1 00:00:01,097 --> 00:00:02,841 Part two of Culture and Team, and 2 00:00:02,841 --> 00:00:05,708 we have Ben Silverman, the founder of Pinterest, and 3 00:00:05,708 --> 00:00:08,530 John and Patrick Collison, the founders of Stripe. 4 00:00:10,510 --> 00:00:13,300 Founders that have obviously sort of, some of the best in 5 00:00:13,300 --> 00:00:15,190 the world at thinking about culture and 6 00:00:15,190 --> 00:00:15,940 how they build teams. 7 00:00:15,940 --> 00:00:18,400 So, there's three areas that we're gonna cover today. 8 00:00:18,400 --> 00:00:20,690 One will just be sort of general thoughts on 9 00:00:20,690 --> 00:00:22,980 culture as a follow up to the last lecture. 10 00:00:22,980 --> 00:00:25,580 And then we're really gonna dig into what happens at 11 00:00:25,580 --> 00:00:27,120 the founding of these companies. 12 00:00:27,120 --> 00:00:28,790 And building out the early team. 13 00:00:28,790 --> 00:00:32,530 And then how that changes and evolves as these 14 00:00:32,530 --> 00:00:35,940 guys have scaled their companies up to 100 plus. 15 00:00:35,940 --> 00:00:37,270 I don't even know how many people you have now, 16 00:00:37,270 --> 00:00:38,380 but quite a lot. 17 00:00:38,380 --> 00:00:39,520 Very large organizations and 18 00:00:39,520 --> 00:00:41,240 how you adapt these principles of culture. 19 00:00:42,910 --> 00:00:46,205 But to start off I just wanna ask a very 20 00:00:46,205 --> 00:00:47,970 open-ended question which is, 21 00:00:47,970 --> 00:00:51,010 what are the core pieces of culture that you found to 22 00:00:51,010 --> 00:00:52,951 be most important in building out your companies? 23 00:00:52,951 --> 00:00:56,720 >> So what are the most important parts? 24 00:00:58,130 --> 00:01:01,829 >> It's on. >> Oh, it's on. 25 00:01:01,829 --> 00:01:02,769 Yeah I mean I think for 26 00:01:02,769 --> 00:01:04,910 us, like we think about on a few dimensions. 27 00:01:04,910 --> 00:01:05,780 One is like who do we hire, 28 00:01:05,780 --> 00:01:07,180 and what do those people value? 29 00:01:08,710 --> 00:01:11,950 Two is what do we do every day, like why do we do it? 30 00:01:13,180 --> 00:01:15,460 Three is what do we choose to communicate? 31 00:01:15,460 --> 00:01:17,480 And then I think the fourth is what we 32 00:01:17,480 --> 00:01:18,460 choose to celebrate. 33 00:01:18,460 --> 00:01:19,220 I guess the converse of 34 00:01:19,220 --> 00:01:20,590 that is like what you choose to punish. 35 00:01:20,590 --> 00:01:23,880 But in general I think running a company based on 36 00:01:23,880 --> 00:01:24,490 what you celebrate, 37 00:01:24,490 --> 00:01:26,360 is more exciting than what you punish. 38 00:01:26,360 --> 00:01:29,040 But I think those four things kind of make up 39 00:01:29,040 --> 00:01:30,040 the bulk of it for us. 40 00:01:35,796 --> 00:01:38,666 >> We've placed a large emphasis on, as Stripe has 41 00:01:38,666 --> 00:01:41,536 grown and probably more than other companies is, 42 00:01:41,536 --> 00:01:43,560 transparency internally. 43 00:01:43,560 --> 00:01:44,940 And I think it's been something that's been 44 00:01:44,940 --> 00:01:45,505 really valuable for 45 00:01:45,505 --> 00:01:48,460 Stripe, and also a little bit misunderstood. 46 00:01:49,590 --> 00:01:53,740 All the things people talk about like hiring really 47 00:01:53,740 --> 00:01:57,070 great people, or giving them a huge amount of leverage. 48 00:01:57,070 --> 00:02:00,220 Transparency for us plays into that. 49 00:02:00,220 --> 00:02:03,335 We think that, if you are aligned at a high 50 00:02:03,335 --> 00:02:07,340 level about what Stripe is doing, if everyone really 51 00:02:07,340 --> 00:02:11,078 believes in the mission, and then if everyone has 52 00:02:11,078 --> 00:02:14,639 really good access to information, and kind of 53 00:02:14,639 --> 00:02:18,401 has a good picture of the current state of Stripe. 54 00:02:18,401 --> 00:02:21,175 Then that gets you a huge amount of the way there in 55 00:02:21,175 --> 00:02:24,230 terms of working productively together. 56 00:02:24,230 --> 00:02:26,760 And it kind of forgives a lot of the other things that 57 00:02:26,760 --> 00:02:29,790 tend to break as you, as you grow a start-up. 58 00:02:29,790 --> 00:02:33,090 As we've grown, we started off two people. 59 00:02:33,090 --> 00:02:33,980 We're now over 170 people. 60 00:02:33,980 --> 00:02:37,160 We've put a lot of thought into the tooling that 61 00:02:37,160 --> 00:02:41,170 goes around transparency, because at 170 people, 62 00:02:41,170 --> 00:02:44,100 there is so much information being produced, 63 00:02:44,100 --> 00:02:46,420 that you can't just consume it all as a fire hose. 64 00:02:46,420 --> 00:02:47,300 And so 65 00:02:47,300 --> 00:02:49,820 how we use slack, how we use email, things like that. 66 00:02:49,820 --> 00:02:51,030 We can go into it more later. 67 00:02:51,030 --> 00:02:52,420 But I think that's one of 68 00:02:52,420 --> 00:02:54,060 the core things that's helped us work well. 69 00:02:55,300 --> 00:02:58,180 >> I think culture to some degree is basically kind of 70 00:02:58,180 --> 00:03:00,790 the resolution to a bandwidth problem. 71 00:03:00,790 --> 00:03:04,480 In the sense that, maybe when you start out working 72 00:03:04,480 --> 00:03:06,270 on something, you're sort of coding all the time, but you 73 00:03:06,270 --> 00:03:08,580 can't code all the things that you think the product 74 00:03:08,580 --> 00:03:10,440 might need, or the company might need, or whatever, and 75 00:03:10,440 --> 00:03:13,150 you so you decide to work with more coders, right? 76 00:03:15,420 --> 00:03:17,870 And so the organization gets larger. 77 00:03:17,870 --> 00:03:19,590 And maybe, in some idealized world, 78 00:03:19,590 --> 00:03:21,750 I don't think this actually true, but kind of ideally 79 00:03:21,750 --> 00:03:23,950 you could be involved in every single decision, and 80 00:03:23,950 --> 00:03:26,370 every single sort of moment of the company, and 81 00:03:26,370 --> 00:03:28,820 everything that happens, but obviously you can't, or 82 00:03:28,820 --> 00:03:30,260 maybe you can if two people. 83 00:03:30,260 --> 00:03:32,440 But you certainly can't at even like five or 84 00:03:32,440 --> 00:03:34,530 ten kind of that point comes very quickly, 85 00:03:34,530 --> 00:03:36,710 then by the time you're 50 it's completely hopeless. 86 00:03:36,710 --> 00:03:39,570 And so culture is kind of how you kind of, 87 00:03:39,570 --> 00:03:44,110 what the strands are that you sort of want to have, 88 00:03:44,110 --> 00:03:46,630 the invariance that you want to kind of maintain, 89 00:03:47,670 --> 00:03:50,840 as you can get specifically involved in sort of 90 00:03:50,840 --> 00:03:53,338 fewer and fewer decisions over time. 91 00:03:53,338 --> 00:03:56,192 And I think when you think about it that way it, 92 00:03:56,192 --> 00:03:59,680 maybe its kind of importance becomes sort of 93 00:03:59,680 --> 00:04:00,610 self evident, right? 94 00:04:00,610 --> 00:04:03,030 Because again, like the fraction of things you 95 00:04:03,030 --> 00:04:07,240 can be involved in directly is diminishing, I mean, 96 00:04:07,240 --> 00:04:08,130 almost exponentially, 97 00:04:08,130 --> 00:04:10,430 sort of assuming your head count growth sort of is 98 00:04:10,430 --> 00:04:14,290 on a curve that looks like one of the great companies. 99 00:04:14,290 --> 00:04:17,700 And yeah that's super important. 100 00:04:17,700 --> 00:04:20,648 And again, it manifests itself in a bunch of 101 00:04:20,648 --> 00:04:21,850 different ways. 102 00:04:21,850 --> 00:04:25,680 Like for example, in hiring, I think a large part of 103 00:04:25,680 --> 00:04:28,810 the reason why maybe the first ten people you hire, 104 00:04:28,810 --> 00:04:29,920 what kind of goes to ship, 105 00:04:29,920 --> 00:04:32,760 decisions are so important is because you're not just 106 00:04:32,760 --> 00:04:34,320 hiring those first ten people. 107 00:04:34,320 --> 00:04:36,330 You're actually kind of hiring 100 people. 108 00:04:36,330 --> 00:04:37,700 Because you should think of kind of 109 00:04:37,700 --> 00:04:39,790 each one of those people as bringing along sort of 110 00:04:39,790 --> 00:04:41,580 another ten people with them, and 111 00:04:41,580 --> 00:04:45,830 sort of figuring out exactly what 90 people, you would 112 00:04:45,830 --> 00:04:48,070 like those first ten people to bring along is obviously 113 00:04:49,340 --> 00:04:53,360 gonna be quite consequential for your company. 114 00:04:53,360 --> 00:04:55,940 But really briefly I think it's largely about sort of 115 00:04:57,520 --> 00:04:58,020 abstraction. 116 00:04:59,280 --> 00:05:00,780 >> So one thing that a lot of speakers in this 117 00:05:00,780 --> 00:05:03,910 class have touched on is how hiring those first ten 118 00:05:03,910 --> 00:05:05,670 employees, if you don't get that right, 119 00:05:05,670 --> 00:05:07,978 the company basically will never recover but 120 00:05:07,978 --> 00:05:10,180 no one has talked about how to do that so. 121 00:05:11,180 --> 00:05:14,000 What have the three of you looked for 122 00:05:15,010 --> 00:05:17,190 when you've hired these initial employees, 123 00:05:17,190 --> 00:05:19,800 to get the culture of the company right? 124 00:05:19,800 --> 00:05:20,835 How, how have you found them, and 125 00:05:20,835 --> 00:05:21,630 what have you looked for? 126 00:05:24,080 --> 00:05:24,610 >> Sure. So, 127 00:05:24,610 --> 00:05:28,542 I guess this answer is different for every 128 00:05:28,542 --> 00:05:32,640 company and I'll say for us it was very inductive. 129 00:05:32,640 --> 00:05:35,390 So I literally looked for people that I wanted to work 130 00:05:35,390 --> 00:05:36,960 with and that I thought were talented. 131 00:05:38,230 --> 00:05:39,960 I think, I've read all these books about culture, 132 00:05:39,960 --> 00:05:41,130 because when I don't know how to do something, 133 00:05:41,130 --> 00:05:42,730 I first go read things and 134 00:05:42,730 --> 00:05:43,920 everyone has all these frameworks. 135 00:05:43,920 --> 00:05:47,090 And I think one bit, big misconception that someone 136 00:05:47,090 --> 00:05:49,270 said once is that people think culture is like 137 00:05:49,270 --> 00:05:51,480 architecture when it's a lot more like gardening. 138 00:05:51,480 --> 00:05:52,650 You know, you plant some seeds and 139 00:05:52,650 --> 00:05:54,050 then you pull out weeds that aren't working, 140 00:05:54,050 --> 00:05:55,570 and they sort of expand. 141 00:05:55,570 --> 00:05:58,650 So, when we first hired people we hired people that 142 00:05:58,650 --> 00:05:59,520 were like ourselves. 143 00:06:01,060 --> 00:06:02,030 I often looked at like three or 144 00:06:02,030 --> 00:06:04,460 four different things that I really valued in people. 145 00:06:04,460 --> 00:06:06,700 You know, I looked for people that worked hard and 146 00:06:06,700 --> 00:06:08,890 seemed high integrity and low ego. 147 00:06:08,890 --> 00:06:10,490 I looked for people that were creative, and 148 00:06:10,490 --> 00:06:12,170 I usually meant that they were really curious. 149 00:06:12,170 --> 00:06:14,430 They had all these different interests. 150 00:06:14,430 --> 00:06:17,120 Some of our first employees are probably some of 151 00:06:17,120 --> 00:06:19,280 the quirkiest people I've ever met. 152 00:06:19,280 --> 00:06:20,250 They were engineers but 153 00:06:20,250 --> 00:06:22,540 they also have all these crazy hobbies. 154 00:06:22,540 --> 00:06:24,390 Like one guy made his own board game 155 00:06:24,390 --> 00:06:26,240 with this elaborate set of rules. 156 00:06:26,240 --> 00:06:28,240 Another guy was really into magic tricks. 157 00:06:28,240 --> 00:06:31,210 And he had coded not only like this magic trick on 158 00:06:31,210 --> 00:06:32,000 the iPhone, but 159 00:06:32,000 --> 00:06:35,290 he had shot the production video with a preview. 160 00:06:35,290 --> 00:06:37,400 And I think that, that quirkiness has actually been 161 00:06:37,400 --> 00:06:38,390 a little bit of a calling card. 162 00:06:38,390 --> 00:06:40,770 And we find that really creative, quirky people that 163 00:06:40,770 --> 00:06:42,710 are excited about many disciplines, and are 164 00:06:42,710 --> 00:06:46,720 extraordinary at one tend to build really great products. 165 00:06:46,720 --> 00:06:49,220 They tend to be great at collaborating. 166 00:06:49,220 --> 00:06:50,610 Then the last thing is, we really look for 167 00:06:50,610 --> 00:06:51,390 people that wanted to, 168 00:06:51,390 --> 00:06:54,180 they just wanted to build something great, and 169 00:06:54,180 --> 00:06:56,720 they weren't arrogant about it, but they just felt like, 170 00:06:56,720 --> 00:06:58,190 it'd be really cool to take a risk and 171 00:06:58,190 --> 00:07:00,530 build something bigger than themselves. 172 00:07:01,670 --> 00:07:02,790 And that, in the beginning, is very, 173 00:07:02,790 --> 00:07:05,250 very easy to select for if you're in our situation. 174 00:07:05,250 --> 00:07:08,260 We had this horrible office, like, nobody got paid. 175 00:07:08,260 --> 00:07:10,530 So there was no external reason, other than being 176 00:07:10,530 --> 00:07:12,810 excited about building something to join. 177 00:07:12,810 --> 00:07:14,770 In fact, there was every reason not to. 178 00:07:14,770 --> 00:07:16,410 And that's something, looking back, 179 00:07:16,410 --> 00:07:17,170 I really, really value, 180 00:07:17,170 --> 00:07:19,080 because we always knew people were joining for 181 00:07:19,080 --> 00:07:20,275 the purest reasons, and in fact, 182 00:07:20,275 --> 00:07:23,680 were willing to forgo other great job opportunities, 183 00:07:23,680 --> 00:07:27,190 market salary, a clean office, 184 00:07:27,190 --> 00:07:29,540 good equipment just for the chance to work. 185 00:07:29,540 --> 00:07:32,540 So, to this day I think a lot of those traits have 186 00:07:32,540 --> 00:07:33,230 been seeded and 187 00:07:33,230 --> 00:07:35,050 are embedded in the folks that we look at now. 188 00:07:37,990 --> 00:07:39,400 >> Yeah the first ten hires are really 189 00:07:39,400 --> 00:07:43,950 hard because you're making these first ten 190 00:07:43,950 --> 00:07:46,500 hires at a point where no one's heard of this company, 191 00:07:46,500 --> 00:07:47,550 no one really wants to work for it. 192 00:07:47,550 --> 00:07:48,530 You're just these, 193 00:07:48,530 --> 00:07:51,140 like two weird people working on this weird idea. 194 00:07:51,140 --> 00:07:53,070 >> And like their friends are telling them not 195 00:07:53,070 --> 00:07:54,190 to join. 196 00:07:54,190 --> 00:07:57,320 For our second employee, I think 197 00:07:57,320 --> 00:07:59,800 he'd accepted the offer or he was just about to, and 198 00:07:59,800 --> 00:08:02,700 his best friends took him out the night before, and it 199 00:08:02,700 --> 00:08:06,596 was like a full on assault for, why you should not join 200 00:08:06,596 --> 00:08:09,250 this company, why this is ruining your life basically. 201 00:08:10,560 --> 00:08:13,720 And so anyway the guy subsequently 202 00:08:13,720 --> 00:08:14,750 continued to join. 203 00:08:14,750 --> 00:08:15,598 And actually one of 204 00:08:15,598 --> 00:08:17,771 those friends also now works at Stripe. 205 00:08:17,771 --> 00:08:20,067 But this is what you're up against. 206 00:08:20,067 --> 00:08:20,934 >> Yeah. 207 00:08:20,934 --> 00:08:24,362 And I mean it's also hard because no batch of 208 00:08:24,362 --> 00:08:27,790 ten people will have as great an influence on 209 00:08:27,790 --> 00:08:31,630 the company as those first ten people. 210 00:08:31,630 --> 00:08:34,640 I think everyone's impression of recruiting is, 211 00:08:34,640 --> 00:08:36,429 you open LinkedIn. 212 00:08:36,429 --> 00:08:38,049 It's sort of like ordering off the dollar menu. 213 00:08:38,049 --> 00:08:40,188 It's, I want that one, that one, and that one. 214 00:08:40,190 --> 00:08:42,543 And, and now you have some hires. 215 00:08:42,543 --> 00:08:46,606 Whereas at least for us, it was very much over a very 216 00:08:46,606 --> 00:08:50,103 long time period, talking people we knew or 217 00:08:50,103 --> 00:08:53,780 friends of friends into joining. 218 00:08:53,780 --> 00:08:56,880 We didn't have huge networks, Pat and 219 00:08:56,880 --> 00:08:58,100 I were both in college at the time. 220 00:08:58,100 --> 00:08:59,630 So there were no people that 221 00:08:59,630 --> 00:09:01,750 we'd really worked with to draw on, and so 222 00:09:01,750 --> 00:09:05,280 a lot of those early Stripes were people we had heard of, 223 00:09:05,280 --> 00:09:08,990 friends of friends, and the other interesting thing they 224 00:09:08,990 --> 00:09:11,665 all had in common is that they were all 225 00:09:11,665 --> 00:09:16,400 early in their career, or undervalued in some way. 226 00:09:16,400 --> 00:09:17,290 Cuz when you think about it, 227 00:09:17,290 --> 00:09:20,430 if someone is a known spectacular quantity, 228 00:09:20,430 --> 00:09:22,570 then they're probably working in a job and 229 00:09:22,570 --> 00:09:23,730 very happy with that. 230 00:09:23,730 --> 00:09:27,720 And so we have to try and find people who were, 231 00:09:27,720 --> 00:09:30,860 in the case of our designer that we hired, he was 18 and 232 00:09:30,860 --> 00:09:33,970 in high school and in Sweden at the time. 233 00:09:33,970 --> 00:09:34,820 In the case of our, 234 00:09:34,820 --> 00:09:39,040 our CTO, he was in college at the time. 235 00:09:39,040 --> 00:09:40,030 You know, a lot of these people, 236 00:09:40,030 --> 00:09:42,820 they were early on in their careers and 237 00:09:42,820 --> 00:09:46,920 the only way we could, you can relax when constrained. 238 00:09:48,020 --> 00:09:49,940 You can relax the fact that they're talented or 239 00:09:49,940 --> 00:09:52,212 relax that it's apparent that they're talented, 240 00:09:52,212 --> 00:09:56,240 and we, not consciously, but we relaxed the latter. 241 00:09:57,420 --> 00:10:03,320 >> Yeah, I think finding kind of people who are, or 242 00:10:03,320 --> 00:10:05,130 just think like a value investor. 243 00:10:05,130 --> 00:10:08,170 You're looking for the human capital that's 244 00:10:09,990 --> 00:10:11,900 significantly devalued by the market, you know? 245 00:10:11,900 --> 00:10:12,940 You probably shouldn't look to 246 00:10:12,940 --> 00:10:15,850 hire your brilliant friends at Facebook and Google or 247 00:10:15,850 --> 00:10:18,050 whatever, because they're already discovered. 248 00:10:18,050 --> 00:10:19,510 You know, if they're wanting to join that's great, 249 00:10:19,510 --> 00:10:21,919 but they're probably harder to convince. 250 00:10:23,492 --> 00:10:25,570 John and I spent a little a while yesterday 251 00:10:25,570 --> 00:10:28,790 afternoon sort of trying to figure out in retrospect, 252 00:10:28,790 --> 00:10:31,490 what kind of traits our first ten or 253 00:10:31,490 --> 00:10:34,320 so people had in common that we thought were significant. 254 00:10:34,320 --> 00:10:37,711 And you know, in general sort of in speaking about 255 00:10:37,711 --> 00:10:41,182 culture, I sort of want to caveat everything we say 256 00:10:41,182 --> 00:10:45,785 with, I that sort of advice is very limited experience, 257 00:10:45,785 --> 00:10:47,785 widely over extrapolated. 258 00:10:47,785 --> 00:10:50,690 And I think there's a lot of truth to that. 259 00:10:50,690 --> 00:10:54,130 But for our particular first ten people, the things we 260 00:10:54,130 --> 00:10:58,350 sort of figured out that seemed to be important were 261 00:10:58,350 --> 00:11:01,760 they were also very genuine and straight. 262 00:11:01,760 --> 00:11:03,900 I think that actually matters quite a lot in that 263 00:11:03,900 --> 00:11:06,880 sort of they're people that sort of that others want to 264 00:11:06,880 --> 00:11:10,120 work with, that they're people that others trust. 265 00:11:10,120 --> 00:11:11,810 They sort of have an intellectual honesty in 266 00:11:11,810 --> 00:11:13,630 how they approach problems and, and so forth. 267 00:11:14,960 --> 00:11:15,650 They were people who 268 00:11:15,650 --> 00:11:18,490 really liked getting things finished. 269 00:11:18,490 --> 00:11:19,290 There's a lot of people who 270 00:11:19,290 --> 00:11:21,400 are really excited about tons of things. 271 00:11:21,400 --> 00:11:23,630 Only a subset of those are actually excited about like 272 00:11:23,630 --> 00:11:24,970 completing things. 273 00:11:24,970 --> 00:11:26,160 You know, there's a lot of talk about, for 274 00:11:26,160 --> 00:11:28,940 example hiring people off their GitHub resumes, 275 00:11:28,940 --> 00:11:31,140 whatever, actually think that doesn't quite ring, 276 00:11:31,140 --> 00:11:32,620 kind of correct to me. 277 00:11:32,620 --> 00:11:35,520 In the sense that place is a large premium on sort of 278 00:11:35,520 --> 00:11:36,680 lots of different things. 279 00:11:36,680 --> 00:11:38,210 I think it's actually a priori, 280 00:11:38,210 --> 00:11:40,630 sort of much more interesting to work with 281 00:11:40,630 --> 00:11:43,750 someone who has spent two years really investing in 282 00:11:43,750 --> 00:11:44,882 going deep in a particular area. 283 00:11:44,882 --> 00:11:49,670 And then the third trait they all seem to have in 284 00:11:49,670 --> 00:11:52,460 common is they just sort of cared a great deal. 285 00:11:52,460 --> 00:11:54,300 Like, it was offensive to them when something was 286 00:11:54,300 --> 00:11:55,670 just a little bit off. 287 00:11:55,670 --> 00:11:56,670 And kind of again, 288 00:11:56,670 --> 00:11:58,590 in hindsight, there are all these like crazy things we 289 00:11:58,590 --> 00:12:03,010 used to do that I mean, do in fact seem crazy like we 290 00:12:03,010 --> 00:12:04,470 probably shouldn't have done them, 291 00:12:04,470 --> 00:12:08,620 but everyone was always like well, was borderline insane 292 00:12:08,620 --> 00:12:10,740 in sort of how much they cared about tiny details. 293 00:12:10,740 --> 00:12:16,340 Like we used to, like every single API request that 294 00:12:16,340 --> 00:12:19,540 ever generated an error went to all of our inboxes and 295 00:12:19,540 --> 00:12:20,760 phoned all of us. 296 00:12:20,760 --> 00:12:23,830 Because it seems terrible to ever have an error that 297 00:12:25,120 --> 00:12:26,120 didn't go and 298 00:12:26,120 --> 00:12:28,650 get a resolution from the user standpoint. 299 00:12:28,650 --> 00:12:31,440 Or we used to like copy everyone else on 300 00:12:31,440 --> 00:12:32,850 every outgoing email. 301 00:12:32,850 --> 00:12:35,030 And we'd like point out slight grammar or 302 00:12:35,030 --> 00:12:37,070 spelling mistakes to each other, because it's terrible 303 00:12:37,070 --> 00:12:38,910 to ever send an email with a spelling mistake. 304 00:12:38,910 --> 00:12:41,890 So anyway those were the three traits come 305 00:12:41,890 --> 00:12:42,660 from that area. 306 00:12:42,660 --> 00:12:46,466 A genuine, caring a great deal. 307 00:12:46,466 --> 00:12:51,471 And, so what was my second one? 308 00:12:51,471 --> 00:12:52,105 >> The other one. 309 00:12:52,105 --> 00:12:55,442 >> Yeah so. 310 00:12:55,442 --> 00:12:58,903 Sorry yes completing things like list of three items. 311 00:13:00,990 --> 00:13:02,730 >> Yeah, I'm only gonna say I just don't think 312 00:13:02,730 --> 00:13:06,410 there's any wrong place to find people, so when I 313 00:13:06,410 --> 00:13:09,080 look back at our first few folks that we hired. 314 00:13:09,080 --> 00:13:10,090 They came from all over the place. 315 00:13:10,090 --> 00:13:12,159 Like I put up ads on Craigslist. 316 00:13:12,159 --> 00:13:14,162 I went to random tech talks. 317 00:13:14,162 --> 00:13:15,642 You know, we met people at, 318 00:13:15,642 --> 00:13:18,190 we used to throw weekly barbeques at the office. 319 00:13:18,190 --> 00:13:21,490 It was like bring your own food and drinks. 320 00:13:21,490 --> 00:13:23,070 And then we would just talk to folks. 321 00:13:23,070 --> 00:13:24,360 I think every time I ever went and 322 00:13:24,360 --> 00:13:25,330 got coffee in Palo Alto. 323 00:13:25,330 --> 00:13:28,060 Like one of you guys was recruiting at 324 00:13:28,060 --> 00:13:29,570 Coopa because their office was like 325 00:13:29,570 --> 00:13:32,605 strategically situated next to the best coffee shop. 326 00:13:32,605 --> 00:13:36,050 But I think that the really good people generally, 327 00:13:36,050 --> 00:13:38,390 they're generally doing something else, and 328 00:13:38,390 --> 00:13:39,910 so you have to go seek them out, 329 00:13:39,910 --> 00:13:42,394 rather than expecting that they're gonna seek you out. 330 00:13:42,394 --> 00:13:45,000 Triple so, when no one's ever heard of or 331 00:13:45,000 --> 00:13:47,301 is using the product that you're working on. 332 00:13:47,301 --> 00:13:52,110 >> Yeah it's probably really important to 333 00:13:52,110 --> 00:13:55,140 have a great elevator pitch. 334 00:13:55,140 --> 00:13:56,847 Not even for investors, but 335 00:13:56,847 --> 00:13:59,979 just because everyone you run into right now is maybe 336 00:13:59,979 --> 00:14:03,730 six months, a year down the road, a potential recruit. 337 00:14:03,730 --> 00:14:06,280 And so, the right time to have gotten them 338 00:14:06,280 --> 00:14:08,200 excited about your company, the right time for 339 00:14:08,200 --> 00:14:09,620 them to have started following us, 340 00:14:09,620 --> 00:14:11,820 and be thinking about it if they think about what 341 00:14:11,820 --> 00:14:15,900 they're going to do next, is as soon as you can start. 342 00:14:15,900 --> 00:14:18,220 It's gonna take a very long time to recruit people so to 343 00:14:18,220 --> 00:14:20,190 being able to consistently get people excited about 344 00:14:20,190 --> 00:14:23,330 what you're doing will pay back dividends later. 345 00:14:24,650 --> 00:14:28,190 >> Maybe this is a little tangential but John and 346 00:14:28,190 --> 00:14:29,920 I were also chatting yesterday afternoon sort of 347 00:14:29,920 --> 00:14:31,850 like a bunch of our friends have sort of 348 00:14:31,850 --> 00:14:33,862 started companies right out of school. 349 00:14:33,862 --> 00:14:35,660 We were sort of thinking about, 350 00:14:35,660 --> 00:14:37,785 what seems to go wrong in those companies. 351 00:14:37,785 --> 00:14:39,685 And I think something, 352 00:14:39,685 --> 00:14:43,010 that may be the most common failure mode. 353 00:14:43,010 --> 00:14:44,680 Since we said of doing something kind of 354 00:14:44,680 --> 00:14:46,940 overly niche, or overly sort of specific and banded. 355 00:14:46,940 --> 00:14:50,250 I think maybe it comes from sort of like there's a major 356 00:14:50,250 --> 00:14:51,930 shift in time horizon. 357 00:14:51,930 --> 00:14:54,327 As you go from classes to building a start-up. 358 00:14:54,327 --> 00:14:55,543 Right? A class kind of plays out in 359 00:14:55,543 --> 00:14:57,193 a quarter or a semester or whatever. 360 00:14:57,193 --> 00:15:00,116 Whereas a start-up is like a five or ten year thing. 361 00:15:00,116 --> 00:15:03,163 And I think this is really problematic because it's 362 00:15:03,163 --> 00:15:05,379 actually quite hard to hire people for 363 00:15:05,379 --> 00:15:08,009 niche things, in that if you tell somebody, 364 00:15:08,009 --> 00:15:11,290 look we're going to build a rocket that goes to Mars. 365 00:15:11,290 --> 00:15:13,170 Like I mean that sounds almost impossible, but 366 00:15:13,170 --> 00:15:14,470 also sounds fucking awesome. 367 00:15:14,470 --> 00:15:15,680 Right? And so it's actually pretty 368 00:15:15,680 --> 00:15:16,730 easy to convince people to work on it. 369 00:15:16,730 --> 00:15:17,740 Of course, if it's well, you know, 370 00:15:17,740 --> 00:15:20,050 we're going to build this, I don't wanna single out any 371 00:15:20,050 --> 00:15:21,710 particular idea, because probably sound like I'm 372 00:15:21,710 --> 00:15:23,090 picking some actual start up that's doing it. 373 00:15:23,090 --> 00:15:24,740 But you know, if you pick something pretty narrow, 374 00:15:24,740 --> 00:15:27,010 something that maybe kind of inductively comes out of 375 00:15:27,010 --> 00:15:28,450 the kinds of problems you solve as part of a class 376 00:15:28,450 --> 00:15:30,370 project, that's actually a much harder effort. 377 00:15:33,460 --> 00:15:36,410 >> One specific question that has come up a lot is, 378 00:15:36,410 --> 00:15:39,630 how as a relatively inexperienced founder do 379 00:15:39,630 --> 00:15:41,980 you identify who the really great people are? 380 00:15:41,980 --> 00:15:43,770 So, you know, you meet people at these barbecues or 381 00:15:43,770 --> 00:15:44,860 for your friends or whatever, and 382 00:15:44,860 --> 00:15:45,970 maybe you've worked with them a little bit. 383 00:15:45,970 --> 00:15:48,480 But what specifically did, 384 00:15:48,480 --> 00:15:51,350 did you guys do in your processes to identify like, 385 00:15:51,350 --> 00:15:53,180 you know, what this person's going to be really great? 386 00:15:53,180 --> 00:15:54,550 Or when did you really get it wrong? 387 00:15:54,550 --> 00:15:55,950 But what have you learned about how to 388 00:15:55,950 --> 00:15:58,670 identify raw talent if you can't just say, well they 389 00:15:58,670 --> 00:16:00,270 worked at Google or Facebook, they must be good? 390 00:16:03,220 --> 00:16:04,550 >> Well, I mean, you'll never 391 00:16:04,550 --> 00:16:07,272 like 100% know obviously until you work with folks. 392 00:16:07,272 --> 00:16:09,260 Which is why the flip side of it is, you know, 393 00:16:09,260 --> 00:16:11,210 if someone you hired just wasn't a good fit, 394 00:16:11,210 --> 00:16:12,550 you owe it to the company and 395 00:16:12,550 --> 00:16:15,650 to them, to tell them how they can improve, and 396 00:16:15,650 --> 00:16:18,060 if they're not working out to fire them. 397 00:16:18,060 --> 00:16:19,207 But I think that the generally, 398 00:16:19,207 --> 00:16:21,350 the question of talent falls into two big buckets. 399 00:16:21,350 --> 00:16:23,320 Like, one is you have some sense of 400 00:16:23,320 --> 00:16:25,340 what makes them good at their job, and 401 00:16:25,340 --> 00:16:28,560 there are some areas where you have taste in that area, 402 00:16:28,560 --> 00:16:30,080 and there's some where you don't. 403 00:16:30,080 --> 00:16:30,780 And the ones where you 404 00:16:30,780 --> 00:16:32,289 don't are actually much more difficult. 405 00:16:33,570 --> 00:16:36,160 So what we would do is, is we would do a few things. 406 00:16:36,160 --> 00:16:36,700 Like first, 407 00:16:36,700 --> 00:16:39,650 before talking to anyone we try to get a sense for like 408 00:16:39,650 --> 00:16:42,290 what is really world class in that discipline mean. 409 00:16:42,290 --> 00:16:43,810 And this becomes very important later, 410 00:16:43,810 --> 00:16:46,140 when you're hiring things like Head of Finance and 411 00:16:46,140 --> 00:16:47,850 you don't anything about finance. 412 00:16:47,850 --> 00:16:49,930 Except what was contained in like a library book you 413 00:16:49,930 --> 00:16:51,800 got about like an introduction to finance, 414 00:16:51,800 --> 00:16:53,295 or head of marketing. 415 00:16:53,295 --> 00:16:56,420 So, I always made it a habit of, like, talking to 416 00:16:56,420 --> 00:16:58,510 people that I knew de facto were world class and 417 00:16:58,510 --> 00:16:59,020 then asking them, 418 00:16:59,020 --> 00:17:00,890 specifically, what are the key traits or 419 00:17:00,890 --> 00:17:02,030 characteristics that you look for? 420 00:17:02,030 --> 00:17:04,170 What are the questions that you ask, and 421 00:17:04,170 --> 00:17:05,819 how do you find them? 422 00:17:05,819 --> 00:17:06,809 And if you're looking for 423 00:17:06,810 --> 00:17:08,319 the next person that's as good as you, like, 424 00:17:08,319 --> 00:17:10,699 where are, where is that person working right now, 425 00:17:10,700 --> 00:17:12,290 and, like, what's, what's her phone number? 426 00:17:14,420 --> 00:17:17,280 I think that, like, learning what good and bad is during 427 00:17:17,280 --> 00:17:19,329 the interview process, is extremely expensive. 428 00:17:19,329 --> 00:17:20,579 It's an expensive use of your time, and 429 00:17:20,579 --> 00:17:22,138 it's an expensive use of everyone else's time, 430 00:17:22,140 --> 00:17:25,460 so precalibration of that really matters. 431 00:17:25,460 --> 00:17:26,950 And then, once you have someone in, sort of, 432 00:17:26,950 --> 00:17:28,470 the interview process, 433 00:17:28,470 --> 00:17:30,012 you'll build the process over time, 434 00:17:30,012 --> 00:17:33,380 to both screen quality. 435 00:17:33,380 --> 00:17:36,640 And so at Pinterest, you know, we have 436 00:17:36,640 --> 00:17:38,520 an evolving set of standard questions that we're 437 00:17:38,520 --> 00:17:40,440 always rotating through, and we're always measuring, 438 00:17:40,440 --> 00:17:43,360 are these good indicators or bad indicators of quality? 439 00:17:43,360 --> 00:17:44,900 But the other thing that the questions are meant to do, 440 00:17:44,900 --> 00:17:46,710 is they're supposed to give a sense for, 441 00:17:46,710 --> 00:17:47,740 is this the right place for 442 00:17:47,740 --> 00:17:49,220 that person to come and work? 443 00:17:49,220 --> 00:17:50,800 And this to the point you guys made about being very 444 00:17:50,800 --> 00:17:53,250 transparent about what's gonna be easy or hard. 445 00:17:53,250 --> 00:17:55,460 Really great people wanna do things that are hard, 446 00:17:55,460 --> 00:17:57,530 they wanna solve tough problems. 447 00:17:57,530 --> 00:18:01,560 And so, there was a certain brilliance in 448 00:18:01,560 --> 00:18:03,250 Google setting out these interview questions that 449 00:18:03,250 --> 00:18:04,410 were thought to be really difficult. 450 00:18:04,410 --> 00:18:06,250 Because then people who like solving really 451 00:18:06,250 --> 00:18:10,280 difficult problems, they come out and seek those. 452 00:18:10,280 --> 00:18:11,780 I think it's really important, 453 00:18:11,780 --> 00:18:12,880 even as companies get bigger, 454 00:18:12,880 --> 00:18:14,870 that you don't whitewash the risks. 455 00:18:14,870 --> 00:18:18,120 I heard that PayPal, you'd go in and 456 00:18:18,120 --> 00:18:19,940 after interviewing with like Peter Thiel and 457 00:18:19,940 --> 00:18:22,780 Max Levchin, then they would say by the way like Visa and 458 00:18:22,780 --> 00:18:23,920 MasterCard wanna kill us, and 459 00:18:23,920 --> 00:18:26,140 we might be doing something that's illegal. 460 00:18:26,140 --> 00:18:29,220 But if you succeed, you'll redefine payments. 461 00:18:29,220 --> 00:18:30,410 Or when they were recruiting for iPhone, 462 00:18:30,410 --> 00:18:31,906 they didn't even tell people what they were doing. 463 00:18:31,906 --> 00:18:34,927 They were like you won't see your families for 464 00:18:34,927 --> 00:18:37,408 three years, but when you're done, 465 00:18:37,408 --> 00:18:40,970 your kid's kids will remember what you built. 466 00:18:40,970 --> 00:18:43,890 And I think that's a really good thing in recruiting as 467 00:18:43,890 --> 00:18:46,100 well that you're very transparent about why you 468 00:18:46,100 --> 00:18:47,950 think it's an amazing opportunity, but 469 00:18:47,950 --> 00:18:51,000 you lay out in gory detail why it's gonna be hard. 470 00:18:51,000 --> 00:18:53,090 And then the right people select in, or 471 00:18:53,090 --> 00:18:55,580 they select out of that opportunity. 472 00:18:57,710 --> 00:18:59,010 >> Evidence suggest it's worth a lot of 473 00:18:59,010 --> 00:18:59,980 people to see their kids, though. 474 00:19:03,220 --> 00:19:05,610 I feel like one thing you have to do as you try to 475 00:19:05,610 --> 00:19:12,340 identify talent is have the confidence to interview for, 476 00:19:12,340 --> 00:19:13,922 in a way that works for you. 477 00:19:13,922 --> 00:19:17,412 I think you know, if you're, say you're not, 478 00:19:17,412 --> 00:19:19,410 the world's best engineer and 479 00:19:19,410 --> 00:19:21,850 you're trying to interview engineering candidates. 480 00:19:21,850 --> 00:19:23,220 I think it's tempting to try to 481 00:19:23,220 --> 00:19:24,880 color code what everyone else does. 482 00:19:24,880 --> 00:19:26,320 And you know, get them to call it on white board. 483 00:19:26,320 --> 00:19:28,750 And do other engineering, interviewee things. 484 00:19:28,750 --> 00:19:30,140 You know in the case of Stripe, 485 00:19:30,140 --> 00:19:31,990 hiring our first engineer. 486 00:19:31,990 --> 00:19:35,730 We flew the guy out, and we spent a weekend coding with 487 00:19:35,730 --> 00:19:37,760 him, and you know looked over his shoulder and 488 00:19:37,760 --> 00:19:40,830 kind of, it was the only way we could really tell and 489 00:19:40,830 --> 00:19:45,070 get ourselves confident that, that person was good. 490 00:19:45,070 --> 00:19:47,720 And I think you can actually extend that to other roles, 491 00:19:47,720 --> 00:19:49,310 where again you're not an expert, and 492 00:19:49,310 --> 00:19:52,500 that you know I'm no business development guru. 493 00:19:52,500 --> 00:19:54,170 But when we interview people for 494 00:19:54,170 --> 00:19:55,360 business development roles, 495 00:19:55,360 --> 00:19:56,962 we'll ask them to do a project, you know? 496 00:19:56,962 --> 00:19:58,780 Where they talk about how they would 497 00:19:58,780 --> 00:20:01,080 improve an existing partnership that Stripe has. 498 00:20:01,080 --> 00:20:02,810 Or which new partnerships they would go out and do. 499 00:20:02,810 --> 00:20:04,140 And again, you know, 500 00:20:04,140 --> 00:20:06,520 even though it's not my domain area, I am actually 501 00:20:06,520 --> 00:20:09,390 confident enough that I can judge those pretty well. 502 00:20:09,390 --> 00:20:12,160 And I think people often have this imposter syndrome 503 00:20:12,160 --> 00:20:14,580 when it comes to interviewing for roles. 504 00:20:14,580 --> 00:20:16,135 Yeah. I think a pretty specific, 505 00:20:16,135 --> 00:20:18,483 just kind of just tactical thing, and to do for 506 00:20:18,483 --> 00:20:21,313 again the first 10 people, is to work with them as much 507 00:20:21,313 --> 00:20:23,620 as you can before committing to hiring them. 508 00:20:23,620 --> 00:20:24,370 I mean, once you 509 00:20:24,370 --> 00:20:26,070 reach a certain scale it's kind of impractical, 510 00:20:26,070 --> 00:20:27,700 because it's a huge time commitment on their side. 511 00:20:27,700 --> 00:20:29,630 And ultimately I mean it just would be unskilled, 512 00:20:29,630 --> 00:20:31,590 would be expensive from your side. 513 00:20:31,590 --> 00:20:33,210 But it's really worth it 514 00:20:33,210 --> 00:20:34,330 to get the first ten people right. 515 00:20:34,330 --> 00:20:35,180 And so, for 516 00:20:35,180 --> 00:20:38,120 a majority, maybe in all of the first ten people, 517 00:20:38,120 --> 00:20:39,680 we worked with them in some capacity. 518 00:20:39,680 --> 00:20:41,480 Usually for a week in advance, and 519 00:20:41,480 --> 00:20:43,777 it's pretty hard to fake it for a week. 520 00:20:45,010 --> 00:20:49,020 It tends to be pretty clear, quite quickly. 521 00:20:49,020 --> 00:20:52,840 And another thing I thought of you know, to the question 522 00:20:52,840 --> 00:20:54,830 of sort of how do you know like who is great or 523 00:20:54,830 --> 00:20:55,870 who is good enough or whatever. 524 00:20:55,870 --> 00:20:57,740 And people always talk about this notion of 525 00:20:57,740 --> 00:21:00,350 the ten x person, that whatever that skill set is. 526 00:21:00,350 --> 00:21:02,210 I don't really know what ten x means, 527 00:21:02,210 --> 00:21:05,550 I think maybe the slightly more kind of helpful or like 528 00:21:05,550 --> 00:21:09,180 intuitive version of that is, is this person probably, 529 00:21:09,180 --> 00:21:12,090 the best out of all of their friends at what they do. 530 00:21:13,560 --> 00:21:15,600 And so you know, it's a little bit sensitive to you 531 00:21:15,600 --> 00:21:18,290 know, how well they chose friends, how many they have. 532 00:21:18,290 --> 00:21:20,150 But for me at least, I find that kind of 533 00:21:20,150 --> 00:21:21,550 a more helpful way to think about it, like 534 00:21:21,550 --> 00:21:23,209 is this the best engineer this engineer knows. 535 00:21:24,510 --> 00:21:25,230 And the other thing I 536 00:21:25,230 --> 00:21:26,570 think that's actually probably just worth 537 00:21:26,570 --> 00:21:28,320 mentioning in all of those kind of first ten people, or 538 00:21:28,320 --> 00:21:31,550 even more generally on the culture and hiring topic. 539 00:21:31,550 --> 00:21:34,650 I think everyone sort of doesn't realize until they 540 00:21:34,650 --> 00:21:36,620 go through it themselves how it important it is, 541 00:21:36,620 --> 00:21:40,210 in large part because the, like in life and the media 542 00:21:40,210 --> 00:21:42,580 and everything, people focus way too much on founders. 543 00:21:42,580 --> 00:21:43,780 And they're like, here we are. 544 00:21:43,780 --> 00:21:44,430 And so kind of, 545 00:21:44,430 --> 00:21:46,760 we're re-enforcing the sort of structural narrative that 546 00:21:46,760 --> 00:21:48,660 like, stripe is about John and Patrick. 547 00:21:48,660 --> 00:21:50,510 And Pinterest is about Ben and so forth. 548 00:21:50,510 --> 00:21:51,830 Whereas, I mean, obviously, 549 00:21:51,830 --> 00:21:54,320 sort of the vast majority of what our companies do 550 00:21:54,320 --> 00:21:56,860 like 99.9% is being down by people who are not us. 551 00:21:56,860 --> 00:21:57,490 Right? 552 00:21:57,490 --> 00:21:58,930 And I think, I mean, that's kind of, 553 00:21:58,930 --> 00:22:02,410 it's obvious when you say it, but it's at a very much, 554 00:22:02,410 --> 00:22:03,840 not just the macro narrative and 555 00:22:03,840 --> 00:22:04,970 you know, companies are abstract. 556 00:22:04,970 --> 00:22:06,670 So you kind of need to associate them with people. 557 00:22:06,670 --> 00:22:08,230 But I think it's worth bearing in mind that like 558 00:22:08,230 --> 00:22:12,010 for, you know, Apple, you know everything was 559 00:22:12,010 --> 00:22:14,840 you know, Steve Jobs was a like tiny, tiny detail at 560 00:22:14,840 --> 00:22:17,240 the end, right, or Google was that way, and so forth. 561 00:22:17,240 --> 00:22:19,590 >> So don't screw it up, is that what you're saying? 562 00:22:19,590 --> 00:22:20,430 >> Something like that. 563 00:22:20,430 --> 00:22:22,870 >> I think, you know, one other thing I'll mention is 564 00:22:22,870 --> 00:22:25,360 that I think referencing people is really important, 565 00:22:25,360 --> 00:22:30,270 and referencing people is just what it sounds like. 566 00:22:30,270 --> 00:22:31,780 You're basically asking people who have real, 567 00:22:31,780 --> 00:22:34,890 material working experience, for their honest opinion. 568 00:22:34,890 --> 00:22:38,220 And we do that really aggressively, and 569 00:22:38,220 --> 00:22:38,740 what we're trying to 570 00:22:38,740 --> 00:22:40,900 figure out is what's this person like to work with? 571 00:22:42,590 --> 00:22:44,070 We're not trying to validate like whether they 572 00:22:44,070 --> 00:22:45,260 told the truth on their resume, 573 00:22:45,260 --> 00:22:47,400 cuz we assume that they're telling the truth. 574 00:22:47,400 --> 00:22:50,740 So, very standard questions that I'll ask somebody who's 575 00:22:50,740 --> 00:22:53,460 in an interview, I might say, hey, 576 00:22:53,460 --> 00:22:55,500 we both know Jonathan in common. 577 00:22:55,500 --> 00:22:57,240 I'm gonna talk to him, you know, 578 00:22:57,240 --> 00:22:59,090 in a couple weeks, cause we're both social friends. 579 00:22:59,090 --> 00:23:02,560 Like if I asked him, what you're the best at, or 580 00:23:02,560 --> 00:23:03,690 what you would be most proud of, or 581 00:23:03,690 --> 00:23:06,230 what you were kind of working to improve. 582 00:23:06,230 --> 00:23:08,770 What would he or she say? 583 00:23:08,770 --> 00:23:10,760 Because it sort of tests the level of self awareness, and 584 00:23:10,760 --> 00:23:12,900 creates a bit of social accountability. 585 00:23:12,900 --> 00:23:15,550 And then I'll try to ask something that makes the 586 00:23:15,550 --> 00:23:18,170 question which is typically very soft feel a little bit 587 00:23:18,170 --> 00:23:20,300 more quantitative, and then calibrate that over time. 588 00:23:20,300 --> 00:23:21,890 So you know, 589 00:23:21,890 --> 00:23:24,680 in evaluating this person or this dimension, do you think 590 00:23:24,680 --> 00:23:26,110 this is the top 1% of people you've ever worked with? 591 00:23:26,110 --> 00:23:28,090 The top 5% and the top 10%? 592 00:23:28,090 --> 00:23:31,410 And it forces a scarcity that gives a materially 593 00:23:31,410 --> 00:23:34,110 different reference, then if you just say hey, 594 00:23:34,110 --> 00:23:36,060 like what's the best thing about John? 595 00:23:36,060 --> 00:23:37,680 He told me you know, he's good at these things. 596 00:23:37,680 --> 00:23:38,250 Can you validate? 597 00:23:38,250 --> 00:23:38,840 And you're like yeah, 598 00:23:38,840 --> 00:23:40,220 sure, Cuz they don't gain anything. 599 00:23:40,220 --> 00:23:43,132 So I think that's just a tool that people should 600 00:23:43,132 --> 00:23:43,920 take seriously. 601 00:23:43,920 --> 00:23:44,559 >> Yeah. 602 00:23:44,559 --> 00:23:45,149 >> Yeah, and 603 00:23:45,149 --> 00:23:48,168 referencing isn't obviously easy to begin with, but 604 00:23:48,168 --> 00:23:51,400 it does actually prove really useful over time. 605 00:23:51,400 --> 00:23:54,160 And you just have to, people I think especially for 606 00:23:54,160 --> 00:23:56,350 named references, people really want to be nice. 607 00:23:56,350 --> 00:23:58,370 So you have to do things like create an artificial 608 00:23:58,370 --> 00:24:00,600 scarcity by saying you know, where would you rank this 609 00:24:00,600 --> 00:24:03,230 person amongst the people you worked within this role? 610 00:24:03,230 --> 00:24:03,780 Or, you know, 611 00:24:03,780 --> 00:24:05,730 you just splitz out on the enchant of how to reference, 612 00:24:05,730 --> 00:24:07,630 but usually I aim to spend, you know, 613 00:24:07,630 --> 00:24:09,470 15 minutes on the phone with that person. 614 00:24:09,470 --> 00:24:12,520 Not just let them say, yeah this person is awesome, and 615 00:24:12,520 --> 00:24:13,720 then hang up. 616 00:24:13,720 --> 00:24:16,190 >> That's also a tremendous source of new recruits. 617 00:24:16,190 --> 00:24:16,800 >> Yeah. 618 00:24:16,800 --> 00:24:18,530 >> Those references are also a tremendous source of 619 00:24:18,530 --> 00:24:19,120 new recruits. 620 00:24:20,900 --> 00:24:22,690 What have each of you learned about, 621 00:24:22,690 --> 00:24:25,770 once you hire these first people and they joined? 622 00:24:25,770 --> 00:24:28,300 What have you done to make them effective quickly? 623 00:24:28,300 --> 00:24:31,280 To get them sort of, you know, 624 00:24:31,280 --> 00:24:34,190 to the right cultural place, you know? 625 00:24:34,190 --> 00:24:36,600 Cuz sort of hiring is usually very difficult, but 626 00:24:36,600 --> 00:24:38,430 then not as difficult as making them 627 00:24:38,430 --> 00:24:39,150 happy and effective. 628 00:24:39,150 --> 00:24:39,860 So what do you do 629 00:24:39,860 --> 00:24:41,500 with these early employees to accomplish that? 630 00:24:44,640 --> 00:24:46,860 >> Well, ways that it's changed when we've gone from 631 00:24:46,860 --> 00:24:48,790 really small to bigger. 632 00:24:48,790 --> 00:24:49,800 When we first started, 633 00:24:49,800 --> 00:24:52,970 we were generally hiring because we needed 634 00:24:52,970 --> 00:24:54,700 that person, like, a long time ago. 635 00:24:54,700 --> 00:24:57,090 And so their whole on-boarding was like, 636 00:24:57,090 --> 00:24:57,690 here's your computer, 637 00:24:57,690 --> 00:24:58,880 we already set up your environment, 638 00:24:58,880 --> 00:24:59,500 don't worry about it, 639 00:24:59,500 --> 00:25:01,690 this is the problem that we have to solve together. 640 00:25:01,690 --> 00:25:04,810 And then because of the nature of the start ups, 641 00:25:04,810 --> 00:25:07,380 we were all in this, like, tiny, two bedroom apartment. 642 00:25:08,770 --> 00:25:09,760 All of the other things, 643 00:25:09,760 --> 00:25:11,250 like building personal relationships, 644 00:25:11,250 --> 00:25:12,680 spending time together, getting to know one another, 645 00:25:12,680 --> 00:25:13,810 it just all happened. 646 00:25:13,810 --> 00:25:14,990 Kind of automatically like, 647 00:25:14,990 --> 00:25:16,145 you didn't have to do anything. 648 00:25:16,145 --> 00:25:19,730 The one thing that I would add to that is, though, 649 00:25:19,730 --> 00:25:21,420 we would always try to remind people of like, 650 00:25:21,420 --> 00:25:22,600 where we wanted to go some day. 651 00:25:22,600 --> 00:25:24,370 Cuz it's really easy when you first join to drop 652 00:25:24,370 --> 00:25:26,380 somebody into a problem, and then they think the whole 653 00:25:26,380 --> 00:25:28,520 world is like this little problem in front of them. 654 00:25:28,520 --> 00:25:29,080 We'd always say, 655 00:25:29,080 --> 00:25:31,320 you know, hey, some day we wanna do for 656 00:25:33,130 --> 00:25:34,830 Discovery what Google did for search. 657 00:25:34,830 --> 00:25:37,300 This is our plan for trying to get that done. 658 00:25:37,300 --> 00:25:40,690 Now as the company, grows I think that process has to 659 00:25:40,690 --> 00:25:42,300 get a little bit more formalized. 660 00:25:42,300 --> 00:25:45,660 And so we spend a lot of time thinking, and honestly 661 00:25:45,660 --> 00:25:48,770 trying to constantly refine what is that person's 662 00:25:48,770 --> 00:25:51,560 experience look like from the day they came in, 663 00:25:51,560 --> 00:25:53,240 to their very first interview, 664 00:25:53,240 --> 00:25:54,960 through 30 days after they joined. 665 00:25:54,960 --> 00:25:57,590 Like do they have somebody whose name they know? 666 00:25:57,590 --> 00:25:58,730 Do they know who their manager is? 667 00:25:58,730 --> 00:26:01,100 Have they sat down with people on their team? 668 00:26:01,100 --> 00:26:03,070 Do they know what the general architecture of 669 00:26:03,070 --> 00:26:05,700 the company is, and what the top priorities are? 670 00:26:05,700 --> 00:26:08,390 And we have a program and it's a week long, and 671 00:26:08,390 --> 00:26:10,370 then there are some function specific programs that 672 00:26:10,370 --> 00:26:11,740 go in deeper. 673 00:26:11,740 --> 00:26:13,638 And that's something that's always been refined. 674 00:26:13,638 --> 00:26:15,190 And the alpha metrics that we look on that, 675 00:26:15,190 --> 00:26:17,400 it's when we ask people like what did you think 676 00:26:17,400 --> 00:26:19,090 afterwards and then 30 days afterwards. 677 00:26:19,090 --> 00:26:21,030 And then we also ask their peers and their manager, 678 00:26:21,030 --> 00:26:23,330 like hey, is this person kind of up to speed? 679 00:26:23,330 --> 00:26:24,090 Do you feel like we've done 680 00:26:24,090 --> 00:26:25,290 a good job making them productive? 681 00:26:25,290 --> 00:26:28,330 And if we haven't, that's a key A, that, that team 682 00:26:28,330 --> 00:26:30,760 shouldn't be hiring anymore people, because they're not 683 00:26:30,760 --> 00:26:32,720 doing a good job bringing in new people. 684 00:26:33,740 --> 00:26:35,655 And B, that we need to retool that. 685 00:26:35,655 --> 00:26:37,520 And so I think those things are important. 686 00:26:37,520 --> 00:26:40,450 I just wouldn't discount just how important it is to 687 00:26:40,450 --> 00:26:42,380 get to know the person as a person. 688 00:26:42,380 --> 00:26:43,840 Like what are their aspirations? 689 00:26:43,840 --> 00:26:44,730 What's their working style? 690 00:26:44,730 --> 00:26:47,250 How do they like to be recognized? 691 00:26:48,450 --> 00:26:51,810 Do they really prefer being in total silence? 692 00:26:51,810 --> 00:26:53,070 Are they a morning person or night person? 693 00:26:53,070 --> 00:26:56,400 Like knowing those things I think, it just 694 00:26:56,400 --> 00:26:58,270 demonstrates that you care about them individually as 695 00:26:58,270 --> 00:27:00,320 well as collectively, like what you goals are. 696 00:27:00,320 --> 00:27:04,540 >> I think there are two things that are important at 697 00:27:04,540 --> 00:27:06,680 any stage through the implementation of a change. 698 00:27:06,680 --> 00:27:08,330 The first is to get them up and 699 00:27:08,330 --> 00:27:10,480 running, doing real work quickly. 700 00:27:10,480 --> 00:27:12,560 Because that's only when you can find the problems, 701 00:27:12,560 --> 00:27:15,110 that's how progress is measured, 702 00:27:15,110 --> 00:27:16,340 is how much real work they're doing. 703 00:27:16,340 --> 00:27:19,000 And so, you know, when we have engineers start, 704 00:27:19,000 --> 00:27:20,800 we'll try to get them committing on the first day. 705 00:27:20,800 --> 00:27:21,620 When we have people you know, 706 00:27:21,620 --> 00:27:23,150 people in business roles start, 707 00:27:23,150 --> 00:27:26,750 we will have them in real meetings on the first day, 708 00:27:26,750 --> 00:27:28,390 on what they're meant to be working on. 709 00:27:28,390 --> 00:27:30,260 But I think sometimes there's a tendency to 710 00:27:30,260 --> 00:27:32,430 be tentative, and help ease people in. 711 00:27:32,430 --> 00:27:36,930 We're much more of the push people off the cliff school. 712 00:27:36,930 --> 00:27:38,820 And then I think the second, is to try and 713 00:27:38,820 --> 00:27:41,690 start quickly giving people feedback, and 714 00:27:41,690 --> 00:27:43,710 especially feedback on how to adapt to the culture. 715 00:27:43,710 --> 00:27:46,020 Because when you think about it, like, you know, 716 00:27:46,020 --> 00:27:49,490 if you have built a strong culture as, you know, 717 00:27:49,490 --> 00:27:52,310 all the companies up here are trying to, 718 00:27:52,310 --> 00:27:55,080 then it's going to take some adapting for the person. 719 00:27:55,080 --> 00:27:58,210 It's not going to be necessarily easy, you know? 720 00:27:58,210 --> 00:27:59,150 One thing we have at Stripe, 721 00:27:59,150 --> 00:28:02,210 is just the culture was a lot more written. 722 00:28:02,210 --> 00:28:03,050 And so, you know? 723 00:28:03,050 --> 00:28:04,480 You have people who are right next to 724 00:28:04,480 --> 00:28:05,790 each other each with headphones on, and 725 00:28:05,790 --> 00:28:07,360 just like IM'ing away to each other. 726 00:28:07,360 --> 00:28:08,170 And for you know, 727 00:28:08,170 --> 00:28:09,490 a lot of people coming in who hadn't worked in 728 00:28:09,490 --> 00:28:11,210 an environment like that, it's, it's sort of hard. 729 00:28:11,210 --> 00:28:12,050 >> In normal places. 730 00:28:12,050 --> 00:28:12,680 >> Exactly. Yeah. 731 00:28:13,980 --> 00:28:16,520 And so, everything from kind of high level how you're 732 00:28:16,520 --> 00:28:19,780 doing at your job, to kind of minor cultural pointers. 733 00:28:19,780 --> 00:28:22,530 The more feedback you can give them, 734 00:28:22,530 --> 00:28:23,440 the better they'll do. 735 00:28:23,440 --> 00:28:24,150 And it's unnatural. 736 00:28:24,150 --> 00:28:24,660 Right? Because it's 737 00:28:24,660 --> 00:28:27,850 unnatural to be constantly telling people that 738 00:28:27,850 --> 00:28:28,980 they're doing a good or bad job. 739 00:28:28,980 --> 00:28:30,870 You don't do that, hopefully, 740 00:28:30,870 --> 00:28:34,210 in your normal life, you're more restrained. 741 00:28:34,210 --> 00:28:35,570 But when you have employees, 742 00:28:35,570 --> 00:28:38,470 that's kind of what you owe them for them to do well. 743 00:28:38,470 --> 00:28:40,170 >> So I think this is sort of a good transition too, 744 00:28:40,170 --> 00:28:41,609 as your companies have scaled. 745 00:28:43,490 --> 00:28:46,420 What are the biggest changes you've had to make to your 746 00:28:46,420 --> 00:28:47,560 hiring processes? 747 00:28:47,560 --> 00:28:50,140 And also how you sort of manage and 748 00:28:50,140 --> 00:28:53,020 run the team, as you've gone from, you know, 749 00:28:53,020 --> 00:28:54,980 two to 10 to 100 to 1000 employees? 750 00:28:59,010 --> 00:28:59,950 >> There a lot of changes. 751 00:29:01,650 --> 00:29:03,550 I mean, I think one thing that we try to do on 752 00:29:03,550 --> 00:29:07,040 the team side is, the goal is to make the teams feel as 753 00:29:07,040 --> 00:29:09,130 autonomous and nimble as possible within 754 00:29:09,130 --> 00:29:11,140 the constraints of having a large organization. 755 00:29:11,140 --> 00:29:14,020 And that means over time, that we're always trying to 756 00:29:14,020 --> 00:29:16,430 make it feel like a start-up of of many start-ups, 757 00:29:16,430 --> 00:29:18,690 rather than this monolithic organization with a set of 758 00:29:18,690 --> 00:29:21,170 foreign processes that cut horizontally through it. 759 00:29:21,170 --> 00:29:22,290 And it's easier said than done. 760 00:29:22,290 --> 00:29:24,280 I think that, like we're not all the way there. 761 00:29:24,280 --> 00:29:27,830 But one of our goals is that each team has, they control 762 00:29:27,830 --> 00:29:29,772 the resources that they need to get things done, 763 00:29:29,772 --> 00:29:32,960 they know what the most important thing is, and 764 00:29:32,960 --> 00:29:34,150 how it's measured. 765 00:29:34,150 --> 00:29:35,320 And in that way, 766 00:29:35,320 --> 00:29:37,490 then the management problem becomes somewhat tractable, 767 00:29:37,490 --> 00:29:39,410 otherwise it feels completely impossible if 768 00:29:39,410 --> 00:29:41,610 you can't decompose it into autonomous units. 769 00:29:41,610 --> 00:29:43,609 Like you just look at it, and you're like, oh my gosh, 770 00:29:43,609 --> 00:29:45,752 communication complexity is increasing geometrically. 771 00:29:45,752 --> 00:29:46,810 Management complexity, 772 00:29:46,810 --> 00:29:48,250 like it's never gonna work, and so 773 00:29:48,250 --> 00:29:50,411 you have to sort of create these abstracted units, or 774 00:29:50,411 --> 00:29:52,247 that's at least what we're gonna try to do. 775 00:29:52,247 --> 00:29:54,741 And at Pinterest in particular, the real 776 00:29:54,741 --> 00:29:57,811 challenge of creating those abstracted units is. 777 00:29:57,811 --> 00:29:59,296 We want to actually have 778 00:29:59,296 --> 00:30:02,419 units that encompass a super-strong designer, 779 00:30:02,419 --> 00:30:06,060 a super-strong leader in engineering, often a writer, 780 00:30:06,060 --> 00:30:08,830 often sometimes like a community manager. 781 00:30:08,830 --> 00:30:10,890 We want them all to be self-contained. 782 00:30:10,890 --> 00:30:12,190 And so that makes the problem really hard, 783 00:30:12,190 --> 00:30:14,860 but that's kind of core to our philosophy of how we 784 00:30:14,860 --> 00:30:15,480 build products. 785 00:30:15,480 --> 00:30:17,560 We try to put people together that have all of 786 00:30:17,560 --> 00:30:18,520 these different disciplines. 787 00:30:18,520 --> 00:30:20,056 They're curious about lots of things. 788 00:30:20,056 --> 00:30:21,680 And then we anchor them on a single project and then we 789 00:30:21,680 --> 00:30:24,115 try to remove barriers to let them go fast. 790 00:30:24,115 --> 00:30:25,940 When we find new barriers, we sit down and 791 00:30:25,940 --> 00:30:28,180 we think, how do we speed that up? 792 00:30:28,180 --> 00:30:30,850 I think hiring is a little bit different. 793 00:30:30,850 --> 00:30:33,710 And I think the biggest change and 794 00:30:33,710 --> 00:30:35,200 the biggest asset that you get as you get more 795 00:30:35,200 --> 00:30:38,890 people is referrals become like more and more and 796 00:30:38,890 --> 00:30:40,660 more and more the life blood depending on 797 00:30:40,660 --> 00:30:41,998 the network of all the people that you bring in. 798 00:30:41,998 --> 00:30:43,380 So one of the really lucky, 799 00:30:43,380 --> 00:30:46,087 and in hindsight, great decisions that we made, was 800 00:30:46,087 --> 00:30:48,794 actually the 14th or 15th person that we hired was 801 00:30:48,794 --> 00:30:51,888 a professional recruiter and she had worked at start-ups. 802 00:30:51,888 --> 00:30:54,580 She had worked at big companies like Apple. 803 00:30:54,580 --> 00:30:57,672 But she knew where that pipeline breaks down, 804 00:30:57,672 --> 00:30:59,812 knew the early indicators, and 805 00:30:59,812 --> 00:31:03,460 taught everyone not just how to screen for talent, but 806 00:31:03,460 --> 00:31:06,949 how scalably to identify the people that are gonna be 807 00:31:06,949 --> 00:31:09,900 culturally, really good for the company. 808 00:31:09,900 --> 00:31:12,076 And I think looking back on that, 809 00:31:12,076 --> 00:31:15,350 it's something that I personally really value. 810 00:31:16,770 --> 00:31:17,894 >> You know, there's just a huge amount of 811 00:31:17,894 --> 00:31:18,495 stuff here, right? 812 00:31:18,495 --> 00:31:19,632 In that I guess, 813 00:31:19,632 --> 00:31:23,408 this is all under kinda the rubric of managing growth. 814 00:31:23,408 --> 00:31:27,737 And either your company fails very quickly or all of 815 00:31:27,737 --> 00:31:31,210 your problems become in some way about managing growth. 816 00:31:32,800 --> 00:31:35,030 One thing that I think tends to take people by surprise, 817 00:31:35,030 --> 00:31:36,550 and certainly took me by surprise, 818 00:31:36,550 --> 00:31:39,330 is kind of how quickly just like the time 819 00:31:39,330 --> 00:31:40,370 horizon has changed. 820 00:31:40,370 --> 00:31:44,720 In that, in your first month you're largely thinking 821 00:31:44,720 --> 00:31:48,660 about things maybe one month ahead, right? 822 00:31:48,660 --> 00:31:50,168 In that, maybe that's kind of 823 00:31:50,168 --> 00:31:52,721 what your development road map is oriented around, 824 00:31:52,721 --> 00:31:54,820 even in terms of who you're working with. 825 00:31:54,820 --> 00:31:55,370 Maybe it's like, 826 00:31:55,370 --> 00:31:58,280 informal relationships where they haven't fully committed 827 00:31:58,280 --> 00:32:00,160 to remain full time or whatever, right? 828 00:32:00,160 --> 00:32:02,372 And then, the more time goes by, 829 00:32:02,372 --> 00:32:05,058 I think that kind of has a reciprocal, or 830 00:32:05,058 --> 00:32:08,220 corresponding increase in the time horizons. 831 00:32:08,220 --> 00:32:09,920 After one year, you're kind of thinking one year ahead. 832 00:32:09,920 --> 00:32:13,675 After four years, you're thinking four years ahead. 833 00:32:13,675 --> 00:32:16,870 But that increases very quickly, right? 834 00:32:16,870 --> 00:32:17,879 And so after you know, 835 00:32:17,879 --> 00:32:20,440 it's after one month, it's again, super short term. 836 00:32:20,440 --> 00:32:23,180 And then just 11 months later you should now 837 00:32:23,180 --> 00:32:25,650 actually be thinking and planning a year ahead, and 838 00:32:25,650 --> 00:32:27,250 thinking about human structures on 839 00:32:27,250 --> 00:32:28,162 that time horizon. 840 00:32:28,162 --> 00:32:31,315 Thinking about the stuff Ben talked about, like where you 841 00:32:31,315 --> 00:32:34,530 want to be going long term and things like that. 842 00:32:34,530 --> 00:32:36,773 I think that also plays into your hiring and 843 00:32:36,773 --> 00:32:38,775 the kinds of people you hire, right? 844 00:32:38,775 --> 00:32:40,499 In that in the really early days you kind of 845 00:32:40,499 --> 00:32:42,322 have to hire people who will be productive, 846 00:32:42,322 --> 00:32:43,520 essentially immediately. 847 00:32:44,670 --> 00:32:47,812 You don't have the luxury of hiring people who are really 848 00:32:47,812 --> 00:32:50,496 promising, but they're not quite gonna be up to 849 00:32:50,496 --> 00:32:52,420 speed for another year or two. 850 00:32:52,420 --> 00:32:54,090 They have to be able to contribute immediately. 851 00:32:54,090 --> 00:32:57,330 But after two or three years, now it starts to get 852 00:32:57,330 --> 00:32:59,550 much more reasonable to make those investments and 853 00:32:59,550 --> 00:33:01,110 in fact, if you're not making those investments, 854 00:33:01,110 --> 00:33:03,380 you're probably being much too short term. 855 00:33:03,380 --> 00:33:06,060 And so I think that's really important. 856 00:33:06,060 --> 00:33:07,880 And a lot of it also just comes from 857 00:33:09,420 --> 00:33:11,355 all these problems in some sense are easy, 858 00:33:11,355 --> 00:33:14,240 like how do you build good social bonds between people? 859 00:33:14,240 --> 00:33:17,795 I mean, we all do it every day, right? 860 00:33:17,795 --> 00:33:21,355 It's kind of, how do you make it systematic and 861 00:33:21,355 --> 00:33:22,968 effective at scale? 862 00:33:22,968 --> 00:33:26,597 It's always a very severely imperfect approximation of 863 00:33:26,597 --> 00:33:29,620 what you would ideally do if you were small, and 864 00:33:29,620 --> 00:33:32,566 what hacks can you pull to make it work as well as 865 00:33:32,566 --> 00:33:35,460 you possibly can at a larger scale. 866 00:33:35,460 --> 00:33:36,804 Like a rapidly growing company, 867 00:33:36,804 --> 00:33:38,244 say growing headcount here, two or 868 00:33:38,244 --> 00:33:40,412 three x a year, it is a very unnatural thing. 869 00:33:40,412 --> 00:33:43,327 And so it's, what's the least bad way of sort of 870 00:33:43,327 --> 00:33:45,370 managing that period of growth? 871 00:33:45,370 --> 00:33:46,620 Human organizations aren't designed for 872 00:33:46,620 --> 00:33:52,240 it and I think it's worth being quite systematic about 873 00:33:52,240 --> 00:33:53,360 thinking about ways to do that. 874 00:33:53,360 --> 00:33:56,650 But realizing that you probably can't do 875 00:33:56,650 --> 00:33:59,420 much better than adequation. 876 00:33:59,420 --> 00:34:01,464 You know, for Stripe, it's things like we 877 00:34:01,464 --> 00:34:03,120 have three meals every day at sort of 878 00:34:03,120 --> 00:34:06,008 long tables where everyone can sit together, right? 879 00:34:06,008 --> 00:34:09,190 And if you think of net, how much more total human 880 00:34:09,190 --> 00:34:12,168 interaction happens as a result of having these kind 881 00:34:12,168 --> 00:34:15,232 of randomly mixed meals, it's vast, right? 882 00:34:15,233 --> 00:34:17,900 And it's kind of a whole list of things like that. 883 00:34:17,900 --> 00:34:20,260 But, I think that's kind of the general framework. 884 00:34:22,020 --> 00:34:23,500 >> One thing I'm really curious about you guys value 885 00:34:23,500 --> 00:34:24,053 transparency. 886 00:34:24,053 --> 00:34:26,577 How have you scaled it over time? 887 00:34:26,577 --> 00:34:28,320 I know we think about figuring it out all 888 00:34:28,320 --> 00:34:29,318 the time, just curious. 889 00:34:29,318 --> 00:34:33,404 >> So you know, start-ups, I can't remember who it was, 890 00:34:33,404 --> 00:34:37,576 defined it as a, start-up is an organization that's not 891 00:34:37,576 --> 00:34:41,510 yet stuck with all of these principle agent problems. 892 00:34:41,510 --> 00:34:43,616 That at most large companies, 893 00:34:43,616 --> 00:34:47,261 what is locally optimal for you is very frequently not 894 00:34:47,261 --> 00:34:50,204 what is globally optimal for the company. 895 00:34:50,204 --> 00:34:56,850 And so there are, as a consequence of that, 896 00:34:56,850 --> 00:34:58,480 probably a lot of ways in which a start-up can 897 00:34:58,480 --> 00:35:00,270 work differently to a big company. 898 00:35:00,270 --> 00:35:04,130 At a big company, a lot of the things that are good for 899 00:35:04,130 --> 00:35:07,460 you, well, you couldn't do them in 900 00:35:07,460 --> 00:35:09,640 a completely transparent environment, right? 901 00:35:09,640 --> 00:35:11,350 Because people would think less of you or 902 00:35:11,350 --> 00:35:13,180 you're doing things you're not supposed to. 903 00:35:13,180 --> 00:35:15,005 But because everyone is kind of sailing in the, or 904 00:35:15,005 --> 00:35:16,916 rowing in the same direction at a start-up, you can 905 00:35:16,916 --> 00:35:19,022 actually just make all of the information transparent. 906 00:35:19,022 --> 00:35:23,369 And so, I guess I mentioned earlier, Stripe used to BCC 907 00:35:23,369 --> 00:35:27,626 every other person at Stripe on basically every single 908 00:35:27,626 --> 00:35:31,789 outgoing email unless you opted out of it, because we 909 00:35:31,789 --> 00:35:35,230 thought that would be much more efficient. 910 00:35:35,230 --> 00:35:37,470 You wouldn't need to have as many meetings, 911 00:35:37,470 --> 00:35:39,370 if you just kind of keep abreast of what's happening. 912 00:35:39,370 --> 00:35:42,590 And over time we've sort of built an increasingly 913 00:35:42,590 --> 00:35:45,700 intricate framework of mailing lists and 914 00:35:45,700 --> 00:35:48,680 we now have a program for generating Gmail filters. 915 00:35:48,680 --> 00:35:51,390 And for like a pretty rocky patch where 50 people or 916 00:35:51,390 --> 00:35:54,840 so, to Ben's point of like, asking people how they're 917 00:35:54,840 --> 00:35:56,380 getting on after the first couple of days. 918 00:35:56,380 --> 00:35:58,671 They're reported terribly because like they couldn't 919 00:35:58,671 --> 00:36:00,722 even find all the emails that people were sending to 920 00:36:00,722 --> 00:36:03,710 them and they were missing things and everything. 921 00:36:03,710 --> 00:36:05,080 >> Gmail broke at one stage. 922 00:36:05,080 --> 00:36:06,584 >> Oh, right, at one point Gmail broke because we 923 00:36:06,584 --> 00:36:07,838 were just like sending too much email. 924 00:36:07,838 --> 00:36:10,804 It is hard to scale, because I mean, 925 00:36:10,804 --> 00:36:15,210 you might contact somebody outside of the company with 926 00:36:15,210 --> 00:36:18,657 like some great idea, and maybe the person 927 00:36:18,657 --> 00:36:22,776 sitting across the way from you thinks that's like 928 00:36:22,776 --> 00:36:26,818 the stupidest idea they've ever heard, right? 929 00:36:26,818 --> 00:36:29,124 And you're kind of subject to the scrutiny of 930 00:36:29,124 --> 00:36:31,369 the entire organization, to some degree, 931 00:36:31,369 --> 00:36:33,080 with all of your communication. 932 00:36:33,080 --> 00:36:35,125 Like kind of the challenging side of it. 933 00:36:35,125 --> 00:36:38,110 Then the good side is people are much more informed about 934 00:36:38,110 --> 00:36:38,930 what's happening. 935 00:36:40,212 --> 00:36:45,312 I guess I don't feel that I can give much 936 00:36:45,312 --> 00:36:50,510 stronger endorsement of it than it has worked so far. 937 00:36:50,510 --> 00:36:52,216 >> That's a pretty good endorsement. 938 00:36:52,216 --> 00:36:55,550 >> Yeah, I'm actually really curious how or whether it'll 939 00:36:55,550 --> 00:36:57,940 work when we're 5,000 people or something like that. 940 00:36:57,940 --> 00:36:59,950 If we're ever at that scale. 941 00:36:59,950 --> 00:37:01,110 >> I think the two things that have 942 00:37:01,110 --> 00:37:04,400 helped us scale it are one, changing the tools and 943 00:37:04,400 --> 00:37:06,060 two, developing the culture around it. 944 00:37:06,060 --> 00:37:08,010 And so on the tools front, you know, 945 00:37:08,010 --> 00:37:10,760 used to be the case, or the infrastructure, it 946 00:37:10,760 --> 00:37:13,670 used to be the case that you could keep abreast by what's 947 00:37:13,670 --> 00:37:15,700 happening in the company by reading all the email. 948 00:37:15,700 --> 00:37:18,480 Now we have weekly all-hands sometimes on the deck, 949 00:37:18,480 --> 00:37:20,720 and we actually have to put all this work into 950 00:37:20,720 --> 00:37:22,830 developing a deck to, to communicate to people what's 951 00:37:22,830 --> 00:37:25,640 going on in the company, since there's so much more. 952 00:37:25,640 --> 00:37:29,820 And the second is on the cultural side, so much 953 00:37:29,820 --> 00:37:33,660 information being available internally, you have to 954 00:37:33,660 --> 00:37:35,950 develop cultural norms around how it's treated. 955 00:37:35,950 --> 00:37:38,680 Obvious things like the fact that a lot of 956 00:37:38,680 --> 00:37:40,590 it's confidential to Stripe. 957 00:37:40,590 --> 00:37:44,260 But even less obvious things like when emailing someone 958 00:37:44,260 --> 00:37:46,190 or talking in Slack or 959 00:37:46,190 --> 00:37:48,860 IRESE, when that is viewed by now 170 people, 960 00:37:48,860 --> 00:37:52,524 it's pretty easy to get stage fright. 961 00:37:52,524 --> 00:37:55,970 And it's, it's pretty easy for what you thought was 962 00:37:55,970 --> 00:37:59,110 a reasonable proposal, you get this drive-by criticism. 963 00:37:59,110 --> 00:38:01,890 And you're now less likely to share in the future. 964 00:38:01,890 --> 00:38:05,200 And so, we've had to create norms around when it's 965 00:38:05,200 --> 00:38:06,650 reasonable to jump into discussions and 966 00:38:06,650 --> 00:38:08,010 how that interaction works. 967 00:38:08,010 --> 00:38:11,110 Because people are on the stage so much more. 968 00:38:11,110 --> 00:38:12,300 >> I'm pretty sure it's not good, 969 00:38:12,300 --> 00:38:14,530 but, not to put her on the spot but 970 00:38:14,530 --> 00:38:16,180 Emily interned at Stripe this summer. 971 00:38:16,180 --> 00:38:17,140 I'm curious, like, 972 00:38:17,140 --> 00:38:19,480 as an intern, what you thought of it? 973 00:38:19,480 --> 00:38:20,920 >> I think, overall, it's great. 974 00:38:20,920 --> 00:38:22,351 I think, like, my first week, 975 00:38:22,351 --> 00:38:24,259 I spent most of the time reading Hackpad and 976 00:38:24,259 --> 00:38:26,670 getting caught up on what the company was doing. 977 00:38:26,670 --> 00:38:28,870 It can often be quite distracting, 978 00:38:28,870 --> 00:38:31,460 from your own work, as there are oftentimes other parts 979 00:38:31,460 --> 00:38:33,330 of the company you're really interested in. 980 00:38:33,330 --> 00:38:35,050 >> Hackpad, by the way, is like Google Docs, but 981 00:38:35,050 --> 00:38:35,760 with a news feed and 982 00:38:35,760 --> 00:38:37,600 that way you can just like see all the documents. 983 00:38:37,600 --> 00:38:38,860 >> Yeah, and you're encouraged to make 984 00:38:38,860 --> 00:38:40,229 everything public, everything you work on. 985 00:38:41,410 --> 00:38:43,310 But overall, it gets you spun up really quickly. 986 00:38:43,310 --> 00:38:45,172 And we also have things called spin ups where 987 00:38:45,172 --> 00:38:47,990 every single leader of a team at the company, 988 00:38:47,990 --> 00:38:51,280 whether it's sys or it's product gives like a 30 989 00:38:51,280 --> 00:38:53,360 minute talk on what their team is currently doing and 990 00:38:53,360 --> 00:38:55,120 how you can contribute, if you're interested. 991 00:38:55,120 --> 00:38:56,925 >> Do you think email transparency was net good? 992 00:38:56,925 --> 00:38:59,982 >> Yes I remember having a hard time understanding what 993 00:38:59,982 --> 00:39:02,410 I should and should not subscribe too. 994 00:39:02,410 --> 00:39:05,480 And the first week having 2,000 emails in my inbox and 995 00:39:05,480 --> 00:39:07,480 then by the end there are three of four teams you 996 00:39:07,480 --> 00:39:11,940 actually want that information coming in from. 997 00:39:11,940 --> 00:39:13,903 >> All right, audience questions. 998 00:39:13,903 --> 00:39:14,971 Yes. 999 00:39:14,971 --> 00:39:17,974 >> So this question for Patrick and John. 1000 00:39:17,974 --> 00:39:21,766 Is it your experience that your early hires begin to 1001 00:39:21,766 --> 00:39:22,383 grow and 1002 00:39:22,383 --> 00:39:26,900 evolve into leadership roles as the company scales? 1003 00:39:26,900 --> 00:39:28,421 And a question for 1004 00:39:28,421 --> 00:39:33,330 Ben, how is the difference, what difference is today? 1005 00:39:33,330 --> 00:39:36,660 How is a difference in your initial bid run of 1006 00:39:36,660 --> 00:39:39,630 the product and the audience would be? 1007 00:39:39,630 --> 00:39:41,080 >> All right, we'll go first on. 1008 00:39:41,080 --> 00:39:42,858 The people, actually all three of you are welcome to 1009 00:39:42,858 --> 00:39:44,050 answer this, have the people that you 1010 00:39:44,050 --> 00:39:46,960 hired early been able to grow into leadership roles? 1011 00:39:49,200 --> 00:39:52,470 >> In the Stripes case, yes, in that quite a number of 1012 00:39:52,470 --> 00:39:55,410 the first ten people are in leadership roles now. 1013 00:39:55,410 --> 00:39:59,610 I think that's one thing that organizations, again 1014 00:39:59,610 --> 00:40:02,548 it's an unnatural skill that they have to get good at, 1015 00:40:02,548 --> 00:40:05,800 is realizing that people don't necessarily come out 1016 00:40:05,800 --> 00:40:09,140 of the womb being good at managing or being good at 1017 00:40:09,140 --> 00:40:13,420 leadership and being able to develop that in people, and 1018 00:40:13,420 --> 00:40:16,510 being able to help people progress as 1019 00:40:16,510 --> 00:40:18,880 they spend a number of years at the company. 1020 00:40:18,880 --> 00:40:22,450 It's a lot of work at a time when everyone is running 1021 00:40:22,450 --> 00:40:25,730 around with their hair on fire, but it's also damaging 1022 00:40:25,730 --> 00:40:27,130 if the company can't develop that skill. 1023 00:40:30,160 --> 00:40:30,740 >> Yeah, I think for 1024 00:40:30,740 --> 00:40:33,280 us the answer is some yes and some no. 1025 00:40:36,220 --> 00:40:37,320 I think one of the big benefits of 1026 00:40:37,320 --> 00:40:38,690 working at a start-up is that you can be 1027 00:40:38,690 --> 00:40:40,310 handed a challenge that no one else would be 1028 00:40:40,310 --> 00:40:43,340 crazy enough to give you the opportunity to take on. 1029 00:40:43,340 --> 00:40:45,160 And that could managing people. 1030 00:40:45,160 --> 00:40:46,970 It could be taking on a project. 1031 00:40:46,970 --> 00:40:49,520 But the implicit contract with that is that if 1032 00:40:49,520 --> 00:40:51,780 your going to ask somebody to take a really big risk on 1033 00:40:51,780 --> 00:40:54,720 that, it shouldn't be like one-way through the door and 1034 00:40:54,720 --> 00:40:55,380 if you don't succeed, 1035 00:40:55,380 --> 00:40:57,545 otherwise it creates fright to give it a shot. 1036 00:40:57,545 --> 00:41:00,730 So we have some folks that are managing large teams 1037 00:41:00,730 --> 00:41:03,800 that started as individual programmers, where they were 1038 00:41:03,800 --> 00:41:05,680 engineering and they said hey, I would love to try. 1039 00:41:05,680 --> 00:41:07,220 I'd love to try leading a project and 1040 00:41:07,220 --> 00:41:07,970 then leading a group and 1041 00:41:07,970 --> 00:41:08,940 then taking responsibility for 1042 00:41:08,940 --> 00:41:10,576 management, taking a group. 1043 00:41:10,576 --> 00:41:11,710 And we've had other folks that tried it and 1044 00:41:11,710 --> 00:41:14,130 they were like, I'm really glad I tried it because I 1045 00:41:14,130 --> 00:41:15,750 never wanna do that again. 1046 00:41:15,750 --> 00:41:17,740 We try to make sure that, for those people, like you 1047 00:41:17,740 --> 00:41:20,660 can have just as much impact at a company through your 1048 00:41:20,660 --> 00:41:22,510 individual contributions as an engineer or designer. 1049 00:41:22,510 --> 00:41:23,130 You don't have to manage. 1050 00:41:24,810 --> 00:41:25,640 But it's really hard to 1051 00:41:25,640 --> 00:41:27,102 predict until you give people a shot. 1052 00:41:27,102 --> 00:41:30,800 And so my strong preference is that you give as many 1053 00:41:30,800 --> 00:41:34,140 people a shot as possible, in the few areas where you 1054 00:41:34,140 --> 00:41:38,070 really feel like there's too much learning curve relative 1055 00:41:38,070 --> 00:41:39,630 to the business objective you're trying to achieve. 1056 00:41:39,630 --> 00:41:42,750 That's when you look for somebody who might be able 1057 00:41:42,750 --> 00:41:44,470 to walk in and really execute well in that job. 1058 00:41:46,420 --> 00:41:48,910 Are you going to answer the other question? 1059 00:41:48,910 --> 00:41:50,100 >> Oh, so the question was 1060 00:41:50,100 --> 00:41:52,170 how has the vision changed since we initially started? 1061 00:41:52,170 --> 00:41:59,364 >> Yes, >> Sure, 1062 00:41:59,364 --> 00:42:04,260 well I think on the vision, when we first started, 1063 00:42:04,260 --> 00:42:06,760 I think we started hiring very inductively. 1064 00:42:06,760 --> 00:42:07,300 We were like, oh, we're 1065 00:42:07,300 --> 00:42:08,420 gonna build this really cool tool. 1066 00:42:08,420 --> 00:42:09,740 People are gonna enjoy it. 1067 00:42:09,740 --> 00:42:10,690 I love collecting things. 1068 00:42:10,690 --> 00:42:12,670 Maybe other people like collecting things. 1069 00:42:12,670 --> 00:42:14,100 And what we didn't expect, 1070 00:42:14,100 --> 00:42:15,790 that kind of revealed itself early on, 1071 00:42:15,790 --> 00:42:19,750 was that looking at other people's collections turns 1072 00:42:19,750 --> 00:42:21,460 out to be this really amazing way to discover 1073 00:42:21,460 --> 00:42:23,730 things that you didn't know you were looking for. 1074 00:42:23,730 --> 00:42:25,839 It becomes kind of a solution to a problem that 1075 00:42:25,839 --> 00:42:27,790 a lot of other technologies don't have. 1076 00:42:27,790 --> 00:42:30,980 And so, over the last couple years especially, 1077 00:42:30,980 --> 00:42:33,340 we've poured enormous kind of technical and design 1078 00:42:33,340 --> 00:42:36,600 resources into building out recommendations products, 1079 00:42:36,600 --> 00:42:38,486 search products, feed products. 1080 00:42:38,486 --> 00:42:41,520 Leveraging the unique data that we have which 1081 00:42:41,520 --> 00:42:43,505 are these pins that were all hand-picked by someone and 1082 00:42:43,505 --> 00:42:45,190 hand-categorized. 1083 00:42:45,190 --> 00:42:47,980 And then on the audience side, the first big 1084 00:42:47,980 --> 00:42:51,140 surprise was truthfully when we first started, 1085 00:42:51,140 --> 00:42:52,610 we didn't really know if anyone would use it. 1086 00:42:52,610 --> 00:42:55,890 And we were just happy that anybody that wasn't related 1087 00:42:55,890 --> 00:42:56,570 to us and 1088 00:42:56,570 --> 00:42:59,930 obligated by familial relationship would use it. 1089 00:42:59,930 --> 00:43:03,190 And so one of the biggest surprises has been how many 1090 00:43:03,190 --> 00:43:05,662 people and how diverse those groups of people have been. 1091 00:43:05,662 --> 00:43:07,235 I think that's been one of 1092 00:43:07,235 --> 00:43:09,554 the things that's really exciting. 1093 00:43:09,554 --> 00:43:11,290 And, the funny thing is, 1094 00:43:11,290 --> 00:43:14,460 that is often like the company goes farther along, 1095 00:43:14,460 --> 00:43:16,162 your aspirations therefore get bigger. 1096 00:43:16,162 --> 00:43:19,440 So there's this gap that always exists, 1097 00:43:19,440 --> 00:43:20,840 I tell my team, between where we are and 1098 00:43:20,840 --> 00:43:21,910 where I think we should be, and 1099 00:43:21,910 --> 00:43:24,100 even though objectively, we're much farther along, 1100 00:43:24,100 --> 00:43:26,390 I feel like the gap has widened even farther. 1101 00:43:26,390 --> 00:43:30,290 But I think that's a really common trait among 1102 00:43:30,290 --> 00:43:32,795 people who found companies. 1103 00:43:32,795 --> 00:43:34,197 >> Yes? 1104 00:43:34,197 --> 00:43:35,348 >> You said that like, 1105 00:43:35,348 --> 00:43:37,986 while selling a vision of the company you had to 1106 00:43:37,986 --> 00:43:38,731 describe and 1107 00:43:38,731 --> 00:43:41,340 go into detail that how hard it is going to be. 1108 00:43:41,340 --> 00:43:43,290 And like they won't see their families for 1109 00:43:43,290 --> 00:43:45,650 three years but then they will also get a thing that 1110 00:43:45,650 --> 00:43:47,610 their grandchildren can be proud of. 1111 00:43:47,610 --> 00:43:49,050 But how do you really know? 1112 00:43:49,050 --> 00:43:51,010 Both of you know that the second part is 1113 00:43:52,180 --> 00:43:53,680 not guaranteed but the first part, 1114 00:43:53,680 --> 00:43:55,200 if they're not gonna see their family for 1115 00:43:55,200 --> 00:43:57,380 the next year, it's gonna be a different show. 1116 00:43:57,380 --> 00:43:59,800 So like, how can you be authentic by selling that 1117 00:43:59,800 --> 00:44:02,514 vision of the company or is it about giving high returns 1118 00:44:02,514 --> 00:44:04,992 for the high risk they are taking when the number of 1119 00:44:04,992 --> 00:44:08,331 equity that already >> Great question. 1120 00:44:08,331 --> 00:44:09,565 So the question is, 1121 00:44:09,565 --> 00:44:12,090 most start-ups are not the iPhone. 1122 00:44:12,090 --> 00:44:13,810 You can't guarantee that people's grandchildren 1123 00:44:13,810 --> 00:44:15,820 are gonna remember this because most start-ups fail. 1124 00:44:15,820 --> 00:44:17,180 How do you convince people to sort of 1125 00:44:17,180 --> 00:44:19,055 make sacrifices to join a start-up? 1126 00:44:21,510 --> 00:44:25,584 >> I think, part of the way in which it resonates with 1127 00:44:25,584 --> 00:44:29,660 people is because it's not guaranteed, right? 1128 00:44:29,660 --> 00:44:31,720 If it were guaranteed then it'd be boring. 1129 00:44:31,720 --> 00:44:36,784 And so it's that there is the prospect of affecting 1130 00:44:36,784 --> 00:44:42,219 this outcome, but nothing more than that potential. 1131 00:44:42,219 --> 00:44:45,236 To the not seeing their families or 1132 00:44:45,236 --> 00:44:49,709 kids, start-ups often do involve longer hours in 1133 00:44:49,709 --> 00:44:52,945 the beginning, but I think, well. 1134 00:44:52,945 --> 00:44:55,758 I, I think that that particular story is 1135 00:44:55,758 --> 00:44:58,160 probably somewhat overstated. 1136 00:44:58,160 --> 00:44:59,650 I guess, it was, I think 1137 00:44:59,650 --> 00:45:02,490 Scott Forestall was trying to recruit these people. 1138 00:45:02,490 --> 00:45:05,830 But I mean, even though startups, 1139 00:45:05,830 --> 00:45:06,790 especially in the earlier days, 1140 00:45:06,790 --> 00:45:08,830 tend to involve somewhat longer working hours. 1141 00:45:08,830 --> 00:45:10,950 I think it was kind of this tendency to, 1142 00:45:10,950 --> 00:45:11,715 to exaggerate it. 1143 00:45:11,715 --> 00:45:12,320 And sort of, 1144 00:45:12,320 --> 00:45:14,550 I mean it's like the startup version of fishing. 1145 00:45:14,550 --> 00:45:17,460 Like, every startup thinks they worked even more insane 1146 00:45:17,460 --> 00:45:18,570 hours than, you know, 1147 00:45:18,570 --> 00:45:20,530 the next one back in the early days. 1148 00:45:20,530 --> 00:45:23,450 Like, we literally never sleep or slept rather, 1149 00:45:23,450 --> 00:45:24,810 for two years. 1150 00:45:24,810 --> 00:45:27,330 So I don't know. 1151 00:45:27,330 --> 00:45:29,980 I think that realistically for most people, 1152 00:45:29,980 --> 00:45:32,350 it's not that big a sacrifice, right? 1153 00:45:32,350 --> 00:45:35,710 You're, maybe on average, being really realistic about 1154 00:45:35,710 --> 00:45:37,842 it, you'll work two hours more on average per day. 1155 00:45:37,842 --> 00:45:39,570 It's certainly a sacrifice, but 1156 00:45:39,570 --> 00:45:44,770 it is not forgoing all you know, pleasure and 1157 00:45:44,770 --> 00:45:47,820 enjoyment for the next half decade. 1158 00:45:47,820 --> 00:45:49,450 >> Yeah, I mean, I think even the iPhone 1159 00:45:49,450 --> 00:45:53,600 wasn't the iPhone before it got done, right? 1160 00:45:53,600 --> 00:45:55,890 I mean, no smart person you're hiring is 1161 00:45:55,890 --> 00:45:57,890 under the illusion that you have a crystal ball into 1162 00:45:57,890 --> 00:46:00,170 the future that only you have, and 1163 00:46:00,170 --> 00:46:01,550 that joining is a guaranteed thing. 1164 00:46:01,550 --> 00:46:03,590 And in fact, if you're telling him that and 1165 00:46:03,590 --> 00:46:06,520 they select in, you probably shouldn't hire them. 1166 00:46:06,520 --> 00:46:07,452 Because they, 1167 00:46:07,452 --> 00:46:10,394 they didn't pass like a basic intelligence test 1168 00:46:10,394 --> 00:46:12,992 about uncertainty in the future. 1169 00:46:12,992 --> 00:46:15,436 But I, but I think it's fair to say, like you know, 1170 00:46:15,436 --> 00:46:17,883 what's exciting and where you think you can go, and 1171 00:46:17,883 --> 00:46:20,920 where it's going to be hard and chart your best plan. 1172 00:46:20,920 --> 00:46:24,144 And then tell them why their role in it can be 1173 00:46:24,144 --> 00:46:26,930 instrumental, because it is. 1174 00:46:26,930 --> 00:46:29,090 You know I really liked what you said. 1175 00:46:29,090 --> 00:46:31,340 You know, if you tell people like hey we're gonna go 1176 00:46:31,340 --> 00:46:33,750 to Mars, it attracts the best people and 1177 00:46:33,750 --> 00:46:36,250 then you're incrementally closer to getting to Mars. 1178 00:46:36,250 --> 00:46:37,960 And they know that going in. 1179 00:46:37,960 --> 00:46:39,630 What I would discourage doing is 1180 00:46:39,630 --> 00:46:40,720 just whitewashing all of that. 1181 00:46:42,010 --> 00:46:45,250 And if people kind of are joining because they want 1182 00:46:45,250 --> 00:46:47,200 sort of oh, I want all this certainty and guarantee of 1183 00:46:47,200 --> 00:46:49,520 working at Google, plus like the perk of working in 1184 00:46:49,520 --> 00:46:52,560 a small start up and more email, and transparency. 1185 00:46:52,560 --> 00:46:53,440 Like that's, that's a really, 1186 00:46:53,440 --> 00:46:54,365 really negative sign. 1187 00:46:54,365 --> 00:46:57,210 And for 1188 00:46:57,210 --> 00:47:00,330 example, when I interview people they'll often say oh, 1189 00:47:00,330 --> 00:47:01,620 I'm really passionate at what you're doing, and 1190 00:47:01,620 --> 00:47:03,030 then I'm like well, what else are you interviewing? 1191 00:47:03,030 --> 00:47:05,340 And then they'll just list seven companies that have 1192 00:47:05,340 --> 00:47:07,580 nothing to do with each other except they're sort of 1193 00:47:07,580 --> 00:47:08,861 at the same stage we're at. 1194 00:47:08,861 --> 00:47:09,916 They're like, you know, 1195 00:47:09,916 --> 00:47:11,471 I love the problem of discovery, so 1196 00:47:11,471 --> 00:47:15,060 I'm interviewing at Stripe, Dropbox, Airbnb, Uber. 1197 00:47:15,060 --> 00:47:17,200 Now I'm also putting into my resumé into Google X, 1198 00:47:17,200 --> 00:47:18,730 into that part of the division. 1199 00:47:18,730 --> 00:47:19,960 And that's a sign that they're probably not 1200 00:47:19,960 --> 00:47:21,420 being authentic with which they care about. 1201 00:47:21,420 --> 00:47:21,940 >> Yeah. 1202 00:47:21,940 --> 00:47:24,660 >> And those folks often, when things get really hard, 1203 00:47:24,660 --> 00:47:26,855 they won't stick it out and work through it. 1204 00:47:26,855 --> 00:47:28,480 Because they were really signing up for 1205 00:47:28,480 --> 00:47:31,520 an experience, not for achieving a goal. 1206 00:47:31,520 --> 00:47:34,650 >> I think the other thing that motivates people 1207 00:47:34,650 --> 00:47:37,650 a great deal, in addition to the prospect of sort of 1208 00:47:37,650 --> 00:47:40,797 them you know, affecting some outcome is just sort of 1209 00:47:40,797 --> 00:47:42,680 the personal development angle. 1210 00:47:42,680 --> 00:47:43,340 And that a start up, 1211 00:47:43,340 --> 00:47:47,170 just because it's much, kind of more lightly staffed. 1212 00:47:48,920 --> 00:47:50,280 It's much less forgiving, right? 1213 00:47:50,280 --> 00:47:51,038 In that, like, 1214 00:47:51,038 --> 00:47:53,099 even if you're the best person in the world, 1215 00:47:53,099 --> 00:47:54,249 if you're not going to. 1216 00:47:54,249 --> 00:47:55,687 Well, whether or not you're the best or 1217 00:47:55,687 --> 00:47:57,550 the worst person in the world, you're probably not 1218 00:47:57,550 --> 00:47:59,835 going to significantly alter Google's trajectory, right? 1219 00:47:59,835 --> 00:48:02,860 Whereas if you sort of, 1220 00:48:02,860 --> 00:48:05,810 really wanted to benchmark yourself, and 1221 00:48:05,810 --> 00:48:08,880 see how much a contribution and an impact you can make. 1222 00:48:08,880 --> 00:48:10,990 And I think that prospect is quite compelling to a lot of 1223 00:48:10,990 --> 00:48:13,350 the best people, then a start up is 1224 00:48:13,350 --> 00:48:18,120 a much better place to, to go test that. 1225 00:48:18,120 --> 00:48:18,660 >> All right. 1226 00:48:18,660 --> 00:48:20,850 >> Last question, who'd you want? 1227 00:48:20,850 --> 00:48:22,280 Last question. 1228 00:48:22,280 --> 00:48:23,670 So this question is for 1229 00:48:23,670 --> 00:48:25,880 Ben, but you guys can all answer it. 1230 00:48:25,880 --> 00:48:28,590 But how has your user base effected your 1231 00:48:28,590 --> 00:48:29,540 hiring strategies? 1232 00:48:29,540 --> 00:48:32,759 So Pinterest is a site that's used like 80% by 1233 00:48:32,759 --> 00:48:33,569 women, so 1234 00:48:33,569 --> 00:48:36,700 how did that effect your initial hiring decisions? 1235 00:48:36,700 --> 00:48:38,020 >> How has your user base effected your 1236 00:48:38,020 --> 00:48:38,560 hiring strategy? 1237 00:48:40,340 --> 00:48:41,690 >> Yes, so, You know, 1238 00:48:41,690 --> 00:48:44,040 conventional wisdom is like you only hire people that 1239 00:48:44,040 --> 00:48:46,820 religiously use your product every single day. 1240 00:48:46,820 --> 00:48:48,010 And that probably works really well if 1241 00:48:48,010 --> 00:48:50,272 you're making an API, probably amazing. 1242 00:48:50,272 --> 00:48:54,930 For us, we screen for people that are ambitious and 1243 00:48:54,930 --> 00:48:58,780 excited about the vision of cracking discovery online. 1244 00:48:58,780 --> 00:49:00,900 And they have to know exactly how 1245 00:49:00,900 --> 00:49:02,720 our service works and have to have used it. 1246 00:49:02,720 --> 00:49:04,490 But they may not be a lifelong user. 1247 00:49:04,490 --> 00:49:06,770 And that, for us it's this great opportunity because we 1248 00:49:06,770 --> 00:49:08,700 can be like, what is the barrier that's 1249 00:49:08,700 --> 00:49:10,270 preventing you from using it? 1250 00:49:10,270 --> 00:49:12,100 Come join, remove that barrier, 1251 00:49:12,100 --> 00:49:15,303 and help us get closer to that vision. 1252 00:49:15,303 --> 00:49:17,860 I don't know, there's a lot of, 1253 00:49:19,000 --> 00:49:20,420 if you read any sort of book, there's all this 1254 00:49:20,420 --> 00:49:23,050 startup wisdom that sounds like really reasonable. 1255 00:49:23,050 --> 00:49:25,200 But it's only useful if it works in 1256 00:49:25,200 --> 00:49:27,060 your particular circumstance. 1257 00:49:27,060 --> 00:49:28,340 And so for us, we've had to sort of 1258 00:49:28,340 --> 00:49:31,360 broaden the lens a little bit in looking broadly about 1259 00:49:31,360 --> 00:49:33,169 people that are ambitious about the mission. 1260 00:49:33,169 --> 00:49:34,620 That care about the product and 1261 00:49:34,620 --> 00:49:36,290 our approach to building products, 1262 00:49:36,290 --> 00:49:38,509 even if from day one, they weren't our earliest users. 1263 00:49:40,040 --> 00:49:44,140 >> The one thing I'll just tack on to that is that when 1264 00:49:44,140 --> 00:49:45,830 we talked about the fact that, you know, 1265 00:49:45,830 --> 00:49:48,500 it's really hard to hire for those early employees. 1266 00:49:48,500 --> 00:49:49,760 And you know, you have people who have 1267 00:49:49,760 --> 00:49:50,880 other good options. 1268 00:49:52,230 --> 00:49:54,890 You're very much at the ugly duckling stage. 1269 00:49:54,890 --> 00:49:56,714 Finding people who are passionate about your 1270 00:49:56,714 --> 00:49:58,399 product can be a great way to find people. 1271 00:49:58,399 --> 00:50:01,888 Because there you kind of have an unnatural advantage 1272 00:50:01,888 --> 00:50:03,443 over other companies. 1273 00:50:03,443 --> 00:50:04,784 And so I know, for 1274 00:50:04,784 --> 00:50:09,258 sure, in Stripe's case, we definitely, we hired I think 1275 00:50:09,258 --> 00:50:13,690 it was four Stripe users in the very early days. 1276 00:50:13,690 --> 00:50:16,150 And you know, those were people who, 1277 00:50:16,150 --> 00:50:18,540 who we probably couldn't have gotten otherwise. 1278 00:50:18,540 --> 00:50:21,196 I'm sure it was the same in Pinterest's case, 1279 00:50:21,196 --> 00:50:23,830 where you'll get all this benefit of working at 1280 00:50:23,830 --> 00:50:25,630 Pinterest and hey, it's Pinterest. 1281 00:50:25,630 --> 00:50:27,170 And you get to work where you pin. 1282 00:50:27,170 --> 00:50:29,469 Yeah. 1283 00:50:29,469 --> 00:50:30,618 >> Thank you guys very much for coming in today. 1284 00:50:30,618 --> 00:00:00,000 >> Thank you.