[0:00:06] So, as it's been established, I'm Greg. I've been working at Stripe for about two years. I guess my story is I was at Harvard, transferred to MIT, ended up dropping out of MIT. And I've really been working on recruiting at Stripe since I showed up. When I showed up, Stripe really did not look pretty and we really needed a designer. And so I started to lead a search to try to find one. And it was pretty spectacularly unsuccessful. It took us a long time before we were able to find anyone who was...
[0:00:34] Who was willing to come on board and do design with us. But that was a really interesting learning experience for me. Just like doing a lot of things wrong and realizing what we could have done better. And I think in general that recruiting is just an interesting problem because everyone has the same set of problems but you look at a company that's done it well and it's just a black box. You see who's working there but it's really hard to see how did they actually recruit those people. And so today I hope to kind of pry open the box.
[0:01:03] And tell you about how we think about recruiting at Stripe and how we've approached that. So recruiting for us was really ad hoc for the first 10 people. It was mostly that we knew people, we talked to our friends, and we kept nagging them until occasionally one would decide to join. And that got us a decent...
[0:01:22] A decent way. But once we were 10 people, we started thinking, wow, there's no way that we could possibly have enough friends to continue this and to hire more people. So I guess we should go and try to figure out whatever it is that every other company does to scale. And so we started doing the career fair circuit because that's the thing you're supposed to do. We went around to a bunch of different schools and came home with 300 resumes. And so we looked at the stack and we were like, what we've been doing to date is just not going to scale.
[0:01:47] to screening and interviewing 300 people. We need to actually figure out how to do this. And so we spent a bunch of time trying to divide an interview process in a way of filtering people. And so we worked really hard, and we went through these 300 resumes, and at the end, we ended up with a grand total of zero hires from that. So here's a graph of where everyone who's at Stripe today came from. As you can see, basically the only thing that's worked really, really well for us has been referrals. The thing that we missed
[0:02:16] When we were initially trying to figure out how are you going to get more than 10 people is that every time you hire someone, that person comes with their own network, their own set of people that they know. Every time you talk to someone, like a lot of the people that you just know also know people. And so through internal and external referrals, you can actually scale a lot further than you think if you really put a lot of effort into it. And I'll get more into that later.
[0:02:37] Another thing that's been mildly successful for us has been Outbound. And so this is mostly people in the Stripe community. So we're a developer-focused product. And so a lot of those people that we interact with every day through support or other mechanisms are actually people that we'd like to hire. We also have run a bunch of events and things like our online Capture the Flag competition, which have been really good for identifying people. We haven't put that much time or effort into Outbound. And so the fact that it's actually been very successful is pretty promising to me.
[0:03:07] The other side of the spectrum is inbound. So this is people seeing your jobs page and sending it in an email. That is by far the place we've spent the most time, that we've put a ton of effort into screening and interviewing those people, and we just haven't gotten return there. So something to keep in mind is that inbound is just a really easy thing to see and to collect, but it's also just not going to be very fruitful.
[0:03:30] And then finally, the recruiter pipeline is an interesting one, because a lot of people think, oh, what's the right way of getting people? Well, recruiters, their job is to get you people. And that's been something that's been spectacularly unsuccessful for us. We've hired exactly one person through a recruiter, and even that was a bit of a special case. That was Hacker School, if you're familiar with that.
[0:03:51] It's funny because I probably get one to two emails from a recruiter a day saying, I've got this candidate I want you to take a look at. And I used to think, wow, look at this. Here's a free source of resumes. Here's a free source of people. I should just go and look at this resume and maybe it'll turn out to be good. And it would be just a waste to not do anything with this.
[0:04:11] And so I did that for a long time, and we just never were getting anyone who was really good. We were getting people who looked plausible, and maybe we'd get them on the phone and then talk to them, but they just would never turn out to be someone that we really, really wanted to hire. And I still kept doing this thinking that maybe it's okay because eventually we'll get someone, because this seemed somewhat promising, until we started hearing that one of the recruiters that we had done this to, just saying, well, take a look at a resume, but that's it, had started recruiting in our name amongst people that we knew.
[0:04:38] And so I think that the purpose there was that person was going to funnel those people back. So there is some brand risk associated with going down the recruiter route. And you should make sure that if you're going to do that, that you're aware that you're acquiring that. I think the biggest takeaway for me from looking at this is that there are high-value pipelines, but there are high-quality pipelines. And these are often very, very different things. At one point, I was looking through probably hundreds of resumes a week.
[0:05:05] Just going through various resume stacks that we obtained through either our VCs or other sources. And we never ended up hiring anyone there. But at the end of each week, I would look back and say, what did I do this week? Well, I looked through hundreds of different people. That is so successful. I'm doing such a great job of recruiting. But at the end of the day, the only thing that matters is how many people you're hiring. And so it's really important that you pay attention to how are you spending your time and how well is that converting into actual hires.
[0:05:36] For us, there's no question that the most successful thing we've done is talking to people that we already know. And if that's what you're going to do, then there's the question of, okay, I already know these people, but how do I get them to be interested in my company? Because really, anyone you talk to who's good is going to already be doing something and is going to seem pretty entrenched in that. It's just never the right time. I would say that the main thing we do is we just try to start out very casually. I think that the most successful recruiting tactic that we have is just adding someone on IM.
[0:06:05] Once you have someone on IM, you start just having casual conversations and things about your company naturally come up and that's a really great way to get someone excited and to open the door to a more formal conversation. Whether it's going to coffee with people or just happening to be in their town when they say that they're going to be there and you magically schedule a flight without telling them that's what you're doing, that's also worked out pretty well for us. You really have to be willing to play the long game though.
[0:06:32] If you're going to try to recruit people that you already know and you're willing to invest that effort, you have to be willing for it to take a long time. There's really no way to rush it. I think that our longest recruitment to date was someone that we interviewed a year before we ended up hiring him. So we actually interviewed him, really liked him, and then his company got acquired. And the entire time that he was with the acquiring company, we were constantly in touch saying, hey, we're still interested. We'd love to chat whenever you're ready.
[0:06:59] And then at the end of the year, he was ready to go, and there we were waiting for him. Another thing that I found has worked pretty well is just, again, a very casual context. So maybe there's an email thread going with the person, sticking in a PS line of, hey, I'd really love to work with you if you're ever looking for something, and just leave it at that. And it turns out that even though people are very good looking, everyone is at some point looking. And if you're already there in someone's mind, then that gives you the opportunity.
[0:07:28] To go and recruit them when they're actually ready. So that's how you find people. Keep in mind that up to a certain scale, you mostly are going to be looking for people you already know. And then the question just becomes, how do you actually get these people to join your company once you've got them interested? So I think that the largest piece of general advice that I have is to just think about recruiting the same way you think about marketing. At the end of the day, they're both doing the same thing. You're trying to get people to be really interested and excited about your company.
[0:07:58] The kinds of things that we do is that a lot of our recruitment email, basically every recruitment email, ends up going up on Hackpad or some collaborative editor, and then a bunch of people will work on that. And definitely the best recruitment emails that we've sent have been ones where we put this on the editor, we have like five different people go and edit it and really tune it and try to craft it and try to get someone really, really excited about Stripe. Another place that a lot of people, I think, don't pay attention to but is really important is the rejection email.
[0:08:26] You're almost certainly going to be sending a lot of these. And a lot of the people that you're rejecting are people who are almost good enough to work at your company, but just not quite the right fit. But they almost certainly have lots of friends who are. And so you should make sure that your rejection email is something that leaves people very excited about your company to the point where they're very excited to refer their friends. And that's something that we've seen. And one way to measure this is what are people saying after you reject them?
[0:08:54] We over time have had various rejection emails. Some of them have just like no one ever replies to them. And some of them people reply saying, hey, thanks a lot. This means a lot to me that you sent this one. So that's a very good metric.
[0:09:07] And I guess one thing that you should be really careful about, and it's a fairly obvious thing, but a lot of people get this wrong, is be very careful about things like typos in emails. There's nothing better to alienate a candidate than that. Also keep in mind that things like HTML email is a really great way to signal that you don't really understand who you're talking to. Similar to that is that in everything you do, whenever you're talking to candidates, whenever you're posting something online, that you are
[0:09:37] You have a bunch of subtle signals that you may not be aware of that you're sending to people. The kinds of things that you have to worry about are, one, don't post any of the Rockstar, Hacker, Engineer, Ninja postings. Those just make you sound like you're clueless. Similarly, the ones that are of the form, we only hire the best engineers, like everyone says that, that's completely useless, and anyone who's clueless will understand that.
[0:09:58] There are also ones that are less obvious. If you post about how you need people who are willing to dedicate their entire lives to this company, you're basically only going to be appealing to people in their mid-20s. And so it's very important to realize that things that you're not intending to leak through are going to. But this also extends to the actual interview process and all the way up to when you're actually making an offer to someone. Be very careful about everything, about how you're exposing things to the candidate.
[0:10:27] I think transparency is something that's a really great way to convey to a candidate that you're a great place to work. So when you're making an offer to someone, being very upfront about how you calculated the numbers and doing things like telling them the number of outstanding shares, which is something that actually a lot of startups that I've talked to do not do. And more generally, I think that your job when you're recruiting someone is to prove to them that they're going to be really happy working at your company. At something like a startup,
[0:10:55] That that's really what you're trying to sell. There are two parts. There's the vision that you're going to build something big, and also that this is just a really great place to work. And if a candidate doesn't leave your process being absolutely convinced that this is a place they're going to be happy, then it's going to be way harder to hire them. And so the kinds of things that you can do here are, I mean, I guess things that we have done.
[0:11:18] We had a candidate coming out who could only come on a weekend. And so we all got together for brunch, met up with a candidate, and then met up individually afterwards. And he remarked afterwards that this to him was a sign that we really wanted him, that we were all willing to take time out of our own days to come and meet with him, and that we were all so excited about the company that we're building and the culture that we're building that we're willing to do that, and that was something he wanted to be a part of. Other things that you can do, I guess when I was first flying out,
[0:11:44] Thinking about dropping out, I asked, so what hotel should I book? To which the reply was, well, you could book a hotel or you could just come crash with us, which to me was this great sign that they wanted to get to know me as a person and that that's the kind of company I wanted to work with is something that cared about me as more than just an engineer. And so I actually did that. It was a great bonding experience. We actually ended up stopping to do that after we had a candidate coming out for two days who was like, well, I guess I'll stay in a hotel for one night and then stay with you guys the other night.
[0:12:13] Which was a signal to me that they felt really pressured to do it, which was definitely not what we were going for. So you can only take these things so far sometimes. So interviewing is an interesting one. At the end of the day, it's really easy to tell if someone's not good or good. But it's a lot harder to tell if someone's good or if they're actually great. And that's what your interview should be focused on, is trying to distinguish the good from the great.
[0:12:43] One thing I think a lot of people get wrong is taking circumstantial evidence and taking that as authoritative. So maybe this person worked at Google on this project. Therefore, they're going to know a lot about distributed systems and be able to build really cool things. Every time we've made that assumption, we've turned out to be wrong. I think that it just does not correlate well with actual success because there's so many factors at play there. When you're working at Google, you have access to all this infrastructure. You don't have to have any idea how it actually works. You just magically plug into it and it happens.
[0:13:11] Instead, what you should do is, if someone's worked on a cool project, try to find someone who worked on that project with them and get a better sense of what they were actually working on, what it was like working with them, how many of the ideas were actually theirs, and things along those lines. Personal references are an incredibly high signal. Basically, whenever we get a candidate in the door, the first thing we do is try to find someone who's worked with them. Similarly, I think that
[0:13:39] The one thing that's really easy to do wrong with interviews is to think that interviews are like, I'm assessing this person's general intelligence, or I'm just trying to figure out if this person's really impressive. But at the end of the day, what you're actually trying to do is figure out, how good will this person be at this job? And so there's really no point in asking brain teasers that have nothing to do with a job. Instead, you should focus on things that you think will be strongly correlated with performance.
[0:14:02] And so for us, the high signal interview that we do is a collaborative hack project where we sit down with the candidate and pair program something. And usually we pick the project in advance and tune it to be something that's well suited for their interests and skill set. And then at the end, we walk away and think about what was the working dynamic like.
[0:14:22] What was the quality of the code that was produced? How well did the thing work? Did they kind of have good ideas about what to build? Is this the kind of person that we could trust to go off in a corner and build not just good things, but build things in the right way and choose the right problems to work on? And that basically everything else that we do is just supplementary info, but the hack project for us is the core thing that just tells us everything that we need to know. So one thing that happened early in the history of Stripe
[0:14:53] was that we were interviewing this engineer who was just clearly way better than any of the candidates we had talked to. He was incredibly talented, very, very technically good, and just was able to build really awesome things during the hack project. It was very clear that he'd be able to churn out code and be able to just fix a lot of the problems that we were facing at the time.
[0:15:15] The one problem was after we went and hung out with him for a while, we realized that we weren't that excited about working with him. We would much rather that we put him in a box and just let him turn out code rather than have to actually interact with him on a day-to-day basis. And then there's a big dilemma there. In that case, what do you do? Do you hire this person? How can you justify not hiring this person? And for us, the big realization was if we hired this person, we would be less happy.
[0:15:44] And that's significant. But even more significant is it would be way harder for us to hire really good people in the future. And so we ended up not hiring that person. And as a result, we devised what we call the Sunday test, which is if this person were alone in the office on a Sunday, would that make you more likely to come in and want to work with them? If the answer is not a clear yes, then we don't make a hire. And that's been a hard and fast rule for us, and I think it's been really, really successful. Now, if you're going to go down that route,
[0:16:13] You have to be really okay with missing out on great people because there's no question that we suffered in the short run because we didn't hire this person and that in general there are lots of pipelines that we just don't look at anymore because we know that there are great people there but it's just not worth our time. So college career fairs are a great example of that where there are amazing people coming out of college but in order to find them you have to sift through all these people who are just not going to be fit for your company.
[0:16:39] If you're serious about hiring only really, really good people, you have to be okay with missing out on them. That's just the cost of doing business. But one thing that I've found that's really useful in recruiting is trusting your instincts. It turns out that certainly in figuring out if someone's not good, that your instincts are basically always right. I've sort of gone through the same scenario dozens and dozens of times where there's someone we're talking to who seems really promising,
[0:17:10] But we talked to them and they just didn't seem that impressive. And then we think, well, maybe we just didn't ask them the right questions. And then we go and we do more evaluation, more evaluation, and it turns out that our initial instinct was just correct. And so something that I've realized is that if you think someone's not that good, they almost certainly are not. And so that's something to keep in mind that takes a while to see. Another piece of just hiring strategy for us has been...
[0:17:36] Whenever we're hiring the first person for a particular role or division, we always hold out for the person who's good at all sub-pieces there. One way you can tell that you've done a good job of hiring that really awesome person is that for us, our BD recruiting strategy right now is that we just get our BD candidate into the same room as our head of BD, and they will walk out being like, wow, I really want to work with that person.
[0:18:01] And so that's something that you should really aim for. There's a real benefit to hiring that person. Because at the end of the day, the first person you make there is going to be an anchor for what you're able to hire in the future. And the flip side is that...
[0:18:14] It's really painful because you have to wait a lot longer before you're able to hire that person. And so actually until like two weeks ago, Stripe did not have anyone working on sales. We had no salesperson because we were trying to find someone who could do both the day-to-day execution of going out and selling to customers, but also the higher level strategy of figuring out like who we should be selling to, how should we be building this team. Turns out those people are really hard to find. And then in the meanwhile, it was just really painful because we had no one working on this full time.
[0:18:42] I think my last piece of advice is that you should be thinking about hiring people rather than roles. So don't think of it as, I've got this opening that I need to fill. I need to find someone who fits this criteria. Because that's a great way to hire people who are in this narrowly defined box. And maybe your needs will change, and it turns out that this isn't actually the right person to be at your company. And one thing that's worked out well for us is hiring people who weren't quite sure where they'd actually fit into the organization. So for our COO, actually,
[0:19:12] Basically, everyone we talked to who had worked with him was like, you need to hire this guy. That would be insane if you did not hire him. And when we talked to him, we really, really liked him. But we just weren't really sure what he would do at the time. But we could point to at least one thing that he would do. It didn't seem like a full-time thing, but we knew there was an existence proof of him being useful.
[0:19:33] And the thing is that as a startup, if you could think of one thing that this person could do, there are probably 10 more things you're just not thinking of they could do. And two months from now, you're going to have 100 things that really, really need doing that this person would be a great fit for. And so you should be willing to hire great people as long as you can kind of see where they might fit. And I guess that I think of recruiting as less of an act and less of a job and more of a mindset.
[0:20:00] As I go through the world, I have this background process constantly going and thinking about, is this someone that I want to work with? How can we think about this in a recruiting context? And so I think that's something that's really important. Great. Thank you very much.