't fund it off of your customers, or if you can't just bootstrap it. But when all is said and done, you need to build and sell product.
In the very early days, that's really the core thing. Yeah.
Geoff Ralston: We've talked a lot about product market fit and finding product market fit in Startup School. One of the most common questions that everyone asks is how do you know when you have it? It's not so obvious, it seems.
Elad Gil: Yeah.
I think it's obvious if you think about it the right way, but I think it's very hard otherwise.
I'll give you three examples of signs that you have product market fit from three different companies.
If your product is broken, and people are still using it very actively, or you have high retention on a broken product, that's a clear sign of product market fit. When Twitter was constantly going down and the Fail Whale days, and yet nobody moved off of Twitter, that was a sign of just raw market adoption.
A second sign is if you're a SaaS company, and you have major brands finding and using you organically and paying for your product, that's a sign of product market fit. Examples of that would be PagerDuty, which had, I think, Apple as an early customer.
There's Zeplin, which had, I think, Facebook using them very early.
Airtable has all sorts of brands that have just sort of adopted it.
I think there's a number of sort of more recent breakout e-companies where big brands are just showing up and starting to pay.
I think that's a clear sign of product market fit.
The last one is just whether you have very strong customer feedback, even from a small group of people.
At Color, for example, very early on, we were getting effectively love letters from our customers.
The company, Color, started off focused on providing people with information about their hereditary risk of certain genetically-driven cancers. We were getting emails from people where they were saying, "Thank you so much for helping me, for making this affordable [crosstalk]
Geoff Ralston: For saving their lives, even.
Elad Gil: For saving their lives in some cases.
They felt that Color had helped with that, and so we got very strong customer feedback very early on it.
Geoff Ralston: Yeah, so it seems like it's some mix of ...
Again, I think some of this is intuition, but some mix of retention, customer feedback, and maybe you should break your product and see if people keep on using it.
Elad Gil: I think there's other metrics, too. For example, people tend to discount growth off of small basis even though the growth is compounding. So, if you're growing organically, 20% a month, even if you're going from 100 to 120 users, that is a real sign of product market fit that a lot of people tend to ignore and not believe because they're like, "Oh, the numbers are so small, it doesn't matter." But if you're consistently adding 20 people, and then 30 people, and then 40 people month after month, that's usually a sign that the thing is really going to work.
Geoff Ralston: This is a while ago, but what went wrong, and what goes wrong for companies during this phase when they're trying to find product market fit?
Elad Gil: Yeah.
I think almost everything you can imagine can go wrong.
If you don't have product market fit, that may or may not derail the company.
If you have product market fit, then it kind of matters less, although you still want to correct those things.
I think, for example, at Mixer Labs, the very first product was a wrong product. We initially started off building almost like content aggregation sites or wikis for locations.
Then, we quickly morphed it into an infrastructure product where we were selling infrastructure to enable geolocation application.
This is before there was a big wave of API companies, and so it was one of the first, alongside things like Twilio or some of the really early developer tool companies.
The starting point for their product was much more of a consumer site that allowed you to aggregate content, and we built out very deep infrastructure for that.
Then, we decided that really what we should be doing is selling the infrastructure.
That was a good example of us changing direction as a startup.
Geoff Ralston: What else have you seen go wrong at this early stage that might be instructive, whether around product market fit, the folks you hire, even sort of the business moves, the marketing moves you might make early on?
Elad Gil: Yeah.
I think on the hiring side, a lot of people will let not-great people stick around for far too long.
Then, at my first startup, we actually let somebody go within the first month, one of our first six or seven employees.
The person was great, and they did well in their career after, but it just wasn't a good fit relative to what we were looking for as a company at the time, so I I think there's a ...
There's a old adage that you should either hire extremely well or fire very well. Optimally, you do both.
I think a lot of people tolerate a bad hire for way too long.
They'll wait six, seven, eight months and say, "Hey, we have to give them another chance, another chance." But really, you want to have an initial feedback conversation. See if they react. Maybe have one more, and then you have to take action.
I think the way you should think about your startup from a hiring perspective is it's almost like a life raft. Your boat is sinking. You're on this little raft, and there's five spots. Who are in those five spots? If the people on the raft aren't the five people that you should have on a raft, you should find the five people. Because you can only support so much from a burden perspective, or a productivity perspective, or a coordination perspective.
I think that's a very important mindset to be in in terms of who's on your team.
Geoff Ralston: This might be a hard question to answer, but I was sort of thinking about, a lot of times, we focus on what goes wrong.
A lot of things go wrong. You guys have seen out there, things go wrong. You've invested in so many epic companies, and your companies have done great. What goes right in those? What stands out as a set of things that people are doing or steps they're taking that kind of gives you that clue that this might be one of those great companies?
Elad Gil: Yeah, I think the focus on customers and what customers need very early is an important signal, and then I think building something immediately is really important, too. Now, that doesn't mean you shouldn't be doing analysis. You should absolutely be thinking about the market. You should be thinking about your market strategy. You should be asking who your customers are and defining out your segments. But you should also just build stuff and try it and start to see reactions rapidly.
I think where I've seen a lot of people fail is they wait and wait and wait in terms of actually building something. Sometimes they just wait too long, where you see these companies that stay in stealth mode for three, four years, and you're like ...
That's a terrible sign because you're not actually getting real customer feedback potentially.
Geoff Ralston: It seems like there's also something about iterating fast in these great companies where there's always something new and better, and it's always better, too.
Elad Gil: Yeah, I think that's definitely true.
The other thing I've seen some people do, which I think actually helps a lot is from day one, they focus on the velocity of the team. So, they may be focused on the build environment early and making it so that anybody can run a full instance of the product with a single line of code just to set it up, right? There are actually things that you can do to make your team dramatically more productive that a lot of people don't do.
Anytime they onboard somebody, it takes them two weeks to get the damn thing running on their laptop instead of actually saying, "How do we make it so that any incremental person who hires or anytime we push a change, we can instantly see its effect?" I think there's almost like little things you can do along the way there, too.
Geoff Ralston: Yeah. Since almost all of the companies that we're talking to are driven by human capital, people don't pay enough attention to making that human capital as effective and efficient as possible. Speaking of human capital, you are CEO in both your companies.
In fact, you've given a lot of advice to CEOs both as an investor and in High Growth Handbook, but later on, advising a CEOs and other executives how to be most effective. How do you think about that for early-stage startups?
Elad Gil: Yeah.
I think for an early-stage startup, ultimately the focus should be on how do you make your team as effective as possible if you have a team? Effective means two things. One is people are all moving in the same direction, and there's clear goals, and there's a clear sense of what you're actually building and why. But also it means, to your point, iterating quickly.
It means talking to customers early.
It means, again, focusing on product market fit. Because I think the mistake that people make is you could have the happiest, most wonderful group of people in the world, but if you never find product market fit, your company is going to die because you're just not going to have the revenue to support it, or the cash to support it, or you won't be able to raise more money. So, whether you have kombucha or not is secondary to just building something that people really want.
I think people get distracted along the entrepreneurial journey, either in terms of eventually if they get funded, and they get invited to all these events, and all these other things start to distract them.
There's a lot of sort of entrepreneurial distractions that could get your attention, when really the core thing is are people buying your product, and at what scale? And are you growing at a good rate? Now, that's if you're on the venture track, and you want truly a high gross breakout company.
There's all sorts of other startups you could start. You could start a lifestyle business, and that could be great.
There's a variety of different reasons to do a startup. So, part of it, too, is asking yourself what do you actually care about as a founder, and what's your objective? And then, are your actions actually mapping against those objectives? I think too many people just go on a default path that isn't the right path for them.
Geoff Ralston: What should a CEO be spending most of their time doing?
Elad Gil: I think it differs early versus late. For an early company, it's going to be building, so are you actually helping to build a product? You may not be the person that should be building the product.
I'm just saying if it's one or two of you, you're probably building or selling. Second is hiring and making sure that the team is productive and focused on the right direction.
Then third, making sure that you don't run out of money. Lastly, I mean, it's avoiding fights with your co-founders and things like that.
I often say that an early-stage startup largely fails for three reasons or only has to do three things. One is you need to find product market fit, which is incredibly hard. You need to make sure that you have the money to survive, and you need to not fight with your co-founder because that'll slow everything down. For an early stage startup, that's ultimately everything you need to do.
I mean, you need to hire, and there's different aspects of selling that you really need to focus on, selling investors, selling customers, selling hires. But it's either build or sell. For a late-stage company, it gets way more complicated. You're still selling for all sorts of reasons, but then you're also much more focused on how do I enable this thing to scale in terms of processes, in terms of people? How do I hire executives? How do I internationalize? How do I launch new products? How do I buy companies? Then you realize quickly that that has to be a team effort because no one person can do all those things.
Then you really start asking, okay, how do I build out and manage a really effective team?
Geoff Ralston: One of the hard roles that a CEO has, I suppose, is to figure out, when you don't have product market fit, whether you need to tweak the product or you need to do more of a pivot. You mentioned that at Mixer Labs, you guys had to do a pivot.
A lot of people ask that question, like, how do you know? Like, when do you know that you've hit the wall with what you're doing, and you need to do a more significant shift?
Elad Gil: Yeah, it's really hard because you ... Most startups build something for six months to a year.
They launch it, and then they hit a wall.
Then the question is is that a real wall, or is that a fake wall? I think it's incredibly hard to know.
I think if you're three, four years into something, and you're just not seeing any traction or momentum, you clearly waited too long.
There's counterexamples to that, and there are always going to be counterexamples. For example, TomTom, that European GPS company that ended up for a while being one of the most valuable companies in Europe ...
I think it was literally four people in a room for, like, six years, iterating through different ideas until they came up with the concept of doing a GPSs, which, of course, now are subsumed by phones. But they took a really long time to get there. Most of the companies that I've seen work tend to work early, at least in terms of adoption, or at least in terms of some core metrics. So, I think it's possible that in today's environment where capital is really loose, people actually wait too long to pivot or to shut down because they can keep raising more money.
Actually, I think that's a bad thing because it's locking up great people at the height of their creative point in their lives and preventing them from going and doing something else. So, I do think at some point, you need to have that really frank assessment and maybe do it periodically. But should we really keep going? Then, if we're going to keep going, I think the key question is when you pivot, the one thing I would advocate is being willing to restart completely. Because typically when people pivot, they pivot within the same market versus across markets.
If you're in a bad market, that means you're going to die because you're just going to go onto the next thing. You say, "Hey, we have all those sunk costs." You start iterating on that new thing, and it doesn't work either.
Geoff Ralston: So, it's sort of a decision between ...
An evaluation of the market being bad versus it's just you have the wrong product. Could be great market and bad product, but ...
Elad Gil: Yeah, you should assess it from first principles.
Then, if you decide that you're done with the business, you have three options. You shut it down, you sell, or you pivot.
If you sell, my suggestion would be sell to somebody that's clearly breaking out.
Geoff Ralston: I think I've heard you say before that, speaking of being in a great market, that that's sort of your fundamental test.
If you believe that there's this incredible market opportunity, you want to be in there. You want to be fighting that fight. Even if you don't have the right product at first, you can find it.
Elad Gil: Yeah.
I think the singular determinant of startup success is the market, and it's not the strength of the team.
I think a lot of people talk about how great teams will always figure it out.
I've seen great team after great team in a terrible market just die because they're in a bad market. So, they never have the opportunity or the time to really figure it out or to iterate as much as they could if they were in a very good market where there's just cash inflows hitting you despite what you're doing.
Andy Rachleff, one of the founders of Benchmark, has what some people now call Rachleff's law, which is great team, terrible market, market wins. So, it doesn't matter how good the team is if the market's awful.
Terrible team, great market, market wins. You can actually be not very good at what you're doing and still do very well. Your product's half broken, but it's still getting massive adoption because you're in a great market.
Then, there's great team, great market. Something magical happens, which is like a Google or Facebook where you had people who then executed, they added new product lines, and they were really great at what they were doing in terms of scaling it to the next level.
I do think that's very true.
I've seen many, many great teams just die in a terrible market, and then I've seen a few bad teams do very well.
That to me was very surprising.
It feels very non-egalitarian or merit based.
Geoff Ralston: Why do you think it is that so many teams choose such bad markets?
Elad Gil: I think it's maybe three reasons. One is sometimes people just jump into things, and they just start building without actually thinking it through. Sometimes that works really well because it's a product you want or need, and so you're building it for everybody else. But sometimes it's something that has been tried many, many times before.
That doesn't mean it won't work in the future, but maybe you should look at the history and ask why didn't it work, and what should I be doing differently versus just doing the same thing again that's failed five times? The second thing I've seen a lot of people fail at is what I'd call multi-miracle startups.
I think every startup needs at least one miracle to succeed. Because if it was obvious that it was going to work, everybody would already be doing it, and so you wouldn't have an opportunity.
In some sense, startups only have opportunities in markets that are non-obvious, which means they have to overcome some obstacle.
The distribution needs to magically work, or the product and the price point need to magically work, or whatever it is. But there's some companies where they come up with two miracles for their startup.
They say, "Our stage one of the startup is we're going to do X, and then stage Y is we're going to do this other thing that's completely unrelated. But because we did X, we can suddenly do Y." Usually those fail because you're saying, well, you have to miracles, and you're compounding very low probability, so you're just bound to fail.
An example would be there was an era where people said that the way to compete with Yelp was to start an events product, which is an open space, win in events, and then it's adjacent enough to what Yelp does that you could add local listings.
Then, you'd win in local listings.
That's a terrible approach. You should say either, "I want to win in events," and just double down on that, or, "I want to go after listings, and then I'm just going to do that." But this sort of multi-step thing tends to fail.
Geoff Ralston: Which is, I think, an object lesson in how not to approach your investor meetings.
If you're talking to an investor, and you have a whole bunch of miracles that have to happen before you're a big company, then that investor's probably not going to be that interested.
Elad Gil: Yeah.
The most common miracle that's quoted today is a data moat. People say, "Well, we'll generate tons of data, and then we'll be differentiated." I think in genomics, something like that could work, but outside of that, I actually think I've never seen a company pull that off in terms of a broader ... Data is the only thing that's their asset, and then they're going to somehow do something with that asset.
Geoff Ralston: Yeah, or we're going to get ... Our AI is going to get the smartest because we'll have the most data.
Then our AI will be the best, and therefore, we'll win."
Elad Gil: Yeah.
Geoff Ralston: Just back to the CEO question, how should teams choose their CEO? Who should be the CEO?
Elad Gil: It's a good question.
I mean, there's different ways that you could describe or argue it. Ultimately in some cases it's a clear, hey, this person is the person we're choosing.
In some cases it's a huge argument.
I think the most important thing is to resolve it even if you get it wrong. Because you can always correct the wrong thing, but at least you have somebody in charge.
I think a lot of founding teams end up with a lot of fights that then slow down the company dramatically if they don't have somebody clearly in charge.
I think if it's a very technical product, it's often better to have the technologist in charge. So, say that you had a businessperson and technologist. Because if it's two technologists, I don't know how to differentiate between who ... Maybe somebody who's a little bit better at selling or recruiting or that part of the job. Maybe that's the right person for the CEO if it's two technologists.
If it's a businessperson and a technologist, usually the businessperson is going to be better at those things, but that doesn't necessarily mean they should be the CEO. Because often the iteration of the product and the understanding of the product is driven by the technologist, and that really impacts how you sell it, how you iterate on it, and the decisions that you make along the way. But honestly, I think any configuration can work.
I think sometimes a founding pair will be very self-honest.
They'll just be like, "Oh, this person is better at the things that we need from the CEO today, which means better at recruiting, better at fundraising, better at selling and better at sort of setting vision and direction." Then it's a clear shot in terms of who should be.
Geoff Ralston: You might even say that you should put in the position of the CEO the person who's going to help you find product market fit, who's going to drive the company to product market fit sooner and better?
Elad Gil: That's a good point. Yeah.
Geoff Ralston: I wonder.
Elad Gil: Yeah.
Geoff Ralston: Why were you the CEO? Because it was your idea? Because ...
Elad Gil: Yeah.
It's interesting. For example, with Color, I stepped down as CEO after four years.
The origin of the company was a bit of a mix, but ultimately it was driven by my co-founder's personal story where he himself is a BRCA2 carrier. His mother has had breast cancer twice. He's had members of his family die of the disease.
And so the company was started in some sense as a form of patient advocacy where we said it's ridiculous that people can't afford and access these really important pieces of genetic information that can really drive what they should be doing from a health perspective in their lives.
I think that-
Geoff Ralston: So, it's his passion and his drive more than anything?
Elad Gil: It was his passion and drive.
I mean, I had a lot of passion for it, having worked in biology and things. But fundamentally, I think one of the reasons the shift we made is so good is because this is something that he's truly lived and cares about deeply.
I think that really comes through as well in the choices that are made in the product, and how do you think about patient data privacy, and how do you think about all these things? I think he had a very deep intuitive sense of what it would mean for him, which I think was very important in the context of what Color was doing.
Geoff Ralston: It seems like a very modern way to look at the CEO as the ultimate product advocate, Steve Jobs being the obvious example there.
Elad Gil: I think in tech, we're very imprinted on that style. But there's also great examples of CEOs who are amazing salespeople. Benioff started off in sales in, I think, Oracle or something, right?
Geoff Ralston: Right.
Elad Gil: There's all sorts of examples of non-technology CEOs who've done extremely well. Or, you have a situation where you do a mix. Like at Google, Eric Schmidt had a PhD in CS, but he'd also run Novell before. He'd run a public company, and so it was somebody who'd done the business side of things as well as the technical side.
Geoff Ralston: Yeah.
I haven't figured out whether it's a trend or not, but it seems a little bit that way.
I always find I prefer if I'm looking at investing in a company or looking at a company for YC that a product person ...
That the CEO really has a feel for the product anyway.
Elad Gil: Yeah, I totally agree.
Geoff Ralston: Let's transition a little bit to fundraising. We've been talking a fair amount about that recently in Startup School.
I don't know. Did you have interesting experiences fundraising that you can talk about in your two companies, Mixer Labs and Color?
Elad Gil: Sure. We took a very different fundraising strategy. For the first startup that we did, we ended up raising a seed from a brand-name venture fund, plus a bunch of angels.
Then we did a second round from sort of a super angel or small-fund group.
The biggest issue we had there is we had a bad investor who, as we were exiting the company, basically tried to claw more value for themselves as part of our exit.
This is sort of a Midas List investor who's very well known now.
The second company, my-
Geoff Ralston: Hang on.
This is a successful investor who tried to screw other people over while they-
Elad Gil: Yeah.
Geoff Ralston: Huh.
Elad Gil: Yeah, I mean-
Geoff Ralston: And their name ... No, don't tell me. Shouldn't go there, I guess, but-
Elad Gil: I'm always tempted, but I always stop myself. But I do think that there are some brand name investors who are dicks, right? So, you should be careful in terms of who you work with. You should diligence them and ask the founders where either things that didn't go well or where there was some issue over time or whatever it is.
That's more likely to reveal the truth about an investor than asking in a situation where everything worked.
Geoff Ralston: We like to say a YC that being nice is a big advantage, but it doesn't always work.
Elad Gil: Yeah.
Geoff Ralston: Sometimes it's not true.
Elad Gil: Yeah, and I think I would differentiate between two things.
I think you want investors who will push you, or question what you're doing maybe is a better way to phrase it. Because I do see some founders sometimes uncomfortable with being asked about some of their decisions, and you should be asked about your decisions.
That's a way to get better.
It may be a naive question that an investor asks, but that may be the thing that makes you think about something that you should be doing differently, or maybe it just reaffirms what you're doing is right. Right? I would differentiate between that and somebody acting badly in terms of clawing value for themselves or doing other things.
I mean, we did have ... We had a pool of investors.
There's only one who acted badly out of a dozen people, so it's not ... Or 15 people or something.
Geoff Ralston: How easy was it to raise money at Mixer Labs, your first company?
Elad Gil: Yeah.
It took us about two or three months.
In general, I think you hear the news about these fundraisers that happen in a week, and you think every fundraiser's going to take a week. But if you're actually being thoughtful on your side as well, it should take a couple of months because otherwise, you're not getting to know the people that you're fundraising from. Usually what happens is you'll find that until somebody really tips, most people are scared to jump in, which is really weird. Right? In general, you'll find that most investors are fear driven versus ambition driven.
They're kind of scared of missing something versus really excited and believe vigorously in something. So, high conviction investors are actually very valuable in my opinion.
Geoff Ralston: Right.
They're few and far between.
It's hard to be the first money in on a company. You have to just really believe, especially with a new founder.
Elad Gil: Yeah.
A lot of investors fall for the opposite of that, which is there's a brand name investor in. So, they go in, but it's actually a bad investment, because everybody makes mistakes.
Geoff Ralston: They're sheep.
Elad Gil: Yeah.
Geoff Ralston: How about with Color? You'd been a successful entrepreneur already. You were a well-known executive at that point, and so you were sort of in a different spot raising for Color?
Elad Gil: Yeah, I think for many founders on their second company, it's much easier because they have relationships. My co-founder and I did most of our seed, and then we brought in people for the A and as people that we'd known for a while and that we really trusted and respected. We had Joe Lonsdale and Vinod from Khosla Ventures come in for our A. For the C for Mixer Labs, we really had to network to get intros. We met Michael Dearing as he was just starting to invest. We were one of the first companies he invested in, and we met him through Kim Scott, who wrote the book Radical Candor.
I worked with her at Google.
I just asked people I know, "Who do you know that's investing right now?" And then we ended up having Sequoia in the round, Reid Hoffman when he was still an individual and evolved from AngelList. We had a bunch of really great people line up, but we had to network for the first few.
I think it was Dearing maybe who tipped first, or somebody else, and then everybody kind of catalyzed.
Geoff Ralston: How does someone get a meeting with Elad Gil?
Elad Gil: You make it sound like that's a valuable thing, but I think a ...
Geoff Ralston: Well, it depends whether those companies that you invested in would have been successful without your investment. Maybe it's the most valuable thing you can get.
Elad Gil: I think that the most successful companies and 95% of the cases are successful despite their investors.
I shouldn't say despite.
I should say irregardless. Now, the investors may help optimize certain things, or they may bring in the key exec.
I'm not trying to minimize the impact of investors.
I just think product market fit is such a raw force that it's helpful to have people around you who can help with scaling and other things.
I've seen companies go the other way where they have a lot of bad capital all the way through all their rounds, and the founders are drowning for help. So, you really do see a difference.
I'm not saying that there isn't a difference.
Actually, I know one company in particular where they had a series of initial weak investors, and their board meetings are completely ineffectual.
They haven't gotten a lot of help.
They haven't built out an executive team. Like, it shows. So, I'm not saying that.
I'm just saying to get from zero to one, it's not your investors who will do it most of the time.
There are some counter-examples, like I think a Paul Graham helped actually with some of the very early ideas for Reddit and things like that.
Geoff Ralston: Airbnb especially.
Elad Gil: Yeah.
Geoff Ralston: But it's pretty unusual for an investor to be transformative for a company.
Elad Gil: Yeah.
Geoff Ralston: So, getting that meeting?
Elad Gil: Oh.
I think the best way is to get an introduction through a founder that I already know or somebody else that I already know so that it's in network.
I think if you can't find a way to do that, then you're probably not going to find a way to get customers. You're not going to find a way to get hires. You're not going to find a way to do all sorts of things, so I think that's the best way.
Geoff Ralston: Yeah.
It seems like Paul Graham's original idea of doing things that don't scale works across different domains, not just building product, but figuring out how to get to the right investor and get that first person to tip?
Elad Gil: Yeah.
It's a component of almost everything you end up doing as a startup. Hiring is also not going to be scalable early, and you're just going to have to grind through giant lists of people. For example, at Color, because we are a CLIA, CAP lab, and we're focused on the regulatory compliance of what we do, we had to hire specialists called CLSs into the lab very early because they had licenses that allowed them to operate certain types of tests.
In the state of California, we could only find two or 300 CLSs total, and we had to hire two or three of them. We had to hire a 1% of all of these people in the state of California, and we wanted to find people who are very good.
If you have a license that's very rare, it means you can get hired really easily, which doesn't necessarily mean that you're great.
I literally ground through 200 people. We put them all on a spreadsheet, and I and the recruiter I work with contacted every one of those 200 people and had a conversation with all of them. So, you just have to grind through stuff.
Geoff Ralston: Yeah.
It's hard work doing a startup, huh?
Elad Gil: That's why I'm taking a break.
Geoff Ralston: Yeah, right.
There's so much I want to talk about. We'll do it ...
I have a few more things, and then we'll open it up for questions. You sort of already answered this a little bit, but maybe you can elaborate on your decision process when you're looking at an investment.
Elad Gil: Yeah.
I think I have three really simple criteria.
They're going to sound very generic, but I think about them deeply, hopefully.
The first one is market, or product market is a better way to phrase it.
Is the team building something that the product will actually want or need? There's different ways to check that.
It could be something that I would have used myself at my company.
It could be you see clear growth or traction from something that's launched, or sometimes I'll literally call potential customers and ask them.
I actually do have a lot of market diligence for things I invest in even though I'm just an angel.
The second thing is are they people who are very high caliber, they can learn quickly, all the rest of it? There is the founder of component, which I think is really important.
I just think the market dominates that.
Then lastly, if they were to call me at 10:00 at night on a Friday or Saturday night, would I take the call, and would I be excited to talk with them? Are they good people? Are they ethical? Would they be easy to work with? Because life is short, and there's lots of companies that'll work.
The last thing you want is a long conversation with somebody terrible. Speaker 2: [inaudible]
Geoff Ralston: That's true. What are the things you see, excuse me, companies doing wrong when they come and meet with you?
Elad Gil: Good question.
I think it's a few things. One is not getting down to the details quickly enough. Sometimes people will have the preamble of their personal story, how when they were seven, they lived in Canada.
Then they were nine.
Then they were 11, and they started learning to write code.
Then something else happened.
Then eventually they came up with the idea in college, but they didn't work on it for seven years while ...
I would just make it a very crisp, concise story.
This is what we're building.
This is why we're building.
This is who I am, but here are the three highlights. Here's who my co-founders are. Here's the market that we're in. Here's a product we're building. Let's show you a demo.
That would be sort of the optimal early-stage conversation because the demo means they actually built something instead of just talked about it.
The quick analysis of the market/how does this fit in or what's the use case is really important because they've thought through that dynamic.
Then the team, who are you, what do you care about, why are you building this is important, versus I was seven in the Appalachian Mountains ...
Geoff Ralston: Get to the point, in other words.
Elad Gil: Yeah. Get to the point quickly and also make efficient use of time.
If the meeting should only take 20 minutes, that's great. What I've found is that very busy people will spend more time with you if you're very efficient with their time because they don't feel that you're just going to sit there and take up the time because they can.
I've had meetings where ...
I'll tell you one example meeting, which I thought was amazing, and I didn't think it would happen.
I was looking into funding things and anti-aging or longevity-related technologies.
I think it's a big gap in biopharma, and I think it's important socially.
There is this founder, Ben [Caymans] , who I met with. We had, I think, 45 minutes set up, and we talked for 10 minutes.
At the end of 10 minutes, he said, "Okay, I got everything I need out of you.
I'm going to go think about this stuff, and then I'll follow up in four days." I was like, "Okay, this guy is just going to disappear." Like, nobody does a 10 minute meeting and actually comes back with anything useful.
Then, he came back in four days with this big analysis of everything that we'd discussed, and what he was going to do next, and how he was going to approach it.
That was a 15, 20 minute meeting.
I was like, "Oh my gosh, this person is amazing.
They're extremely efficient.
They have amazing follow through.
They say what they're going to do, and they do it.
They realize that this is a multi-interaction thing, and therefore, how do I build that confidence in that person that they actually want to keep talking to me?
Geoff Ralston: That's a great lesson for future investors, too, huh?
Elad Gil: Yeah.
Geoff Ralston: Act like that.
I want to turn over to Q&A. But just before we do that, you just wrote a book, High Growth Handbook, which I think is a fantastic book. Maybe not so relevant to running a startup, although it seems like there's a way to sort of track back from all the lessons in there, and it helps you think about your startup, I think, in an effective way.
Elad Gil: Yeah.
The book is called the High Growth Handbook, and it's basically almost like a section-by-section view of a lot of the common things that happen in startups and then tactically what to do about them.
It's a very tactical book that doesn't have a lot of platitudes. Like, there's nothing about A players hiring A players, which doesn't really help anybody.
It's not like somebody's going to write, "I'm an A player," on their resume. But it really helps you understand, okay, you want to build a recruiting process. What's the multi-step process you should build, and how should you go about doing it? Then it's interwoven with interviews with Sam from YC, and Claire Hughes Johnson from Stripe, and Marc Andreessen, and Reid Hoffman, and Rajen Sanghvi, and a bunch of other people around their experiences operating or funding companies that are relevant to each section. So, there'll be a section on board, and then there'll be an interview with Reid Hoffman around what he looks for in board members. He just wrote Blitzscaling, which is sort of a really interesting book around scaling companies that I definitely recommend as well.
The interesting thing is, though, that a number of the sections in the book actually started off as blog posts on my blog. Some of them are actually targeted to early-stage founders, so some of the stuff around hiring maps into it, or some of the stuff around product management maps into it.
I do think there are some things that are universal and then, to your point, some things that are completely divorced, like buying a company, or raising $100 million, or whatever it is, or obviously for much later stages or doing a big reorg, things like that.
Geoff Ralston: But the way the book is structured, it turns out it's easy to skip around. So, you should buy the book ...
Elad Gil: Oh, thanks.
Geoff Ralston: ...
And find the parts you like. Let's open to Q&A, questions for Elad. Speaker 4: This question's about investor updates.
A company that has long sale cycle, let's say four or five months ... What should be the frequency for the initial date, and then what should be put in that initial date? Because now [inaudible]
Geoff Ralston: I'll repeat the question.
The question is about investor updates. How frequent should they be, and what should be in them?
Elad Gil: Yeah.
I have a blog post I actually wrote on this, and I think that you should do it monthly. My suggestion in terms of the sections would be the top section should be what are your asks from your investors, or investors and advisors, or whoever you're sending it to, so it's very clear immediately how they can help.
If they don't read anything else, if they just open the email, they'll see that, and you'll actually get a lot of responses to that.
I think the second section could be metrics, if you have metrics.
It could be sales.
It could be that it's flat, and that's fine. Right? I think burn is important to include, so how much money do you have left? Paul Graham often advocates for default dead or default alive.
Are you going to run out of money at some point or not? I often just phrase it similarly as like how many months of cash do you have left, and how much are you burning? Then, you can have updates around team, around market, et cetera, or product. Did you launch anything new? Press ... Did you get any? I think you can follow that.
The reason I think monthly is important is you want to create high enough frequency that your investors are thinking about you and that they feel that they're up to speed on things, but not so much that it overwhelms you or them.
I think monthly is a nice cadence.
At Color, we started sending out investor updates really early.
I'd call up an investor, and I'd start to give them background in terms of what's been happening.
They'd say, "No, no, no.
I read your update.
I know what's going on. Let's jump straight to it." By being very transparent, it actually helped immensely in terms of not spending the first 20 minutes getting people up to speed, but every investor call suddenly was very efficient.
And they love the transparency because they feel like they're part of the company and the team and that you're open with them.
Geoff Ralston: Yeah. Speaker 5: So, you're a big proponent of the market-first investment strategy. How did you evaluate an obvious market like Airbnb back in 2010? How did you evaluate that market opportunity when it wasn't public?
Geoff Ralston: The question is about market opportunities. Since Elad cares most about that, how do you evaluate opportunities that are not obvious, like Airbnb?
Elad Gil: Yeah.
I invested in a versus their A versus their seed.
I'm guessing the seed was harder.
I think at their A, they had pretty clear traction.
I mean, they were growing at a good clip.
Then, the other side of it was I had actually traveled on a service called Servas, which ...
After World War II, the Esperanto community thought that one way to encourage world peace was to let people stay with each other for free. You'd apply to be a traveler, and they'd literally give you ... You do interviews, and then they give you these little booklets.
If you're traveling to Italy, it'd list everybody in Italy who would be willing to host you, and their phone number, and how to contact them. You'd literally call random strangers up or email them and say, "Hey, can I stay with you, and how long can I stay with you?" And you stayed with them for free.
I traveled on a very similar service throughout Europe, so for me it was a no-brainer.
I thought, wow, you can monetize this.
The most expensive asset that a person has as their home.
If you can help people monetize their homes and make a lot of money off of it, it actually felt like a no-brainer. But part of that was driven experientially, and then part of it was it was working.
Geoff Ralston: More questions. Yeah, back there. Speaker 6: Can you talk about doing due diligence on investors with the bad piece, what things can go wrong, didn't go right? Because investors are not going to do that [inaudible]
Elad Gil: Yeah, I would basically-
Geoff Ralston: The question is about can you talk about due diligence on investors who necessarily give you a bad reference?
Elad Gil: Yeah, I think it's ...
I would just look ... Most of them will have a list of investments either on their website or on AngelList or Crunchbase, and so I would just ...
If you're part of YC, I think there's a Bookface, which has all sorts of information.
Geoff Ralston: We have an investor database, yeah.
Elad Gil: They have an investor database, so you can actually see what everybody says about them.
Geoff Ralston: Elad has a good grade.
Elad Gil: Hopefully.
I would just look at that list and then call the two or three things that you'd never heard of, or that shut down, or were small acquisitions, or whatever it may be.
I mean, sometimes you'll call a successful company, too, and they'll say, "I'll never work with this person again." I know one company ... You can actually see it. Say that somebody's exits their company for a billion dollars.
Then they start another company, and in no rounds do they take their prior investors.
That's usually a sign, and so I would call up that person. Speaker 7: My question is about pricing.
As part of product market fit, I feel like right now we're looking at pricing market fit.
Anything that you can share about pricing journey with your companies or what you advise CEOs?
Geoff Ralston: Any advice on pricing?
Elad Gil: Yeah. Marc Andreessen has this great quote, which ...
If he had to tell startup founders one thing, I think he says that he would tell them to raise prices.
I think in general, especially technology-driven companies tend to really under price their product under the belief that if they price it really low, it'll get them more market share more rapidly, and therefore, they'll win the market. When in reality, they've never run the experiment, and it's really easy to drop prices.
Geoff Ralston: And really hard to raise them.
Elad Gil: And it's very hard to raise them.
I do think that people also overthink pricing and the changes and the impact of them when they have small populations.
If you have 100 customers that are all small customers, and you raise prices and piss them off, that's okay.
I mean, you have to do it nicely, or maybe your grandfather them in.
There's all sorts of ways you can do it the right way. But I'm just saying it's not an existential moment for the company, unlike if you have 100,000 customers, or you have giant enterprises using you, and you're going into these deep negotiations.
I think when you have a very small base, you can actually experiment. You can A/B test things. You can do all sorts of things. But in general, I would start on the high end unless your strategy is explicitly to go for market share and there's a reason for doing that.
There are certain markets where that's very important. Certain areas where you really are generating a data asset, those may be places where you want to have low pricing. Often people will just look at their competitors' websites and say, "Oh, they're charging 9.99 a month.
I should, too." But often that company hasn't really thought anything through either, unless they're a older or more sophisticated company. So again, I think I would just sort of start from scratch.
There's long things around are you doing value-based pricing or cost-based pricing or ...
There's different ways to price products.
I think Michael Dearing has a great online tutorial, which I would definitely look up. You can do a search for it, like Michael Dearing pricing, and there will be a variety of different segments and approaches.
The other thing you could do on the pricing side is good, better, best is a common framework. You have a base product, you have an intermediate product, and then you have the high-end product.
The high-end product exists largely so that people buy the intermediate product.
There's all sorts of pricing tactics that tend to work. Speaker 7: Thank you. Speaker 8: Yeah. When you are looking for a [inaudible] consumer company, product company.
In the beginning, you find it difficult to identify if it's a good product with a bad marketing strategy or communication strategy. Or maybe the content inside the problem that maybe is not fitting the customers right. How do you talk about and identify that problem to see if you have a good product with a bad strategy or a bad product?
Geoff Ralston: I think the question is around, especially in consumer markets, how can you tell what's wrong with the product? Is it your strategy, or is it the content? Or is it some feature of the product that's missing?
Elad Gil: It depends a little bit on what type of consumer product it is.
If you have an e-commerce site, it really is about LTV and CAC. Can you acquire customers at a cost less than what you sell the product to them, and what's the lifetime value of that customer? There, there's a very small handful of distribution strategies that tend to work, be it SEO or SEM or other types of either ad formats or customer acquisition strategies.
If it's more of a social network, then really it's looking at growth rates, and churn, and what are the vectors by which you're pulling people into the product? So, it can get ...
I think there's a dozen different answers depending on the market segment.
Andrew Chen has some really good blog posts, I think, about some of this stuff, so that may be a good reference to go look up.
Geoff Ralston: Okay, one more question. Back there. Speaker 9: In terms of product market fit, what do you consider a real or a fake wall?
Geoff Ralston: In terms of product market fit, what do you consider ... Or, how can you tell the difference between an actual wall where you're about to get smashed or a fake wall that you should smash through?
Elad Gil: I think every startup that succeeds has fake walls. So, to some extent, I think it's back to the discussion we had earlier, which is what is a sign that your product's actually working? And it depends on what you mean by a wall in terms of ... For example, if you're growing 20% a month off a small base, I think that's a sign that it's working. Even though people may say, "Oh, you have a small user base. You're hitting a wall because you're never going to get that big." Then you see these companies become giant companies because the use case is much broader.
Airbnb is a good example where I think that market size was bigger than anybody expected. So, it depends a little bit on what sort of wall you're talking about, but I do think there are clear signs that something is going to work. Often they're ignored by the founders themselves because it's just such a leap of faith to think this small, delicate thing that you're working on that's very breakable will actually turn into this amazing, giant company.
That actually happened to me when I went to Airbnb's Christmas party a few years ago.
I invested when it was about eight people.
Then I went to the Christmas party and there was like, I don't remember how many, 2,000 people or whatever the number was at that time, dressed in suits.
It looked extremely like a very nice corporate holiday event.
I was like, "Oh my God, this is amazing."
Geoff Ralston: What happened?
Elad Gil: No, it was one of those moments where you're like, wow, this really happens. You can go from eight people to thousands over a couple of years and build a substantial institution that's going to survive potentially generationally.
That's very powerful.
Geoff Ralston: Elad, thank you very much.
Elad Gil: Ah, thanks for having me.