Luc Levesque: The Biggest Mistakes Founders Make When Hiring for Growth | 20VC #906

1. 0:00 – 15:00

   Three, two, one, zero.…

   1. HSHarry Stebbings0:00

      Three, two, one, zero. You have now arrived at your destination. Luke, this is such a joy to do. I heard so many great things before the show. I so enjoyed our chat before the show. So thank you so much for having, like, the time to come on the show, and I really appreciate you joining me.

   2. LLLuc Levesque0:20

      Tha- thanks, Harry. Thanks for having me.

   3. HSHarry Stebbings0:22

      Not at all. I want to start there. You've got a pretty fascinating background, so I want to start with how you made your way into the world of some of the most powerful growth orgs with the likes of Facebook and most recently Shopify. And what was that entry into growth for you?

   4. LLLuc Levesque0:37

      So I originally started as an entrepreneur, kind of bouncing between product and growth for a really long time, which I think is pretty common for a lot of growth people. I originally started building my first product when I was 21. I built a travel blogging network called TravelPod that I sold to TripAdvisor. And when I did TravelPod, I really focused on building the best product in the world and not on growing it, which, you know, had its- its own issues, uh, later that I discovered. But what happened was during the acquisition process with TripAdvisor, the CEO of TripAdvisor, I remember he was asking these really deep technical questions like, "What are your meta tags? What- what does the SCO configuration look like? How much traffic do you get from different sources like Google or Google Images?" And I was really thinking, "Why is this guy, the CEO of TripAdvisor, asking so many technical questions about SCO?" And this is, you know, quite a while ago, and back when SCO was a dark art, and growth in general. But that's a moment where I realized how important growth was and really realized that building a great product is important, but if you don't have a strong growth loop for it, it doesn't matter 'cause you can't scale a product. So from that, that was a pretty pivotal moment for me, and one where I really started focusing on viral loops and SCO and paid search and funnel optimization to grow the products that I was building. So after TravelPod, built a Facebook app called the Traveler IQ Challenge. That grew to about 100 million users on the Facebook platform. Also built a travel photo product called TripWow, which grew to 65 million users. Those products were- were very much about viral loops and- and- and growth on the Facebook platform. After that, moved on to growing TripAdvisor itself, which was mainly about SCO. We built a pretty sophisticated SCO growth engine and grew TripAdvisor to be the largest, uh, site on the internet, travel site on the internet, uh, which was a ton of fun. Um, after that, went on to advise a bunch of companies. So I worked with Pinterest, uh, Patreon, Twitter, and shared some of my- my growth playbook with them. And kind of through that, got connected with Mark at Facebook. And, you know, met with Mark, had a- a 15-minute meeting that turned into an hour meeting, and really hit it off, frankly, and, uh, had a- a great chat. And, you know, originally, he was trying to convince me to come down to California to work at Facebook. I was in Canada, wasn't, uh, all that, um, not- not so much interested, but it was a difficult leap to- to make, uh, to leave Canada. So it took a while, uh, and, you know, had a- a long kind of discussion with him. He... The experience was actually pretty im- im-, like, it was quite the experience. We, flew me out to- to California with my wife, we had dinner with him and his wife, and we ended up playing Oculus Rift all night actually. And this is before the Oculus was- was released, so it's pretty cool. So I was, you know, just really having a good time there. And I remember after doing that, he just pulled an envelope out of his pocket with an offer, and I flew back home with an offer to join Facebook, and eventually ended up leaving Canada, coming down to California and working at Facebook.

   5. HSHarry Stebbings3:34

      Can I just unpack a couple of things here?

   6. LLLuc Levesque3:36

      Yeah.

   7. HSHarry Stebbings3:36

      Because there's so much and I don't want to lose, like, strings.

   8. LLLuc Levesque3:39

      Uh-huh.

   9. HSHarry Stebbings3:39

      First off, before we touch on Facebook, you mentioned the viral loops that were actually core to the early products being so successful. Just so we understand, I think viral loops is kind of thrown around as t- two words a lot. What is a viral loop, and what makes one work and be efficient?

   10. LLLuc Levesque3:57

       Yeah. I think of the- of viral loop as basically where you have users on your product that are doing some kind of behavior that's inviting other users to come and register on the product. That's at its simplest level. You can have... The best is when the core product itself has a viral loop. For example, a photo sharing product. Uh, using a photo sharing product and not sharing it is not, uh, a very useful product. So the- the loop, the viral loop is baked into the product itself. You can have kind of secondary loops baked into it, but ultimately, it's about users who are on your product inviting other users to, potential users, to come and use your product and convert. So there's different ways of doing this. It can be, uh, challenging and- and difficult to get right. But when you do get it right, it can lead to some pretty exponential growth.

   11. HSHarry Stebbings4:41

       What are the most prominent viral loops at Shopify?

   12. LLLuc Levesque4:46

       W- we're, it's something we're- we're constantly working on and iterating on. One of our- our best sources is through recommendations. Uh, our merchants love Shopify, and they mention Shopify to their friends, and other merchants come in and- and join through that loop. So we're- we're always iterating on making it easier for merchants to do that. But it, because we have such a great product, and one of the, uh, highest cited reasons for joining Shopify is that somebody had recommended them in, that it's, again, it's kind of part of the- the core product where, uh, it would be hard to do if you didn't have a great product. But, so for Shopify, it's- it's, we can start off that great baseline, and it's- it's an area we're constantly iterating on.

   13. HSHarry Stebbings5:23

       Now, you mentioned Facebook there. I- I do want, have to unpack this incredible journey that you had with Facebook. When you look back on that time, I- I love that playing the Oculus Rift, uh, story. (laughs) But when you look back on that time, what were some of the biggest takeaways for you from- from spending the time you did at Facebook in such a pivotal moment for the company?

   14. LLLuc Levesque5:42

       Yeah, it was an interesting time for the company. I was, uh, hired to, uh, grow the product in- in different ways. Some of the biggest lessons that I took away from that were- were things that I had already done, but is something that I continue to keep inside the playbook. It's that not to look at growth or even product building as one specific thing. To grow your product, you can look at it in terms of SCO or paid loops or viral, but it's also about...... M&A, and building entirely new products. So for example, I had a variety of teams at, at Facebook, but the one that was the biggest win for us was building Messenger Kids and engaging an entirely new segment of the population and by building this kind of, uh, safe family messaging product. So it's, it's, it's about thinking growth, thinking about growth broadly, and really thinking about all the different ways of growing.

   15. HSHarry Stebbings6:35

       Uh, uh, totally with you there. Can I ask, when you think about also the advising that you said there, you mentioned some of the companies you've worked with, from your Patreons to your Quoras to your Twitters. When you think about commonalities between what made them so good, are there commonalities that stand out to you when you reflect back on those experiences advising them?

   16. LLLuc Levesque6:57

       Yeah, I mean, there's, there's a few things that come to mind. The probably main commonality would be around the... we're advising a company where there's a, a lot of success with the product and where there's a good engagement, is that when the growth teams themselves are very curious and very humble and open to being wrong, which is not the case, uh, for all growth teams. But certainly when you have an advisor, uh, relationship with the company is very important. And, uh, some companies have done extremely well with this, and, and advising is a... actually, growth is an interesting thing where it may take 15 years to learn something and, you know, 10 different experiments, but you can communicate it very quickly. So it's, it's a, it's a unique relationship with a company, and the ones that I've seen work the best, the commonality would be, uh, teams that are just constantly questioning assumptions, open to being wrong, very curious. And that's, that's something that I've seen work really well.

   17. HSHarry Stebbings7:48

       I think the humility that comes with that understanding is also just in, in a mutual understanding of what growth actually is. And it means a lot of things to a lot of different people, I think. (laughs) So if you were to kind of put on like a, you know, like a fridge magnet, what is growth, what- how would you define growth today, just so people have a shared understanding?

   18. LLLuc Levesque8:09

       Yeah. It kind of goes back to my comment, um, from my takeaway from Facebook. I, I start with the outcome. What is it that you're trying to do? What is the number one metric or, or activity you're trying to, to create as an outcome? And then what is... then it's all the activities and the all the different approaches that can lead to that, that growth. So I don't define growth very crisply in terms of it's this one thing. It's basically whatever it takes to move that one, um, metric you're trying to, to increase. So I know it's a, it's kind of a broad answer, but it's, it's a reality and it's, um, it's how I've approached growth in general.

   19. HSHarry Stebbings8:48

       No, listen, I love that, but I'm gonna pretend like you're my advisor for a hypothetical startup in this conversation. I get you.

   20. LLLuc Levesque8:54

       (laughs)

   21. HSHarry Stebbings8:54

       So it's about moving that one metric. Luke, how do I know what's the right metric to choose? This seems like an awful lot of pressure. How do I choose the right metric and how do I know that I've got the right North Star?

   22. LLLuc Levesque9:06

       Yeah. I mean, you ultimately wanna find the one that has the highest impact for the company. That'll be different in each case. So for Facebook, it was, um, daily active people engaged on the product. At Shopify, we care a lot about getting as many merchants to be as successful as, as possible on the product. So it comes down to how does the product work, what's the most important lever? And then making a very opinionated decision over this is the one thing that matters. And this one thing, if you get right, the flywheel that it kicks off, will lead to a lot of growth in the right ways for the company.

   23. HSHarry Stebbings9:36

       So how do we think about attacking that in terms of the growth teams? 'Cause the other big question I find is like, do you integrate growth teams into product teams, into marketing teams? Or do you have standalone growth teams like you had at Facebook? How do you think about advising founders when it comes to standalone versus integrated?

   24. LLLuc Levesque9:54

       Personally, I find that it works best when it's standalone. It really comes down to who you're able to hire to bring on the team and the, the skill sets that that person... because that's the constraint. The, the constraint is gonna be the, the who can you find to lead the team who's epic and what are those strengths and how can you build around that person? That's, that's generally how I start. But I do find it works better when the growth org is standalone and that you put as many functions as possible within that, that growth org. The, um, the bottom line is they're gonna have to work either way with the product team, and, uh, it's just a really important relationship between the product team and the growth org. But growth has its own culture that it needs to, uh, it needs to build and it, it does operate a little differently than a traditional, uh, product org. So I, I prefer to have it separately. It can work in many different ways. There is really no cookie cutter, this is how growth works, this is how product works. It depends on a lot of different variables, but I've seen it work best when it's separate.

   25. HSHarry Stebbings10:50

       Okay. So it's separate. Um, I know that I actually really wanna focus on this and have it as a priority for me as the founder. Do I hire a junior growth engineer and we start small and AB testing, or do I go for a big stellar growth leader and build a team around them? How do you advise me there?

   26. LLLuc Levesque11:08

       Well, I generally like to hire as, as senior as possible. Um, you wanna bring in somebody who's had a ton of experience. Growth takes a really long time to build intuition around. So you need to have done a lot of experiments, tried a lot of different things, and growth is a lot about pattern matching. So what, what have you seen work? What are you seeing in the data? And then how can you use your intuition to build other, um, other tactics that'll move the needle? So I, I try to go as senior as possible. That's not always possible, because one of the problems with growth is that there's, there's, there's a limited pool of talent in the growth, um, sector. Uh, so that, that is probably the constraint that a founder will run into first. So one other approach if, if the ability to hire somebody very senior is not possible, then you can hire somebody who's a bit more junior. But in that case, I would surround them with, uh, a set of advisors.

   27. HSHarry Stebbings11:59

       Agreed in terms of surrounding them with a set of advisors. (laughs) Now let's think about the hiring process now, because we know that we wanna get these people on board, but bluntly, I've never hired a growth professional before, so I have no idea how to do it, Luke. How would you literally structure the process for hiring this as h- senior as possible growth person to join the team? What does that process look like for you?

   28. LLLuc Levesque12:22

       Uh, this is an area I spend a tremendous amount of time...... focusing on. In fact, you know, as a growth leader, w- what levers to pull, knowledge, is all important, but if you're building a team and you're not a single person doing growth, the impact of your team will be the most important thing. So hiring and leadership become the most important, um, skills in that case. So it's a craft, in and of itself, that I certainly spend a lot of time refining. Uh, so it's- the process itself can, can vary, but the way I do it is I try to look for, uh, what I call signs of excellence. So past performance is the best predictor of future performance, so what I look for is what signals indicate past behavior and success that would be predictive of a few- future success. That's like the rough scaffolding. Um, so ideally, you want signals that tell you that they're, they're the best. So one signal, for example, uh, I remember when I- a long time ago when I started building teams, really sat back and looked at, "Well, what, what are the signals that somebody who's amazing would, would be highly correlated?" And one that, that I use is that when I meet with somebody, I'll start at the bottom of their LinkedIn page and kind of get them to walk me to the top, and I just listen and, and ask questions along the way. And one thing I'm looking for is, you know, motivations for leaving a role, what w- what did they accomplish at each role. But one signal that doesn't come up all the time but is, is just a great signal is when somebody's leader left a company and went to another company and came back to get them. And if you think about the social dynamic that, that takes place there, the only person that, or one of the only per- uh, people that really knows how good a growth leader is, is their, their leader. And they have left the company and have taken the social risk to come back and, and take that leader with them. So to me, that's such a great sign of, um, that this person is, is high talent, so that's one signal I look for. Uh, I love ex-founders. I love people who have, uh, shown a, a certain degree of grit and had, uh, repeated success, so it's not that they've been successful once or lucky. They've had a, a pattern of repeated succ- success. That's one thing I look for. But my slam dunk is very trusted back channel, uh, somebody who's recommending this, this person that I'm evaluating, uh, without any caveats. I do find when a recommendation comes with, "You know, they're great, but ..." Uh, you have to be careful because sometimes people are trying to tell you something without, um, maybe telling you the, the whole, the whole thing. So great people usually come without caveats. Three signs of excellence and a sense of, "Would I enjoy working with this person?" Because it's such a critical role. But those are my three, my three slam dunk,

2. 15:00 – 30:00

   Well, what are those…

   1. LLLuc Levesque15:00

      uh, things that I look for, and I have a s- kind of series of questions that I also ask to... You're basically trying to get a set of signals that tell you, "This person's really, really good," and there's a few ways to do that.

   2. HSHarry Stebbings15:11

      Well, what are those questions? 'Cause my, my trouble here is like often people work in Facebook's growth team, and clearly, it's a fantastic growth team, and it achieved a lot. But I don't know the difference in attribution between individual and team. What questions do you deliver to really tease out individual performance and quality of candidate?

   3. LLLuc Levesque15:32

      So for me, I try to get a sense of how deeply they understand the craft. Um, were they just executing a playbook they took off the shelf? Um, if they're- if this person's from Facebook, they'll have a very specific set of knowledge because they're supported with a lot of scaffolding at Facebook. Um, so you really wanna get a sense of how, how deep did you go and how well do you understand the craft? So one, one question I love to ask is I'll ask, "Tell me something I don't know about growth." And it actually doesn't matter if they tell me something that I, that I already know. What it does though is it forces the person to dig into their, their knowledge and their experience and pull out the one thing they believe is the top of the stack in terms of, "This is something really clever or really, you know, I know this, I experimented this, I learned this." And it lets you kind of anchor their knowledge in terms of they come up with something really good, and even if you know it, it gives you a sense of, "Oh, wow, they, they actually know that? That's, that's really good." Um, so that's my favorite question. So you're really trying to get a sense of do they really understand why certain things work. And you can ask other questions like, um... I do like asking a question then just following up with a, a long (laughs) string of why's to see if they truly understand what's happening, uh, and that'll give you a sense of, you know, how quickly does that fall apart. If you ask why and, you know, you know, we can't answer, like can you go extremely deep down the knowledge stack to see if they, they understood it? And to me, that differentiates a good growth person from a truly exceptional growth person, somebody who understands all the levers. Because that deep understanding is what will let you use your intuition to launch new experiments and drive new growth without having to, to run another set of experiments. You can use that intuition to drive an experiment that you know 80% will, will work because you've- you truly understand the mechanics of the levers.

   4. HSHarry Stebbings17:20

      Totally with you in terms of really digging deep on the why. In terms of the actual process, again, I- I have no idea. So I find this candidate. I have the first interview. They seem good. What next? Do I, do I do like a test with them? Do I i- integrate them into the team? Like literally, what does that actual meeting-by-meeting process look like?

   5. LLLuc Levesque17:43

      Well, the first meeting I, uh, like I mentioned, I always start at the bottom of LinkedIn page, walk into the top, look for signs of excellence, ask some tough questions, try to get a sense from that first meeting. Do they truly understand growth? What have they done that's exceptional? Usually, within 20 minutes of that meeting, I- I'll know am I, am I... how aggressively am I pitching this and trying to get, you know, as much signal as possible. If the call ends and I don't have enough signals to suggest that they're really good, the call might just- that might just end right there. That might be the last call. Uh, a good meeting where I get a sense of like, "This person actually has some, some pretty good signs," the next step would be I would go to my network and back channel. Uh, back channels are, are, are pretty important s- to me. Uh, at this point, you should really know. If you get, um, a couple back channels and you're getting a strong signal and you see signs of excellence and you've got a, a good sense of what they've done and that they actually understand the levers, that's, uh, that's pretty good at this point.... you also should get a bit of a sense of whether you'd like working with them. If you still don't know, you can go off to ask them to do some, uh, to do a test. Uh, I have done this, not always, because sometimes candidates won't want to do it. Um, and it really depends, you know, how well have you sold the role? What is their current situation? How much do they want the role? Uh, so I'd say in a smaller percentage of times, I get somebody to do a test. But also, growth people are in, uh, very high demand, so, you know, many times candidates, if they're not really interested in the role, um, may, may just choose not to do the test. So, so the test isn't a, a thing I always do. I have done it. It's usually if I've got some, you know, "I'm just not sure about this candidate I want," just like, "I don't want to have the back channel," or, you know, the- the person's been in one job, they've been at Facebook the whole time. Can they actually survive outside of Facebook? I'll want a little bit more color, but, um, that's generally, uh, how I do it.

   6. HSHarry Stebbings19:37

      On reflection, where do you think most founders go wrong when assembling their growth teams when they're hiring?

   7. LLLuc Levesque19:44

      I mean, the biggest mistake you can make as a founder is hiring the wrong person for the role. And I guess it's related to what we were just talking about. It is very hard to get a sense of if the person is any good. Uh, it can be obviously very costly, both in terms of time and money to make the- the- the wrong mistake there, uh, to hire the wrong candidate. Um, so I think that's- that's the biggest mistake. Now, it's- it's very hard to sniff out top talent, so that makes it exceptionally hard. I'd say some common mistakes would be, and this is something that's unique to growth, is that in growth, some of the signals that may suggest that somebody is amazing in a different discipline might not apply to growth. I'll give you an example. In most professions, if somebody's at a company and they're promoted very quickly, that would be a good sign that this person is, uh, you know, exceptional. And in growth, because there's so few people in the space, it could suggest that they're exceptional, but it could also suggest that maybe they were the only person available, the only person this company could hire, um, there was a lot of historical knowledge. There's different reasons because there are so few growth people in the market, uh, that, um, there's a few other... you know, there's other reasons with growth why somebody could be promoted, which, so it can be a bit of a counter-signal as well. Um, another mistake I've made is, uh, well, I've seen founders make, um, is sometimes the ex- the social media halo, uh, or public speaking halo does not correlate with ability. So I've seen people make hires for people that are, you know, on the stage often and great speakers and great evangelists that doesn't always translate to somebody who's- who's great, uh, as a growth person. So I think when it comes to that, you really want to not... again, not just look at what somebody's, um, doing on social media or s- public speaking, you want to look at what they've done, what teams have they led, what impact have they had, uh, and do a lot of back channeling. That- that's another mistake I've seen m- made before.

   8. HSHarry Stebbings21:48

      I- I- I totally agree with you in terms of the loudest microphone doesn't mean the highest quality of candidate. Can I ask you though, should I bring in the head of product into some of these interviews? Should I bring in the head of marketing? They're obviously going to play a crucial role in working alongside them. Should I integrate the team into this process or- or actually do it myself?

   9. LLLuc Levesque22:05

      We definitely want to integrate it with the, uh, the- the product and marketing teams, especially because the relationship between product and growth will be... and marketing, uh, will be critical, because growth and- and product are so intertwined. There'll be work that the growth team will want to do that will affect the product. There'll be work that may potentially, if it's done the wrong way, negatively impact each other's goals. So that relationship needs to be super tight. The relationship between the CEO and the head of growth and the head of product, especially the head of product and the head of growth, are two of the most important relationships.

   10. HSHarry Stebbings22:39

       Can I ask, as a founder, what can I do to make that relationship between my head of growth and my head of product, what can I do to make that as tight as possible?

   11. LLLuc Levesque22:51

       Uh, I think bringing them into the interview process and making them part of the selection, uh, team is- is a good start. Um, making sure that, you know, these- these two individuals have to, uh, be meeting regularly, have a good understanding of each other's goals, and- and work in a- a very kind of collaborative way. It's a bit of... it's a bit of an art, I don't think you can kind of tell both people to- to work well together, but having them work closely together, be aware of each other's goals. Um...

   12. HSHarry Stebbings23:19

       As a growth leader, should I push for growth teams to be included in all product reviews?

   13. LLLuc Levesque23:25

       You know, I've seen it work so many different ways, Harry. Um, I've seen it-

   14. HSHarry Stebbings23:29

       (laughs)

   15. LLLuc Levesque23:29

       ... you know, in- in some companies, yes. In other companies, no. It's worked, it's worked well, and- and-

   16. HSHarry Stebbings23:37

       Why- why did it work in some companies, uh, not in others?

   17. LLLuc Levesque23:41

       ... I think in some companies, it depends on the way the company is structured, how teams work together, how cross functional the- the- the products that are being built are, like how integrated into the- the rest of the product it is. Yeah, I've seen it- it work- work both ways. So in... if you can carve out surface areas, and I think this is- this is important. I think ideally, you want the growth team to own all of the surfaces that matter the most to growth. Uh, that's- that's the ideal state. Senior leader, multi-, uh, disciplined, they know how to do engineering, they know how to do growth, they know how to do product, and they have ideally all these functions, these sur- these surfaces, um, as part of the- the growth team. So they don't need to sit in- inside of another product team's reviews because they... for that specif- for any work happening in that surface because they own the surface. Where that's possible, where you can actually carve it out, that's very important and allows the teams to run independently very quickly. So there's less of a sitting on- on different teams' product reviews in that case. Where the surfaces are owned by... all owned by a product team, then yes, the growth team should be very embedded and reviewing all the work that's being done.

   18. HSHarry Stebbings24:48

       ... question for you, and sorry, I'm, I'm really asking the tough ones. (laughs) Uh, I'm just too interested. You know, growth teams often, like, go into code bases, tinker around, A/B test, and then fuck off. (laughs) Like, how do you ensure that growth teams have the ability to test and iterate without pissing off growth teams, uh, without pissing off engineering teams building sustainable, reliable, durable code?

   19. LLLuc Levesque25:15

       Uh, it... So there's a few ways. First, if the, the relevant surfaces are owned by the growth team, then that becomes less of an issue. Um, and then ultimately, it's through a collaboration with the product team. So, you know, at Shopify, we work very closely with the product team day in and day out, and make sure we have, um, we, you know, we have reviews across both teams where there's overlap, and we involve them as part of the process for the build so that we make sure that there's, there's trust being built. It's when there's no trust that the whole thing falls apart, uh, which is why it really comes down to keeping different teams involved, keeping them in the loop, building trust through, uh, making sure you keep, uh, different teams and different, uh, priorities in mind, and you're not just going in and lowering the quality bar. That's a common, I think, failure state, where you build something that moves the needle, but the quality's terrible, it looks awful, it's bad for the other team's goals, it's bad for consumers. That's a failure state for growth. So, so there's, there's a few ways to do this, but ultimately, it comes through with relationships and trust that are built through, um, working together and just showing this ongoing ability to build products in ways, and, and growth optimizations in ways that doesn't affect other teams negatively.

   20. HSHarry Stebbings26:28

       Luke, how do you structure growth reviews? How, who sets the agenda? Who's invited? How long is the meeting? How often are they? Help me understand what a growth review looks like and what a good one looks like.

   21. LLLuc Levesque26:40

       This is something we're always iterating on. We try at a variety of different... It depends on the size of the team as well. I mean, a growth review is, is, is not really needed when you're a, a, you know, a- maybe a two-person growth team. But ultimately, you wanna look at... And it also depends on what stage. If you're at an early stage, you wanna look at, what do... Just get buy-in on, "Hey, here's the, here's the kind of work we wanna do, uh, from a growth, uh, optimization perspective. Does that make sense?" Uh, we call that the proposal stage. And then when you... everybody's aligned, then we do a review around the, the metrics, the opportunity size, and make sure that, um, we all believe that, you know, this is building the best thing possible. And then... So there's, there's kind of a- an evolution of growth reviews. So it's, they're not all made the same way. Ultimately, they should be very driven around metrics and, um, opportunity size, and also a strong conviction around what we think will work. Uh, but it really depends on the, on the stage of the review.

   22. HSHarry Stebbings27:34

       Okay, so we have a project, and it hasn't gone that well. How do you, as a growth leader, prevent morale deterioration across the team when a project doesn't go to plan and it doesn't work?

   23. LLLuc Levesque27:48

       Well, one, one project not working doesn't generally tank morale, but it's when there's like a series of them. So, there's a few things that are, that are important, and I've seen this in a variety of, of different areas, especially with new products that are looking for that growth lever. Um, because you can go through a period of just iteration and, and learning. So there, there's a few ways to do this. I think one is you do need a culture of grit. You need to, uh, have strong leaders that can keep the team together through those, those phases. I think it's important to celebrate all of the wins, even the smallest of wins, and really focus on the learnings, even though it might take a while to unlock certain levers, and you have to have that grit to get through it, through a series of failures. But reframing, it's not so much a failure, but if something didn't work and you've learned from it, that is also success. And you have to talk about it that way, recognize it that way, and build a culture that, that celebrates that. So I think there's, um, there's some strong leadership required through those phases. Uh, but I, I do think it's important not to frame failed experiments as failure as long as you're learning and it leads to an unlock that's even bigger, and that, that happens quite often.

   24. HSHarry Stebbings28:55

       How long should one give new projects before determining whether it's a success or failure? You, you never want something to be too long and it drawn out, but you also don't wanna cut something off before you have enough data to know whether it's a success or failure. How do you think about enough data to determine conclusion?

   25. LLLuc Levesque29:13

       Talking about projects, now people?

   26. HSHarry Stebbings29:17

       Either or. Uh, I mean, let's do both.

   27. LLLuc Levesque29:20

       Okay.

   28. HSHarry Stebbings29:20

       (laughs)

   29. LLLuc Levesque29:20

       So, projects, uh, when you run out of... Like, six months is a good chunk of time. If you're not e- e- exiting six months with a sense of, okay, we have some levers, we see li- uh, we see line of sight into some levers that, that could move the needle for, uh, an individual product or, or project. Uh, if you can go six months without any hope or (laughs) inspiration that something's working, that's, uh, that's probably a sign. Um, you know, it can be longer, but generally, six months is probably a good, a good milestone even if it's not about unlocking the growth, but just having learnings that are, uh, suggesting that there, there are,

3. 30:00 – 45:00

   What are some big…

   1. LLLuc Levesque30:00

      you know, better wins to come and giving you hope. In terms of people, I think within three to six months, if you have a new leader coming in, they should be able to demonstrate some- something (laughs) that gives you a, a sense of whether you made a good hire or not. We like to say that, you know, the best people come in and leave a- an impact crater when they leave. So they, they get hired, so everything around them gets better. So within, frankly, 30 to 60 days, you should have a sense of whether you made a good hire, and they've either hired in their, their best people. The best people become kind of talent magnets. They pull in the best people with them. Within, you know, three months, they should have an early win. Um, the speed of experimentation should, should increase. So within the first few months, you'll get a sense of whether you've made a good hire or not, and it should be pretty obvious.

   2. HSHarry Stebbings30:48

      What are some big red flags in those first 60 to 90 days? If that's, like, the things that they could do well, for me as a founder hiring my first growth leaders, what should I be nervous about or look out for?

   3. LLLuc Levesque31:01

      There are many. (laughs) The, uh, the two that, uh, come to mind would be, first is w- speed, I think, is very important in growth. The cadence and velocity that you can experiment in and, uh, and, you know, push new code out and, and run things is very highly correlated with your ability to be successful. So if I'm not seeing the speed and progress within the first few months, I would be... or hearing a lot of excuses, I would be pretty nervous. That would be a red flag for me. And you've kind of hinted at this a few times earlier around product teams working with growth teams and how do you structure them and have them working together. Because of how cross-functional growth can be, if there's tension in the first few months between the t- either the teams or the new growth leader and the product teams, that's, that's a, a red flag and, and that, that, that can go sideways pretty quick. So, I'd look for early tension. How are the relationships working out? Um, is this new person coming in with a sense of humi- humility, asking a lot of questions? And, um, that can go sideways pretty quick. I would... You know, speed and velocity can be improved. Uh, a relationship falling apart can be completely dysfunctional and just cause the whole thing to fall apart.

   4. HSHarry Stebbings32:12

      What's been your biggest challenge in scaling the growth org at Shopify, Luke?

   5. LLLuc Levesque32:17

      I mean, it always comes down to, to the team, to building an amazing team. I wouldn't say it's, it's, uh, a- it's a challenge we're having. We, we're, we've built an amazing team. I would say Shopify currently has, has some, some of the best talent in the world, uh, doing growth. Uh, but it's always something I spend a tremendous amount of time, uh, focusing on. And it's one of those spaces... Growth is one of those things where if you make the right hire, it's not a, you know, 0.5% increase. It's literally a, you know, a, a 10 or 100, uh, X increase. So, that's the area, I don't know if I'd call it a challenge, but it's the area I focus a lot of time on and also the area that l- literally making the right decision can be just a game changer for, for growth in, in all aspects. So, if you look at whether it's in paid search or in, um, funnel optimization or in SEO or in product building or we have a, uh, a product org and an engineering org inside the Shopify growth, uh, org. And, um, just bringing amazing talent is the thing that matters the most, so that's what I spend, uh, the bulk of my time on, uh, when I'm not in reviews and, and, uh, getting my hands dirty with the teams. Uh, so that's, that's the most important thing. Uh, I don't know if you classify it as a challenge, but it's, it's definitely the, the area that has the most impact at, at, at, at my, my scale, at the scale that we're operating at.

   6. HSHarry Stebbings33:33

      Where do you think you'll wake as a growth leader, Luke?

   7. LLLuc Levesque33:36

      Hmm. Well, I think it's always a challenge staying high and then going in deep. I do love getting my hands dirty. Uh, I really, really enjoy that. Uh, and, you know, as the team grows, that s- becomes harder to do. And so it's that, that trade-off of how do you stay at the very high level and then zoom in to all the different areas. That's, that's a skill I'm always working on, um, that I think I'm okay at, but I'm not, uh, it's something I'm always, uh, looking to improve and, and get better at.

   8. HSHarry Stebbings34:09

      Yeah. I, I, I totally agree. I, I think with founders actually, for me, the, the best founders have the ability to move from the 5,000-foot master vision to, like, last Wednesday's Facebook ad spend and how it converted. So I totally agree with you there.

   9. LLLuc Levesque34:22

      Yeah. Yeah. (laughs)

   10. HSHarry Stebbings34:22

       Can I ask, when you think about... And... No. Sorry. You go.

   11. LLLuc Levesque34:26

       Well, a subtle tweak to that as well is that, you know, how do you do that without completely deflating the team? Uh, because there's a way to go in and engage with the team and have, you know, get right on the, on the beach and, and with, with, with the troops and do a lot of work to make things better. But, uh, done in the wrong way, it can be very disempowering for the team. So there's, there's an art to how you stay high, empower teams, hire the best people, uh, be involved, but also give them space to execute. So it's, it's an art. It's definitely an art.

   12. HSHarry Stebbings34:54

       You said that about kind of getting your hands dirty in the data, in the trenches there. Can you talk to me about a decision that you made without data and how did it go?

   13. LLLuc Levesque35:04

       So I think data is very important and especially when you pair it with, like, a pattern matching off of previous experience. So I know, for example, uh, so I focus a lot on, on what is the thesis that this experiment is likely going to work? What are the signals that you have that suggests that this experiment's gonna work? Just throwing things on the wall. So you start with the data. What is the opportunity size? Is this worth doing? And then you try to look for signals in a, in a thesis that will increase the likelihood of something working. So I remember one situation at a past company where, uh, somebody on the team wanted to make a change to the website to see if it would impact SEO traffic and I asked them why. And they said, "I just wanna throw things against the wall and see if it works." That is a bad answer. A better answer is, "I did this experiment. It had this result, which would suggest that if we do these two things, because we also saw competitor increase in the search rankings based on a similar change, that we'll likely see, um, a, a positive impact from, from this experiment." So, so lack of data is not good. You certainly wanna start with a good data foundation, but equally important is what is the sound logic behind the thesis that will make sure that this works? And that's something that, um, I think probably doesn't get spent enough focus on in terms of validating not just the data of the experiment, but the thesis behind it and what are the signals... And building up that stack of signals that we know. Those are, those are so valuable in growth. When you build a set of, we call them levers essentially, a set of levers that you know work or experiments that, that didn't work, you build this intuition over time and, uh, I think that's frankly the most important part of growth because it increases your hit rate as you're running experiments.

   14. HSHarry Stebbings36:53

       It's so interesting you say that, because Bangali Kava, uh, formerly of Instagram, who I, I know well, and then obviously Casey Winters, uh, formerly of Pinterest, who I'm sure you know, um, uh, they both actually said that, like, past intuition has been detrimental to their (laughs) performance, because they've relied on previous experiences, expecting them to create the same outcomes that they did before. And actually, it impacted their mindset in the wrong way moving forwards, 'cause it was different products, different circumstances. How do you think about not letting past experiences negatively impact how you think about future growth projects? Because-

   15. LLLuc Levesque37:32

       Yeah.

   16. HSHarry Stebbings37:33

       ... it could be totally different.

   17. LLLuc Levesque37:34

       Yeah. I think there's a few th- few things you have to do. One, you have to constantly test your, your knowledge, 'cause one thing you know now may be different in a different context. And also, I think it's important not just to under- not... It goes back to really understanding what's going on. So if an experiment worked, that's... Just knowing that experiment worked is interesting, but may not be repeatable. The most important thing is understanding the why behind it. Why did it actually work? It goes back to some of my interview questions. Uh, it'd be... Uh, I don't know if easy is the right word, but it would be, say, like, easy for somebody to give me an answer that... uh, uh, of something they've done that's worked well. But it's a little bit harder to dig th- you know, five whys deep to find out, "Well, why did, why did this work, and why do you think..." And once you know the root knowledge behind it, that's more transportable than the actual, um, experiment, uh, result itself. So, uh, but I think it's, it's valid that you need to constantly retest these assumptions and retest these, uh, experiments. But, um, counter to that, once you understand the, the, the deep knowledge of why something works a certain way or why a certain algorithm works a- another way, I mean, I've certainly, um, seen cases where you can apply that very quickly, uh, to, to, to produce wins. And, uh, it's usually a really good starting point. You come in with your playbook, you execute the stuff that you know will work, and then the things that don't work, you just reassess your base knowledge. That's constantly evolving, but you should have this set of knowledge, the, the deep, true understanding of what's going on, um, not so much that when you make... you know, you change X to Y, that creates a lift. More importantly is, why does it create a lift, and how d- and how can you continually do that going forward?

   18. HSHarry Stebbings39:13

       Can you talk to me about a specific growth decision that didn't go to plan? And why didn't it go to plan?

   19. LLLuc Levesque39:20

       A growth decision? Can you define growth decision?

   20. HSHarry Stebbings39:23

       Uh, so it could be we decided to focus on, uh, button color changes, or we decided to focus on a specific, uh, referral program for LATAM, and actually, we got it wrong. Um, it could be we decided to implement a Twitter recommendation for usernames, um, where we suggested your username and we thought that would reduce friction and barriers to entry, and actually, it didn't. It just created usernames that people hated. Um, uh-

   21. LLLuc Levesque39:56

       Yeah. I'm sure there's tons, uh, like (laughs) because it's usually... especially at the beginning when you start a new lever, there's a, a huge failure rate. So, when we were working on TripWow, uh, this viral photo sharing product, one of the things we learned was that... Be- it was a viral loop, uh, but also it was, um... it was, uh... it was a consumer product. And like most products, whether it's SaaS or consumer, the funnel really matters. And we spent a lot of time reducing friction, which is a, a common growth approach. You want to reduce friction, and you want to increase motivation. In fact, that's one good mental model that I, I like to think about products through, is that, how do you reduce friction and use the product, and what is the intent of the user coming in? What is their motivation to use it? And those are two levers you can play with. What I, what I found was that we'd run a series of experiments to reduce friction from, from the sign-up page, from the home page. And that had... ori- originally, at the beginning, had a lot of impact, but you basically bottom out. There's only so much friction you can reduce until it just stops having impact. So, uh, uh, and it... to answer your question specifically, you... I think we did seven iterations of the home page, and after, you know, three, we saw a lot of impact, but then there was very much diminishing returns to whether it's button colors, or changing the UI, or adding "Free" to the button. Um, it just... you get to a point where you've just removed enough friction proportional to the intent of the users coming through, and removing more friction is not going to move the needle. So that's, that's actually a good... Going back to lessons learned from historical testing, that is one lesson that I th- think is important to take forward, which is th- there's, there's a s- there's a bottom to how much friction you can, you can remove from the flow, and it... you know, you can spend a lot of time doing experimentation there that doesn't move the needle at all.

   22. HSHarry Stebbings41:34

       I, I, I totally agree. I actually had Scott Belsky on the show say, "The devil's in the default," when it comes to product and growth, being just accepting what is is where the devil lies. And I, I don't know, I ruminate on that, because I'm like, "Well, users like what they know, and they like consistency." How do you think about that devil in the default in terms of kind of product design and growth understanding?

   23. LLLuc Levesque41:58

       I interpret that, I think, in, in a few different ways, so I'm not sure what he was getting at. But I mean, certainly defaults for a product can have a, a massive, uh, impact on, on growth. Uh, depends on what you're trying to achieve. But, um, but def- removing friction and defaulting user behavior through, through interaction with a product will have a, a, a big impact to how the product, um, uh, grows. I'm not sure-

   24. HSHarry Stebbings42:23

       Yeah.

   25. LLLuc Levesque42:23

       ... if that is exactly what he's talking about, but... (laughs)

   26. HSHarry Stebbings42:26

       No, no, I, I get you totally. I do want to dive into my favorite though, Luke, which is a quickfire round. So I say a short statement, and then you give me your immediate thoughts. Does that sound okay?

   27. LLLuc Levesque42:35

       Sounds good.

   28. HSHarry Stebbings42:36

       Okay. So what tactics have not changed over the last five years in growth?

   29. LLLuc Levesque42:41

       Um, there's, there's a lot that haven't changed. They've, they've, they've changed a little bit, but they're still, they're still valuable. So one is... certainly SEO is still... it's changed a lot, but it's still a very, uh, big fire hose of traffic and conversion that you can get. So SEO is, is a, is a, is a good one. I think like the tried and true funnel optimization work or just product optimization of...... um, making a product easier to use, getting users to a delightful moment in the product as fast as possible. Those principles haven't changed and that's something that, you know, even though some of the channels have changed, doing work to, uh, optimize onboarding or exit flows, that- that- that type of work ha- is still very, uh, very valuable. Um, and then, you know, paid channels in general have become more important because a lot of these services, a lot of these engines like Google, are moving away from showing organic results and showing paid. So paid, certainly, paid search for example, works very well and, um, uh, you know, it's still a very large channel that has to be part of anybody's strategy.

   30. HSHarry Stebbings43:45

       Tell me, what tactics have died a death over the last five years?
4. 45:00 – 53:01

   Tell me, what- what…

   1. LLLuc Levesque45:00

      uh, I- I like using growth tactics that rely on engineering approaches rather than using a, uh, you know, a bunch of spreadsheets to come up with, uh, changes to make and- and doing analysis, but building algorithms and- and systems to scale, uh, growth solutions and- and you really need engineers to do that. So I think it's really important to give growth leaders the engineering, uh, capacity to- to execute, uh, quickly. Um, I'll select to say you need your- your growth leader to be as close to the sun as possible. You don't want to bury it too deep in the org. That's a mistake I've- I've seen done, uh, before. Um, because if there's something really critical to the business that is, uh, uh, a decision, perhaps a controversial decision that needs to be made, sometimes a very time-sensitive decision that needs to be made, you don't want it to be six layers from the CEO. You wanna have a- a short loop there. Um, so I think those are some mistakes that some founders can do. I've seen that.

   2. HSHarry Stebbings45:53

      Tell me, what- what one piece of advice would you give a growth leader today entering a new role?

   3. LLLuc Levesque45:59

      I'm really big on focus. I think focus is the number one lever that you can use at any point and have immediate impact by stopping the work on things that don't matter and starting to work on- on- on different work or work that- that does matter, that will move the needle. So I would advise a growth leader very quickly find out what your North Star is, find out what is that one thing that will be impactful for the business, get everything on your team rallied around that North Star, and then stop doing all the things that don't matter. That is the m- to me, that is the most important thing because it will increase the likelihood of- of having the right kind of impact very quickly. And it's- growth is something that can touch and move every lever and every metric in the company. So there's always this demand for, hey, can it grow this part of the business or this part of the business? And deciding which area that will be focused on and that the teams will- will focus their energy around, uh, is extremely important, harder than it sounds, and I think, uh, you know, a- an important part of leading growth is probably saying no, you know, once a week or at least once a month to something that doesn't align with that North Star. So it's very quickly determine the North Star, rally the team around it, and start showing wins as fast as possible.

   4. HSHarry Stebbings47:10

      Luke, I am your CEO. You are my head of growth. What does a good relationship between us look like? How often should we meet? What can I do to be the best CEO for you?

   5. LLLuc Levesque47:22

      So I've seen this work in a variety of different ways. There's different types of CEO with different DNA. Um, I- I think you want the- the distance between the CEO and the head of growth to be quite short. Uh, if I were designing from first principles, I would say the head of growth reports to the CEO meets weekly, and is flagging issues and also discussing potential, you know, strategy changes and approaches at a very high level. That said, I've also seen it work where head of growth reports to the COO or the president or, um, or another executive of, um, CRO potentially. So it can work in- in different ways but in either of those scenarios with the- the CEO and the head of growth need a very tight relationship and ideally be meeting weekly if possible.

   6. HSHarry Stebbings48:11

      Luke, what would you most like to change about the world of growth today?

   7. LLLuc Levesque48:16

      I think one of the biggest challenges with growth is there's not ... There's a small talent pool in growth. There's not a ton of people doing it and there's a variety of reasons for this. Uh, the schools aren't really teaching it, uh, at scale yet. Um, also to learn growth, the most valuable thing is that, and this is- this is probably a good signal for- for founders to look for, the best way to learn growth is to have a lot of- of reps, to- to have taken a lot of shots on goal 'cause that's the best thing you can do to inform, um, your playbook, to inform your knowledge, to inform your intuition. And there's not a ton of companies out there. There- there aren't an infinite amount of companies that have that scale where people that are practicing growth can learn that high-end growth and really become experts. So what you end up with is, uh, a small number of people who have that experience.... and who know what works and what, who actually know what works and what doesn't work because there's just, um, only so many companies where they can, they can build that experience at. So what I would like to change is to grow the talent pool and have, um, uh, just more people in the industry who are, who are good and who are sharing the knowledge.

   8. HSHarry Stebbings49:28

      I, uh, I totally agree with you in terms of expanding the talent pool. (smacks lips) I want to finish with actually one of my favorites, which is like, what, what one company growth strategy, other than Shopify, (laughs) uh, have you been most impressed by recently and specifically why?

   9. LLLuc Levesque49:44

      The, uh, one company that I'm an investor in, uh, it's called Bounce, and was really impressed with their growth strategy. One of the reasons I invested was because they've been very good at their growth loop, and it's everyone in the company from the founder, uh, to, to everyone else very much understands it. And, you know, I've been an, an advisor for, for a long time. The one thing that, i- it goes back to a lot of things we've talked about today, that I was very impressed with this company is their, the way they leverage advisors and have surrounded th- themselves with people that can help them learn and how they've been, uh, just incorporating that knowledge and iterating on their, their growth loops through, uh, through what they're, they're hearing. So I think that's, that's something I've been very impressed with with this company. It's growing really fast and it has to do, I think, with how they've really put that growth lever and their growth channels throughout the bloodstream of the entire company, everybody in the, in the company, and how they've leveraged, uh, advisors. I really, I actually do think more people, more founders should leverage advisors as probably the fastest way to get, uh, wins and get knowledge, um, especially in this limited talent pool that, that's out there.

   10. HSHarry Stebbings50:57

       Sorry, final, final one. What can a founder do to attract the best advisors? What can they, how do they pitch? How do they br- What makes you say yes versus no?

   11. LLLuc Levesque51:10

       That's a good question. Um...

   12. HSHarry Stebbings51:13

       (laughs)

   13. LLLuc Levesque51:14

       I think, I think different people are looking for different things. Um, what I look for is, because, it, it depends on the f- on the advisor, I suppose. Um, so but most advisors and, and ideally you want to hire an advisor who you're, you're aligning incentives, so you bring in an advisor and you compensate them with equity. That's the idea. I know not all advisors would do that, but that's, that's the idea because then the, then the advisor is, is incentivized in the same way as, as the founder. Um, so if you're able to find, uh, advisors that would take on, uh, exclusively equity or the majority in equity, then really you want to convince them of the, the, in the same way you would convince an investor. That's probably the right mental way to, mental mindset to take here, uh, is to think about how would you pitch the company to an investor? Because an advisor who's coming in and taking equity is basically investing their time. That's how I think about it when I join a company is, what's the path forward? What are the growth loops? Do I believe that these growth loops will lead to the scale that we're trying to build? Um, what, what are the, you know, valuation multiples? Where can this go? What are the comparables? Uh, one thing that I've always found helpful is for a founder to introduce me to their VCs because the VCs invested their time and I'm investing my time, uh, sorry, their, their time and money, and I'm investing my, uh, my time. And the logic that a VC and investor would have taken is very similar to the logic that I'm using, so, so I'd say a founder, introduce them to your, your, um, your top, uh, VCs, uh, and basically pitch to the advisor the way you would pitch an executive coming in and to an investor who's putting their, their money in.

   14. HSHarry Stebbings52:48

       Luke, I cannot thank you enough for putting up with my, uh, quite prying questions today. I so appreciate it. I love doing this, and thank you so much for joining me.

   15. LLLuc Levesque52:57

       Thanks, Harry. It's a lot of fun.

Episode duration: 53:01