Reid Hoffman at Startup School SV 2016

[Applause]

So, uh, up next needs no introduction. I'll give a very quick one. Reed Hoffman, uh, has been in—yeah, please do—round of applause! You know what it sounds like; you all know who he is. I'll skip the introduction.

All right, for the first question, um, tell us something you believe to be true that very few people agree with you on.

Well, uh, this room, um, uh, actually may actually agree with this, but I think this is—and when you count the globe, that question about the globe, not the globe—that question always has to be indexed against the population. But what I would say is that, uh, we have an extremely high likelihood that within the next 1500 years that we create another cognitive species, e.g. a species with a similar kind of or better cognitive abilities than we do. People in this room normally tend to think that's artificial intelligence. But I actually think it's a jump ball between that and a different version of the Homo sapiens. Like, we're in actually a relatively interesting and unique period by which Homo sapiens is the only, uh, part of the Homo genus because there was Florensis and Neanderthal and so forth. And when you consider the longer time frame, it's actually 40, 50,000 years ago is actually not that long ago.

So, um, I think we're going to do that, and it's conditioned upon, uh, the particular breakthroughs in AI, which I think there is still some invention and magic to get a truly generalized AI, which may or may not, you know, happen quickly. If it does, then AI is likely, and if it doesn't, then the Homo genus is likely.

Just for fun, uh, and you don't have to answer this if you don't want, but given that you made the point about indexing that question against the rest of the world and then many people in this room may agree with that, although I would still bet most don't, um, what is something that you believe that very few people in this room would agree with you on?

Yeah, um, let's see in this room, uh, okay, trying to index that. Um, I guess, uh, well, this might be they hadn't thought about it. Um, I think that one of the things that's actually really important for inventing products is a fairly deep sense of a theory of human nature and humanity. And so I think that anyone who is inventing the kinds of products that we typically do, you should actually, in fact, be able to articulate a relatively robust theory about what is human nature, what is humanity like now, and where is it going, and how does your product or service fit into that.

And that may not be so much they disagree as actually most people I talk to can't articulate a particular robust theory of human nature. And that's one of the reasons why, years back, as you know, I started saying this thing: I invest in one or more of the seven deadly sins because I was trying to get people to think about what is actually, in fact, a kind of common human psychology. And if you're trying to create a mass-market consumer application, what are the parts of how what it is to be human that you're actually triggering, responding to, serving, etc.?

That's a great answer! I, uh, that most people in the room don't think about that or don't agree it's important.

Um, so I want to talk about three main topics here. Uh, one is not the mistakes that people make when they start a company but the great things people do. Everyone worries about trying to avoid mistakes; that's actually not enough. You also have to do one thing really right. The second is how to raise money from Greylock, and the third is how to scale up a company and when to start that.

So, so let's start with the first. Um, what are the things that you see entrepreneurs do at the very beginning that are the opposite of mistakes that lead them to great results?

So I think one of the—let's see, uh, we could probably spend the whole half hour entirely on this, but I'll try to do so. I gave you all three questions up front so you could allocate time.

Yes, um, so one thing is build the strongest possible network around you, right? And, uh, and every single thing you're doing, whether it's hiring, whether it's getting investors, whether it's, uh, who you're talking to about your company, obviously YC is an opportunity to generate a really strong network.

Uh, and I think, uh, and in specific, in terms of, it's not just like what do you think of my idea. That's usually a weak question. How can my idea be improved? If my idea were to fail or my plan were to fail or my strategy fails, why do you think it would fail? Right?

Asking those kinds of questions, I think another thing is to actually think about one of the challenges is it's awesome that we have, um, gotten to such a massive entrepreneurial environment here. But, uh, for example, breaking through the noise is really tricky. So you actually have to really test and think about whether or not your idea will be sufficient to break through the noise.

And so folks who really think about, like, some way that they can either, you know, grow on a particular platform or have a particular way that they can actually get good growth into their product is, you know, particularly someone that sometimes—that's virality; sometimes SEO, sometimes other things. What are the early signs that an idea or a particular product may be able to break through the noise? What do you look for when you're evaluating that? Because I think that is this critical thing, but it's so hard to really know.

Um, well, uh, obviously, which, like, you know, okay, the obvious brain-dead thing that everyone knows is once you see traction, you still have some questions: how far does traction go? But that actually answers a bunch of those because you can study the actual, like, okay, are people engaging? Are they re-engaging? How do they discover it? And you can, you can, you can per thing, you can deconstruct the answer to your question.

The real challenging one is when you have the theories beforehand, and the theories beforehand, a little bit of that comes back to what I was saying in terms of a theory of human nature. Because you say, okay, why is it, if it's viral, why will the virality spread? Some of that's technique, but some of that's also—it'll trigger a funny bone in some way.

Now, I wouldn't have looked at Pokémon Go and said, well, actually, in fact, this nostalgia thing will create a craze. I would have missed that, right? But I learned, I, I, I, along with lots of other people, I downloaded and played with it immediately. Say, okay, how do I learn that pattern? What was the way that was fitting into people's identity? Why was this the thing that was kind of, uh, creating something that literally spread through essentially word of mouth?

Did you get really into it?

Uh, well, uh, uh, let me put it this way. I, uh, it became a habitual thing that I would do during an Uber ride. I have no shame; I'll just admit I was one of those people for a couple of days. Like three in the morning, my brothers and I would like be out driving around Golden Gate Park trying to catch Pokémon. Horrible.

Yeah, no, for me, it was when I got an Uber, then I pulled it up, and I go, because you know he had the advantage of someone else is driving you, yeah, you can go through all the Pokéstops and everything else at a reasonably fast clip.

Um, and so the other thing is, is, is, is, is there a principally pre-traction, actually limited edition consumer? Enterprise in consumer, the principal thing is there's actually an interestingly unique theory—there's some kind of theory together with some, like, understanding about how distribution works that you're applying. Like you say, okay, I actually think people will send out, you know, SMS messages, or they'll send out emails, or this will get indexed and searched the following way, or people will actually, in fact, do the following activities, which will cause other people to discover this, you know, app, site, service.

Um, and then that's a coherent theory, and that can sometimes be sufficient.

Uh, for doing it on the enterprise side, actually, sometimes clear mastery of traditional enterprise techniques is what matters, right? Because actually, in fact, if you have a relatively high—a good product in terms of what you could sell, you could measure it, as a result, you know what your sales channel looks like; a sales process can work effectively. But those two things and the ability to pull them off are actually absolutely essential to the things that you do right.

Because the most classic problem, which you've seen—and I know you're talking about the successes here—but the classic problem is, I've got a cool idea, or someone doesn't have a unique product yet, and that's radically insufficient. You're making it just lucky whether or not, actually, in fact, the distribution works. And these companies only grow to enough throw weight if they get to a point where they have enough distribution.

So roughly speaking, one of the rules of thumb that I use is if you don't, if, if you don't see a path by which you're getting to 100 million a year in revenue and then have good growth after that, you're not going to be a stand—you have zero chance of being a standalone tech company.

Shall I continue on the—maybe one more? But then I do want to make sure we move on.

I was just thinking about that. Um, so then the other things I think to do right are, um, I think the key things are, people who, when you're doing this—this is the fail-fast point, which is you're looking for the earliest possible proof to whether or not you're on the right track. It isn't that you're trying to fail, but you're actually hunting for the quickest end.

And, and, like sometimes even argument evidence. Like, for example, when I was literally having the idea for LinkedIn, I was going around and my friends—going, okay, why won't this work? And the common pattern I was hearing is because in a network, there's no value until you have a million people, so no one's going to invite for the first million people.

And the fact that I was—that I knew that that was the thing to test—that was the thing that I was essentially working on from day one to get to the minimum viable product in order to do that.

Do you think that people give up too early related to that, or, or at least they're looking at the wrong metric?

One of the things I see great founders do really well is not give up too early, and it sounds simple, but in this kind of culture of, you know, try something; if it doesn't work, fail fast, stop. I, I do think you can lose a lot of great companies.

Uh, so definitely you can. And it isn't to try to, like, for example, when I'm, uh, talking to an entrepreneur the first time, one of the things I test for is kind of what I what I call flexible persistence, which is they'll both listen and hear to the things I'm saying, questions, and they will, uh, adopt, they'll, they'll think about it; they'll respond; they may be willing to consider the alternate. But then they also have conviction in their point of view. They have a theory—a theory of humanity, a theory of distribution, a theory of product that they go, look, I've really thought through it, and I—there's some really good stuff here, and this is why it is.

And if you don't have both, you almost certainly fail. And part of that is people say, I'll just pivot my way there, and like, no, no, no. Have a deeply thought theory about what you're heading for. And so plan B frequently is not, oh, now I'm doing something entirely different. Usually plans B are slight variations in tests.

So if you list your hypotheses about what it is you think you're going to—why you're going to win at the startup game, then you start testing them. Part of Plan B is to say, well, maybe hypothesis 3 isn't right, but 3 prime is right and I can just shift a little bit of what I'm doing this way, and then it can work.

Um, and so, uh, like, for example, like one of the thoughts that we had very early in LinkedIn is even though we made a very heavy bet on individuals, would we need to shift to enterprising companies and groups as a possibility to make it work? It didn't turn out that we needed to do that, but that was a, okay, as a plan B.

And I think that staying persistent ultimately really—try it. Like you want two or three points reducing your confidence before you really decide to pivot.

So in the interest of time, let's move on to raising money. Um, if I want to come raise money from Greylock, uh, what can I do to be successful? What is step one, and what is the last step before the check and everything in between?

So, um, first—and this is true for, you know, all top-tier VCs—you’re much stronger off getting a referral, right?

So not Y Combinator, though?

Um, well, it can be a referral from one of the partners at Y Combinator, but it's, it's—the here's why you should pay attention to this: I probably get somewhere between 30 and 50 unsolicited decks in my email every working day, right? And so it's kind of like, okay, how many of those have you ever funded?

Uh, none that I'm aware of.

Okay, right? I mean, sometimes—okay, so the intro is really important. The intro is really important.

Um, and who counts as a good intro to you?

Um, uh, counts as a good intro is someone that I, T.R. that I, uh, know and trust and respect, right? So you have to guess it to some degree, but you know—and here I'll throw you under the bus—Sam counts, right?

Uh, and so, um, that's one. And, uh, the second is, uh, you know, look again; think of there's tons and tons of startups, and your average, every venture capitalist looks at like 600 deals a year and funds between zero and two of them, which means you're looking for something that is, uh, a little like your opening question. You're looking for something that's seriously unique, right? You're looking for something that says, look, I, I, uh, it doesn't have to be the only thing in its field, although then if it's a bunch of things in its field, why will this one be the one that will break through? Why will this one be the one that creates the interesting company?

Are you about to say something?

No? Okay, uh, and so that's part of the reason why what I suggest to people who want to come talk to us is to say, well, what's the thing that either has suddenly opened up or the thing that makes your thing going to be very big? When you realize there are thousands of startups now, uh, which is great, but, you know, given you can fund between zero and two, and so, for example, um, you know, one of the ones I did at the beginning of the year is one called Convoy, which is essentially Uber for regional trucking.

And actually, that one took a bunch of work for me because there were a bunch of people doing that, right? And I had to look through to see which one was the one to do because we can only do one.

And what did that founder do to convince you that that was the one?

Uh, well, uh, first I came in through a great referral, which is Ali Portovi, right, who had done a whole bunch of work and had chosen us. Uh, second is in the discussion of, like, literally in the very first discussion, I got a very good sense that the, uh, that the founders—Dan and Grant—understood the game in front of them. They understood what the issues were; they had thought about, like, for example, when I pushed them about what the, uh, comparisons and thoughts with Uber were; uh, what was unique about this space; uh, why actually, in fact, this is a very different network than an Uber network; um, uh, what, uh, the challenges might be.

And they had actually thought about all of it. It doesn't mean they have perfect answers and doesn't change, but they had thought about it; the way they were developing their product was in line with that thinking.

Uh, and they had a story, and this is one of the things that's super important, is like, what is—it's not to say you say, I'm guaranteed—what's uncredible? You say, I'm guaranteed to be successful. I'm coming in as a seed series A, and I'm guaranteed to be successful. And that either says to me that you're crazy, right? Because you don't actually understand the risk, or you're lying to me, and both of those are not great outcomes from a viewpoint of a partnership.

Uh, and so it's much better to say, here, actually, in fact, what I see the game in front of me. Here's what I think the risks are; here's how I'm addressing the risk; here's how I'm measuring them; this is how I'm doing it. And by the way, I have confidence that I'm going to succeed.

Great, perfect. Uh, and so that ability to understand what the game is in front of you, because then you'll parse it the right way.

Um, I think it's very important to have an ability to have, uh, structured conversations early. It doesn't necessarily mean you haven't had a deck; some conversations I have are literally just here's my set of beliefs that make me believe that starts a good idea. It's like the list of the investment theses—the kinds of things that would say why it is this would be, you know, five to seven years from now this would be Airbnb; this would be something that you would go, oh my gosh, this is a world-changing company.

Do you recommend a deck?

Um, I reckon deck is a set of PowerPoints. Oh, that's—yep.

Uh, generally speaking, I think it's good to have at least a lightweight one. Um, and that's lightweight; it's like 10 to 15 slides. Um, uh, some basic—and could you just go quickly through the what you expect to see in those slides? Like what are the main five or six critical topics?

So, um, I presume you guys point people to the LinkedIn series B deck that I published?

We have, but maybe not this audience.

So republished his, uh, LinkedIn series B, uh, deck.

Yeah, it's online, and you should check it out.

Yeah, so this is roughly speaking—you should open with here's my—here are the hypotheses that back the investment thesis, and that should be probably no more than seven bullets. Usually, it's more than three, but something—and that could be eight, if it had to be.

And then your slides are the backdrop of those bullets. Why is it you think you'll win at that, right? So, for example, you know, there is actually space for professional networks, separate from a social one. It'll stall this thing in people's lives.

Um, you know, that sort of thing is the kinds of things that are in it. And it can get distributed through virality, right? And we can solve the viral problem.

Then I think you need to have something about, like, okay, what's the mature business sketch look like? It doesn't mean like, oh, here's my fake Excel revenue projections, which anyone with a brain knows you can make these things look like anything. But it's like—it's a—oh, well, charging people like this, this is why the product will be good. This is why we'll have good margins; this is why we'll have a defensible position; this is why we'll be at a big size; and this is why we'll be growing, right?

And so, for example, that was one of the reasons why the Airbnb pitch was really good from the beginning is like, well, we're eBay for space, and we're starting with travel, and travel itself is huge, and this is what the dynamics work in terms of opening up liquidity between host and traveler, and you're like, okay, I get it, right? That was one of the pitches literally two minutes in—you're like, okay, I understand, right? And by the way, you understood the risk then too: the risks were like, well, actually, in fact, people aren't used to this, and opening your house is a trust and safety issue, and is there a regulation issue, and, you know, all these kinds of things.

But those are then—these are known risks that they were then working against. To claim zero competition is usually not credible. Usually, it's: here's why the competition has a very different angle of attack than I do, and this is why my angle of attack would work.

And then, you know, kind of more or less, for example, an early stage consumer companies—why it is I think distribution is going to work.

And then can you give us a little bit of sense of how you think about terms? What terms are important to a venture firm, to an entrepreneur, and what a series A might look like?

So, um, to a venture firm, part of the key thing is—so there's a big difference between angels and ventures. Angels can invest in a wide variety of things; they don't actually commit to doing that much. They do commit to doing a little bit of help, but not that much help.

Uh, Venture folks commit to, I'm with the company more or less until it gets deported or not. I can only do—really, it's very rare that you can do—that a firm can do ones that are competitive, so I can only do one of the spaces, and I'm going to throw a bunch of resources behind it.

Uh, what—you know, um, you know, injuries and Horowitz, we, other folks have, uh, like great recruiting practices, a number of things like that as a way of doing it. Like a Greylock thing is we help our portfolio companies find and engineer every other—every alternative day, right, during the year. And that's, you know, that's a pretty good flow.

And so, um, typically, uh, the best way to do it is very clean terms, have, uh, either one or at most two leads. Two leads tends to be more on the enterprise side than the consumer—somebody who is really going to work, roll up their sleeves, and work with you. That usually gets for venture to a certain percentage, um, you know typically—and there's a—there's actually—not—it's not an artificial number. Most VCs target a 20% or greater ownership because when they look at, when they do their models of what their baskets of ownership are on exits, if they have north of 20, then they can make their fund multiples work, and so that's the reason why that's the target.

What is this of—a venture firm has a 500 million dollar fund; what is the return on that fund that a firm would like to hit to be really great?

Uh, typically, you would want to have two and a half billion dollars of return gross, so 5x gross.

That's before the GPs take their carry-ins.

Yes, that's—that's a reasonable fund. That isn't the, you know, uh, 10-plus X fund that, you know, all the top tier firms have hit at some point. There are funds where you get that, but that's the—oh, that was a credible fund.

Great! So, uh, now I want to move on to the last topic. When, uh, you've, you know, done things well, you've made a good company, you've got a product people like, you've raised the series A, uh, and now you transition to scale up.

Um, you've popularized that term; what does that mean, and how do you know when it's time to do it?

So, um, we taught a class that's online; you can find it at Greylock and some other places, I think on iTunes, called Blitzscaling. Uh, Sam was our opener; I was, uh, because we were essentially actually building on the class that Sam and Seltham done one or two years earlier. I can't remember exactly what it was.

Uh, and, uh, the key thing is, is, um, first mover means first to scale. And most often, both within the consumer and enterprise, the first to scale is the person who wins, who owns the market, who sets the terms, uh, who gets the best advantages from the capital market, who gets the best advantages of the talent market, is known to customers, etc.

It isn't necessarily the first out of the gate; it's the first to scale. And so what that means, uh, pretty logically, is your question would be—is you know, typically, comfort for investors, for entrepreneurs, and say, well, let's completely resolve product market fit and then scale. If you can do that, and because you're doing it sufficiently by yourself—you don't have aggressive competitors—great! That's the right way to do it.

You can more or less understand here at my distribution model is working, my engagement model is working, I have a revenue model that at least is basically working and maybe can be improved, and now what I'm just doing is figuring out how to accelerate that as much as possible.

More often than not, in Silicon Valley, you've got part of that story right, but you have intense enough competition that you have to make a decision to scale sooner than that. And part of calling it Blitzscaling—like, one of the key things that if you actually look around at part of what's happened in the last 20 years is that actually, in fact, startups will spend a bunch of capital at less efficiency than you would typically think because, like, perfect operational efficiency—we know exactly what our model is; we're rolling it out in order to get to scale fast.

And obviously, Uber is the canonical modern example of this. And the decision to do that is partially a question of both kind of offense and defense. Offense in terms of, well, I really need to establish myself, I need to get to critical mass in the network in order to have the network effects.

It's kind of like early LinkedIn, although we eked it out in terms of—in terms we didn't actually Blitz to the first million; we kind of compounded until we got there, so we were doing the other strategy because there was no one doing our strategy all credibly close to us. So we could take time, or defensive, which is you're worried that someone else is going to get there first, and so you actually have to put in the afterburners and then try to adjust as you go.

And that's the combination of offensive and defensive reasons is part of the decisioning that you're making about what, about how and when to do it. Now, the other thing that's, uh, kind of classic, and this is, you know, kind of a line that probably you haven't seen—that's in the new draft of the book has built to scale—which is when you begin to look at each different key element that can make your business scale quickly, it's like customer acquisition, revenue model, support model, servicing model—how in each of these things, growing your company—how in each of these things can you actually, in fact, get to a global scale relatively quickly?

Now, it doesn't make sense to obsess with that if—before you have, when you have zero idea of product market fit, it's only—but it is important to start thinking about it because the thing that makes the really big businesses are the ones that can scale.

Which direction do you see more entrepreneurs making this mistake? In 2016, are they waiting too long to start scaling, or are they doing it too quickly?

Foreign, it's both. It's not—it's—I don't think it's kind of like, oh, they're all like—both mistakes are made.

Um, and you know, the wait too long is kind of the question of, uh, I'd really like to prove it out more, I'm not really that worried about competition, um, I'd like to get complete certainty about, about what I'm doing.

The going too fast is, um, I haven't—I haven't tested some key elements that are really important, and then I get way over my skis, and it's hard to correct because at that point, for example, you've got investor expectations, you've raised a bunch of capital, uh, you need to pivot—it’s actually difficult to—to set back to, etc.

Have you found any successful way to do that if you do get ahead of your skis because you're scared of a competitor but you find out the product is not quite working yet, and you do need to pull back but you've already raised this mega round on huge projections? Have you ever seen a way to make that work?

One doesn't immediately come to mind. Although the advice that I would give would be this: this is one of the reasons why fundraising is not just the capital thing; it's a partnership question. And if you actually have a good partnership with the folks, and you’re kind of saying, look, we're sharing risk, we're sharing analysis—like, for example, when you've got the—when you've raised money, you're not saying, oh, it's guaranteed; it's perfect; it's all in the bag, and it’s all happening because then, by the way, very naturally, your partners get very grumpy with you, right, when it's not that.

But so look, we're trying this, we have good confidential work, but not perfect, we're going to be a good partner with you. And then if you have that, then it's easy enough to go back and say, look, we need to really pivot, and we need to do something different.

I'm sure there are examples, but it's not, it's not immediately apparent to me.

It's hard for me to think of one too.

Uh, one of the things that I have seen our company struggle with the most when it comes to scaling quickly is how to identify enough good people to hire, and related to that, how to keep the culture as you double in size in a short period of time, um, and then double again and again and again.

So, uh, to that joint question of how do you identify really good people, and how do you bring them together in a way that preserves the company culture that got you here in the first place, um, what can you share with us about that?

So, I'll start with the culture, and then I'll go to the hiring. So culture—part of the really key thing in culture is to—and this is actually, um, like I—I interviewed Reed Hastings about that on Blitzscaling; he's actually a very good cultural person. Part of the—how he created the Netflix deck and how they got to thinking about it was actually, in fact, uh, fairly key, and that was kind of in the vein of they were discovering that they'd hire really good people who are still our players but weren't cultural fits, and they would bounce out.

And so the reason they codified the deck and published it was to make sure they weren't going to have that churn problem where people understood this is what we stand for. Now part of the reason why codifying it is useful is because part of how you scale a culture is that everyone is keeping each other accountable. It's not hierarchical; it's not just, oh, whatever the two or three founders say or one of the three founders say, that's—that's it and whatever changes. It's no, this is who we are; this is how we play; this is what we're doing, and we're all holding each other accountable.

And so, for example, one of the big ones for, uh, for Reed Hastings was, and is, is we're not a family or a sports team, right? And so it's not lifetime loyalty; it's we're all performing. And when you're doing your, your, your kind of reviews of your team, if you wouldn’t fight to keep the person, you don't just go, oh, they're perfectly good. But if you wouldn't fight to keep them, you give them a severance package, and you move on because that's what your target is in terms of what you're doing.

And then there's other elements of culture, like do we really value design? Do we really value, um, an ability to, you know, like for example, in the LinkedIn cases, the individual members first, not the people who are paying us money. It's each individual member, and that's an important part of the kind of culture and how we talk about, like, how we make product decisions.

So that's the cultural part. Then the hiring part, um, one of the key things—it's—you can get—there's—you have to correct for some downsides of this, but obviously using your network is super important because you can get trusted connections in that can sometimes be, you know, a little too, um, a little too much all of the same. Like one of the problems is sometimes you're bad on diversity because of this. It's like, oh, well, you know, we're a bunch of men, so we get a bunch of other men, and it's like, no, no, no, actually diversity makes you stronger, so you should be doing that.

So networks can have that bias problem that you have to—you have to correct for, uh, systemically, but you can still use the network's hunting to make sure you get the appropriate diversity of skills and perspectives and awareness in your—in your company.

Um, then the next thing is, is reference checking, and to give you a specific example of something that I do is usually when I'm considering somebody—if I can, if I know a couple people who know this person even before I decide to reach out, uh, make a connection, I'll frequently do a very light reference check because reference checks tell you a lot more. Peop—everyone's very good at—or most people are very good at seeming very coherent and reasonable in a couple of interactions.

The real thing you want is how does the person work over weeks and months and years, right, in the trenches, and references are very good for that. And of course, references from people you trust or people they trust is part of the—the right way to do that.

And the way that I frequently do that is I'll send a note to people saying, uh, Sam: rate them one to ten. And the reason I do one to ten is because, uh, the—the weak answers are usually seven, seven and a half—that's kind of the, oh, I want to say bad things about Sam, but I'm not going to—and it's an email, and it's totally fine.

And what you're looking for is a combinations of eights and nines, and when I get a ten, I usually write back and say, oh really? Like, Sam's one of the best person—best people you ever worked with, right? Just to make sure that this is, you know, this is kind of real.

And then when you have a blend of them, that can give you a sense of whether or not I should dig in more to this person. So networks, uh, referencing, pre-referencing, and then definitely doing deep referencing.

Now, one of the last things I'd actually say on hiring is this—is one of the ones that I find particularly entertaining is, um, actually spend a bunch of time with the person. So if you’re at the level of importance and hierarchy in the company—like, don't think two or three interviews and you're done when you're hiring.

Uh, like for example, when I was hiring Jeff, I spent probably about 40 hours with Jeff in discussion before we got to, okay, this is going to work in addition to all the referencing. It was super important and important to get right. And executives, it can easily be 20 hours, right? Sometimes it can be 10 to 15, because it's like, okay, do I—do I get a sense of how we play together, not just the references anywhere else, but, but, but how are problem-solving techniques or our values, is the way that we—the things we want to accomplish, are they sufficiently aligned, etc.?

Great! Well, thank you very much for spending the afternoon with us, and, uh, appreciate all the thoughts. Good luck!

Thank you.