+00:00 GMT

Building a Successfully "Spiky" Org (Part 1)

with Jean-Denis Greze

July 16, 2020


Jean-Denis Greze - Head of Engineering @ Plaid

Jean-Denis Greze is Head of Engineering at Plaid, the technology company giving developers access to the financial system and the tools to build many of the most influential applications and services of the modern financial era. Companies such as Venmo + Paypal, Coinbase, Robinhood, Acorns, Clarity Money and hundreds more are built on Plaid.

"You're asking me what makes us different. I think it's that we've been really deliberate about building what I would call a ‘spiky org’ as opposed to a very balanced organization. The reality is when you're in a fast-growing company, it's much easier to do a few things well than to try to do everything."

- Jean-Denis Greze

Prior to joining Plaid, Jean-Denis was Director of Engineering at Dropbox, where he led the growth, identity, notifications, Paper, and payments teams.

Prior to Dropbox, Jean-Denis worked in fintech in New York and has CS degrees from Columbia as well as a JD from Harvard Law School.


Nominations for the 2020 Inspiring Leadership Award are now open!

We created this award to recognize role models of engineering leadership, for the work they do every day to make a difference in their teams and organizations.

Share their story with us and submit a nomination HERE


  • What you should focus on when building an organization: Be a "spiky" org (2:31)
  • How to change and adapt your organization that preserves your strengths, mitigates weaknesses, and develops new capabilities: force yourself to adapt your “spikes” (7:01)
  • Recruiting, Growth and Performance Management as “spike” examples in organization building (and why it's NOT useful to be good at all three of them) (8:15)
  • The org design dilemma between "Hiring Well" vs. "Firing Fast" (12:01)
  • The org design dilemma between business impact vs. craft and quality (18:27)
  • How you know when you should change your strengths, values and build a new "spike" (24:48)
  • The dilemma of building an organization with bottom-up vs. top-down decision making (27:47)
  • How to develop new strengths, capabilities, or “spikes” in your engineering organization (32:20)
  • Jean-Denis's process to create space for questions, creativity, and problem-solving (36:33)


What you should focus on when building an organization: Be a "Spiky" org

Jerry Li: Hey Jean-Denis, welcome to our show we're really excited about this conversation. Been, looking forward to talking to you, to learn all the things you have to share about building an engineering organization at Plaid and how your lead Plaid through rapid innovation.

I've been in payments for a big part of my career, and I know how painful it is to work with banks and to integrate with them. And it can be slow but I'm really impressed by the, engineering culture. And also how fast you guys can move. So, can you tell us what makes Plaid so different?

Jean-Denis Greze: Yeah, I think I can, I can try first of all, Jerry, Patrick, thanks for having me on the show. I'm just really excited to be here and to have this conversation. so I think what makes us successful today is that Plaid and Plaid engineering or Plaid engineering product design support isn't... It's not really one company. It's multiple organizations that have very different strengths.

and I think, you know, you're, you asked me kind of what makes us different. I think it's that we've been really, deliberate about building what I would call a spiky org. Like, as an org that has a spike, as opposed to a very balanced organization. and I'll explain why that's good for the problems that we have. But also the reality is when you're in a fast growing company, it's much easier to do few things well than to try to do everything. Right?

And so a lot of my focus as a leader is to say, Like, what are the two, three characteristics? It could be capabilities. It could be culture. It could be, like technologies. It could be people, whatever. What do we need? The few things that if we, if we are really, really, really good at them, we're going to be successful longterm. So I call this being spiky or being a lopsided organization.

And the reason why it's great is because one it's much easier for people inside of your organization to reason if they only have a few things to reason about right. If they only have a few goals to think about, they don't always have to weigh like 17 dimensions and be very rational and pragmatic, right. Cause it's, it's difficult, for people to have lots of ideas in their heads and to figure out the right way forward.

Number two, it's a lot easier to build a spiky org, I'll talk about this in a bit, like about how to approach recruiting. and performance management and, growth opportunities for your team members. If you want to be world class at all, three of those things that takes like a big people ops team, it takes a lot of time and effort, and often you just don't need to be good at all three to be very successful. So it's easier to build.

And then finally, if you have a very spiky org, you have a lot of clarity about the kind of problems you can solve successfully, and the ones that you can't. And that's really important to know. You need to always know as an engineering organization, the things that you will fail at because you're not structured to tackle them.

And I think at Plaid, one of the reasons we're successful as a business is because we said no, 90 times before we said yes, once, right? There's there's all these other companies and FinTech infrastructure, many of which are like billion dollar companies. And there were points in time at Plaid where we could have said like, "Oh, maybe we should build an API for that. Maybe we should go after this market."

And our founders are great about this, but I think it's something that the entire organization has. We just said no, because it doesn't align with our strengths as an organization.

Jerry Li: Do you have examples for that?

Jean-Denis Greze: for example, there's, there's companies that build, it's like lending as a service, right? They basically provide you like white label loans that then as a consumer facing app, you can offer loans to your customers. For a long time, that seemed like something... like Plaid we asked like, "should we do this right? Or should we become a credit bureau?" Right.

These things I came up and maybe if we'd gone, those routes we'd be more successful, but I just don't think so. I think the fact that we were so focused on just how do we allow developers to connect their users bank accounts. to the apps that they want to use that's it! Let's just do that really, really, really well.

And to build an organization that was good at doing that, but it's not good at building consumer facing apps. We're just not good at that. Like it's, it's not in our DNA. it was, it was a good decision. but there was temptation all along the way.

Yeah. So I can't go, I it's hard for me to go like. "Company X, like we thought about doing that." but, I think I know the one is like, an API for opening bank accounts for creating credit cards.

Right? All these things were things that at one point in time, we were like, maybe we should do it." but we decided that we just didn't have a competitive advantage building those things at the time. And so we didn't.

So, the main point here, right? The, point, and I think it's, perhaps, hopefully controversial for listeners is what are you really strong at? And can you ignore everything else? And that's how you should think as a leader about how you build teams and how you build organizations.

How to change and adapt your organization that preserves your strengths, mitigates weaknesses, and develops new capabilities: force yourself to adapt your “spikes”

The reality of the world though, is that if you only pick very few things that you're good at life is going to be difficult when the world changes. So what I want to talk about here is I'm going to give you some examples of spikes. Spikes that we have at Plaid with that I've seen other companies have. And why you can win if you have those spikes, but I also want to talk about , when you're facing change, how you can adapt your organization in a way that doesn't destroy your initial strengths.

Cause I think , one of the things that happened as companies grow is they become like average, as, as opposed to being great at a few things, they kind of... they become okay at everything. Right? And that's a disaster. You don't want that to happen.

So as you grow, you want to keep your spikes. You want to keep the things that keep you unique, but you obviously need to tackle a broader range of problems and you need some strategies to do that successfully without sacrificing what made you successful in the first place.

So I'll talk about how you can isolate so that you can tackle things that align with your weaknesses, how you can find outlets for things that your organization is not really good at. And then finally I'll talk about how you can force yourself to adapt your spikes.

eventually you will have to change, right. The world changes around you. And so. You know, you, you may be the world's best sprinter today, but if everyone used to run marathons and suddenly you need to learn a new skill set, like how do you go and do that as an organization?

I don't know if you have any questions. Otherwise, I was going to go into some examples of spikes that are pretty extreme, for people to think about.

Jerry Li: Yeah.

Recruiting, Growth and Performance Management as “spike” examples in organization building (and why it's NOT useful to be good at all three of them)

Patrick Gallagher: Yeah, definitely. I think it would be really helpful to dive into some examples of what spikes look like within plaid or within other organizations.

Jean-Denis Greze: let's do it. Alright, so let's talk about first spike for me is around how you think about talent. there's three pillars to talent, generally in tech, the first one would say, how you attract talent, so recruiting.

The second one is how you grow the skills of your team so that they can tackle the challenges that you have. I'll try to call that growth.

And then the final one is performance management, right? How do you make sure people have accountability and that, you know, if people are not meeting the expectations that you have for them within an organization, you find a path forward that doesn't involve them, frankly, generally.

Right? So those are the three pillars, recruiting growth, performance management. I don't think it's useful to be good at all three of them. Now, most people try to be good at all three of them. Right. But that's a lot of work. And the thesis here is actually, you only really need to be good at one of the three.

And there are different ways to skin a cat. Like you could build a successful company picking any of the pillars, and I think you can be successful. It just means your culture will be different and you'll make different trade offs.

So I'll talk about hiring well, meaning focusing really on the recruiting side, because that's the most commonplace template in Silicon Valley, right. It's handed down to this current generation from, you know, 20 years ago, Google that had historically a very high hiring bar.

So hiring well means you do a lot of work upfront to make sure that you only hire people who are going to be really successful within your organization. So this is where you look at, like from a technical perspective, you look at like, what are the technical skills you would look at? How do they collaborate with others? You would look at whether their values and their principles align with the values and principle of your company. Right? You would look, And their work experience and try to decide whether they've tackled problems that have helped them develop the skills that you need.

So you've spent all this time up front, you have very rigorous sourcing process. You have a very rigorous interview process. You give very few offers, right? and then people join your company and they feel like they've joined a special club. Because it was just so hard to get the job in the first place. And, and obviously they're successful within your org cause you spent all this time upfront, like vetting them.

Jerry Li: Yup. And you've built a brand by the, you know, reflected by the quality of the people they bring in.

Jean-Denis Greze: Totally. Yeah. Everyone's like, "Oh, Stripe's got a really high bar. So now everyone wants to hire Stripe engineers..." right. And before that it was Dropbox and before that was Facebook and right.

So, so that's hiring well, that's one way to do it.

If you hire well, you actually don't need to performance manage. Because you've already done all the work upfront to make sure that people are really successful. Sure some people won't be successful within the org, but...

Say, you go and you're like, "Oh, I also want to be the best performance management company out there."

So then you're going to make people feel like, they're in a special club, but Ooh, they could get kicked out of the club... And so you have all these benefits from hiring well, in terms of like this warm culture of inclusivity and then you're adding this thing, that's like, "no, no, you could be performance managed out." Right. "We're going to really pay attention to results".

And why are you doing it? It doesn't make sense. You've spent all the work upfront, forget performance management. Just assume that yeah you'll have 5% of people in your org that aren't performing that well, that's just the cost of doing business. Spend your energy on something else. You've already spent it all on the recruiting front.

And I see this mistakes made all the time. Like my pet peeve is the founder who like finds one person on the team who's not doing a great job. And then they're like, "we're not good at performance management. Like we need to be really good at it." Right. And then they go to like their head of HR and they're like, "we need to be really good."

And they like over-correct based on just the. The statistical chance, like, yeah, obviously you're going to hire some people who don't work, but in doing that, they really, they changed the culture and they force everyone to do work that they don't need to do to be successful.

It's like needless we're spending needless energy on something that's not going to add a ton of value to your organization...

The org design dilemma between "Hiring Well" vs. "Firing Fast"

so that's the hiring well thesis, but there's actually another thesis. The "fire fast" thesis that you do not see as much of an in Silicon Valley for reasons that are understandable. But actually it's used in startups and they're successful companies now that are known to have like very high bars. When, when they started out, their philosophy was much more fire fast.

So here are the problems with hiring. Well, right? Like one of the big problems with hiring well is you don't get diversity. You just don't. Right? Because like, if the interview process is so difficult, you're not taking risks on people that don't have a certain background. You have a lot of trouble hiring people that you haven't seen before. So when you hire for a new role, it takes you like, who knows how many candidates to just like, figure out what you're supposed to do with them. And then it's slow. So if you need to go really fast and you have a really high bar, except if you have an incredible brand right it can, it can take a lot of effort.

So firing fast, solves a lot of this. Firing fast. It's very simple, right? It's like you decide that your interview process will be much more lightweight. You're going to go after a much wider set of candidates. And you will be very good three to six months in at determining. "Are people bringing a lot of value for the company"

You don't necessarily fire faster the whole history of your company. You do it like at a, in a, in a short time period near the beginning.

and this seems like weird, but actually there's a lot of benefits, right? It means people when they join are super competitive, cause they're trying not to prove themselves. yeah, you have a culture of unsafety, which is a little bit dangerous. You might have knowledge gaps before people are leaving. You may have a bad reputation in the Valley, . But you can grow really, really fast. You can take risks on people that you wouldn't be able to attract otherwise. And you can build a really diverse team.

Now you just don't see this playbook a ton, but I think you could. And I think it would be very powerful, right? If you look at law firms, if you look at investment banks, they very much have a fire fast approach. It's up and out, right. And that attracts people who really like competition. Who work really hard. Right? And so that's not, that's not for everybody. Like that's maybe not the type of place I would like to work at, or that most of the listeners want to, but you can build a successful business that way.

And the reality is there are startups, but I will not give you names... who until they were like, well, over a hundred to 200 people... had this kind of mindset, I think end game right now, given the supply and demand and balance in the market, it's difficult to have because it'll hurt your reputation as a company. Right? Great people who are risk averse will not want to go somewhere necessarily where there's a chance that they may be let go in their first, like 6 to 12 months. But again, you know, if you're competitive environment, you're going to also attract the people who like to be part of the competitive culture and show that they can, they can cut it.

so if you're fire fast, you do not need to spend a bunch of time on your recruiting. It's like, just not that necessary.

Jerry Li: And a lot of companies, I think they all claiming that what we want to hire the best engineers, the best talent. And which is not practical, especially for like early stage companies. They want to move fast. and the other approach like the second approach you mentioned, fire fast, if they embrace it then it can work really well, but a lot of people, I guess they didn't realize this is a strategy. Like they sort of feel bad that they have to do this way, but you don't have to be, this is a perfect strategy for people to get started and then you can build momentum and then have a better brand.

Jean-Denis Greze: Yeah. And honestly, just be honest about it with people. I mean, I think today, since almost nobody does fire fast, if you did it and you were open about it, you would automatically have a great pipeline because there's there's people who want that. Yeah. There's, you know, like human beings are cut so many different ways.

So, you know if like the whole market has moved towards one philosophy... having an opposite philosophy, even though it's like maybe a little weird one can be a huge benefit.

Patrick Gallagher: just to talk about the impact of that. We had an interview with Farhan Thawar at Shopify a few months ago. And what he talked about was for some of the companies, he led, they had 15 minute interviews and then. They had the three month period and the results that they saw from team diversity and retention after that three month period.

So the people who stayed on and it was a good mutual fit, the retention was in the high nineties of their ability to retain talent beyond how they recruit people in. So it's just really, really interesting to think about how that thesis, that like really investing in a fire fast thesis as your recruiting strategy can lead to some of those longer term impacts that you don't have to focus on as much.

Like if you, if you nail that your retention might actually also increase as well, which is super fascinating.

Jean-Denis Greze: that's a great example.

I want to talk really quickly about the third strategy. Cause I said there were three pillars, right? Recruiting growth and performance management. You can also focus on growth.

this is, if anybody, any listener, have like read about the of the retail sector, large retailers, like if you read about the history of, of places like Walmart, Some of them had this philosophy, that to be like a VP you had to at one time, way back when, like worked on their shelves and then you'd wear like a store manager and then you were distribution center manager and then you were regional manager, right? Like the idea that people learn the craft and learn about every of the business by growing up in it.

so I haven't seen actually a pure growth strategy in the Valley. Obviously a lot of companies care about skill growth for employees. The careers. are generally not that long and that people now switch jobs every what... like three to five years. Right. It's pretty short. So it's, it's a big cost to invest in someone longterm if you know, you're not going to be able to retain them.

But I actually think it's like an interesting perspective to ask yourself, Hey, could I take a growth approach to how I build my team? I think if there's any CEOs listening, I think that's actually a very good approach across an entire, an entire company. Right. You're good at identify the talent. That's thriving, give them real opportunities to really quickly step up in your org.

so anyway, plaid it's thesis is we're very much in the Hire Well front Right. And when I get later to ways that you can take a spiky org. and add new elements to it without becoming quote "average" I'll talk about how we've introduced the concept of fire fast, actually to some of our funnel and why I'm really excited about it. and I think before I thought kind of systematically like this, about the extremes, I was very much in the camp of "We're really good at hiring. So let's just focus on that and not think about the rest."

And I'm trying to adapt and be like, " Hey, how do we mitigate some of the negatives of, having a really high hiring bar, you know, for example, to provide more diversity on the team.

Jerry Li: before we move forward, I think it's important to point out that there are three different strategies. you don't need to be good at all three, but what is important that identify which one you want to focus and be really good at that.

Jean-Denis Greze: Yeah. And understand the implications, right? The cultural implications and other implications.

Jerry Li: Yeah. And that can change probably over time as you know, the company transitions.

Jean-Denis Greze: Yeah. I'm trying to save people time, right? I'm trying to save someone from, focusing on something that is going to add incremental value to the org. because they've already solved 90% of the problem by going really deep into one avenue of solutions.

The org design dilemma between business impact vs. craft and quality

The second example that I want to talk about, which we'll revisit again later is this idea of building an engineering product design kind of organization, that is focused on business impact versus one that is focused on craft and quality.

And these aren't total opposites. I think a lot of people try to do both of these things well. And I think it's very, very hard. Like the hardest question that I ever get from, companies that I'm advising or other VP Eng is like, "how do you know if you're investing enough in reducing technical debt? how do you know when you've invested too much in moving fast, that it's going to hurt you long term?"

This is like the very, very hard question to answer. I have frameworks for it. We can do another podcast about that, but it's hard. And the reason it's hard is because it's hard at the VP level, but it's hard all the way down to an IC that's working on a project. You know, are they, are you building it well enough? Or are you building it too quickly?

And if you ask people to answer that question all the time, in my opinion, it really slows down your ability to do either one really, really well. Right. If people were optimizing for craft they would build like longterm APIs, everything would have full tests, everything would be super reliable and so on and so forth.

and if you focus on business impact, you would get business results all the time. You would drive ARR, product adoption, whatever your, your top metrics are for your company. It's very difficult to do both.

Jerry Li: And also not just the skills and the focus, but also about the people that are working on that. Because some people are naturally, excited about the impact they're making. Where other engineers, really like the craft. They feel good by having good design and clean code. And, you know, everything's still reliable.

Jean-Denis Greze: Yeah, but also like what does your organization value? Because if, if you're a craft engineer in an organization that supports the business impact, are you going to be able to get, are you going to get promos? are you going to be allowed to work on a 12 to 18 month timeline project to do something really well when that's not what you need to, you know, to get to the next level or to get respect from your manager?

I've heard, it said like, you know, to get some of the really senior promo's you like needed to build a new service. , like at some company that , really cares about like craft and scale, you need to build a new service from scratch.

So people were looking for opportunities to build new services. Even if you didn't actually need to build a new service for X to be successful as a business.

Jerry Li: Yeah, I've seen that a lot.

Jean-Denis Greze: Okay. so you have these two they're kind of extremes, right? And obviously if you drive for impact, you have lots of issues. Like impact seems nice, but impact generally means it's a known ROI, but what if it's a project that has a 10% chance to succeed, but if it's exceeds, it's a home run. Like, are you going to prioritize that in an impact org? Very difficult, right?

Longterm projects and impact course, very difficult, risky projects, very difficult to take on. You're much more likely to be tactical and a high ROI with short horizons. But you're going to have this huge gap. You may not be as creative because you may not be willing to take the risks. You may have revenue pressure leak into EPDs, which is like a dangerous , failure mode. You're gonna have a lot of technical debt. cause the technical debt at any one point in time any project introduces minimal, technical debt, but across all of your projects, you've like halved your ability to ship things.

and then if you're craft world, your negatives are, you could be building bridges to nowhere... projects that have no connection to business outcomes. you are prone to too much craft, which can lead to things like huge deadline slippage, because you don't have a good way to measure against like a kind of a business thing. On the, on the other side, you're going to have solutions that are looking for problems.

BUT if you're a business that cares about reliability and uptime and stability that is a great set of values to have. And there's some businesses that need that, right? There's some businesses where it's all about uptime and quality and speed. Right. and in that case craft, maybe the right thing to do, or you may have a team that is so outsized, you're making so much money that you can have so many engineers that it's okay to focus everybody on craft, because some of them will land by luck on things, you know, by luck on things that have high business impact.

Anyway I'm pointing out the two opposites because I think if you look at smaller startups... often very, very focused on business impact, right. And then quality starts to fall. And what happens is they then try to like do both business impact and craft well.

And then they end up in this like, bad place where they're kind of good at neither. They're not fast anymore, but they still don't have like great reliable systems. So, what I would advocate is when you get to that point, you kind of make a conscious decision and you say, like "Hey, listen, we're about quality now. And we're going to really focus on quality and we're not going to try to do both like moving fast and quality. Like we're now in the quality phase of this business."

And we spent, we spent a lot of time and we spent a year in it. I mean, you don't ship anything new, but you really switch your values towards like leveling up the bar of your systems. And then eventually you can decide, "Hey, the time has come again. We have now a stable foundation. The time has come again for us to reintroduce, like, a business mindset "

And then you can reengage that gear. And I think that's a much better strategy than trying to build an org that's good at both. Cause very, very, very difficult.

How you know when you should change your strengths, values and build a new "spike"

Jerry Li: Yeah, a couple questions here. one Based on your experience at Plaid, when do you know that this is the right point to make a switch? And second question is how do you get people aligned with that? Because they're so used to shipping fast because it's instant gratification .

Jean-Denis Greze: so I think the strategy that we use that we are using at Plaid now is we're trying to have, you remember how I mentioned, we were like three orgs in one. Right. So the reality is we're, creating very different cultures that have a different mindset to get around this.

So I still want to live in a world where no engineer has to like, think about both. There are some engineers that are really focused on business impact, and there are others that are really focused on craft and they live in different cultural and values . I'll go through the history.

you know, we had a pot of gold in front of us, like for the last four or five years as a company. and. For really long time, that goal required a lot of short term business impact projects. And so we really built an organization that was really good at that. and we, I think have pivoted a little late, like about 12 to 18 months ago, we've pivoted like much more to the quality and the craft side. And the leading indicator for us is honestly, we just couldn't get our NPS of a bad place. and I think we waited too long.

And I think now I'm talking to customers a lot more. I'm a lot more in touch with that side of things to try to see the way the wind is kind of sailing. And then what we've done is we've created bubbles of impact within an organization that's more craft and quality focused. That's kind of where we've landed today.

but again, like the goal is to just never have anybody have to make the trade offs like day to day, because I think that's, what's really, really difficult for engineers.

most everyone is like, yeah, okay he just said nothing. Like most companies obviously have some people that focus on quality and some people... but I guarantee you, if you actually look around... way too many people are thinking about both. and that's what I'm trying to get as a takeaway here is like, if you're trying to do both. You're optimizing for things that by definition have tension. And so you've NOT built an org that has a spike towards the thing that's most important for your business. And as a leader, if you can identify that, I think then it opens up a world of possibilities that's, that's pretty exciting.

Jerry Li: I think this has a particular large impact for engineers because unlike other professionals, engineers when they work with day to day. They have to make a lot of decisions... on different levels... constantly. So if that is not clear, craft versus impact is not clear, then they're gonna struggle on all those small, tiny things. the overhead everyone accumulated on daily basis to struggle between the two is huge.

Jean-Denis Greze: Yeah, totally. I mean, it's, it's not even in your head, right? It's like you're writing a technical spec and then a staff engineer drops in and the staff engineer is like all they care about quality and you're like, wait, hold on a second. I thought I was on a team that was supposed to focus on impact and that it was okay to cut these corners? Now I'm really confused. And then there's going to be like a flame war... that doesn't have to happen if people were just totally values aligned on like how we're going to accomplish our goals.

it's a hard job. Right. every level is difficult.

The dilemma of building an organization with bottom-up vs. top-down decision making

the third example is it's kind of shorter. It's about like, whether you want to build an organization, for example, has bottom up decision making or top down decision making. It's really hard to get both of those at one place. Like if you try, right. If you try, if you tell everybody it's bottom up, it's bottom up, it's bottom up. And then the CEO comes in and says like, no, we're going to do X project... so much credibility has been lost so much energy has been lost...

And then if, if you say top down, but leadership is not held accountable. And does it stand behind its decisions? you know, then there's like a blame culture or the bad version of a blameless culture where you don't even know who's like the root cause of decisions that are being made because you have leadership that presents as top down, but expects things magically to happen from below.

and I think on that one, I've tried at Plaid. We've tried be bottom up. and we try to have it bottom up for everything that's below I would say the quarter boundary. And we've been successful at that. we're trying to get it to like six months, but anything that's, any decisions that are made that are like, Like multi quarter or definitely a year. I think we're still in a world where that comes top down, but we're fighting hard to get the bottom up to be able to handle, like, more of the length of the planning cycle.

and so what we try to be explicit with people are like, where is it okay to be top down and where is okay to be bottom up. But I think as you scale bottom up so powerful, if you can get it right, it makes your organization flexible even within whatever spikes you have, cause you have, if problem on the ground, people on the ground who are like talking to customers and understand the domain, they understand the technical domain, the product domain, the design domain, the support domain, but they have that expertise that you quickly lose an a startup as you grow.

So often here, to the extent some of the listeners are VPs at, you know, companies that have say like a hundred to 300 engineers, There will be a point in time when the founders and the initial set of like product engineering and business leaders no longer understand the market.

Right. You know what I mean? They, they think they do cause they're having conversations with a few customers and they understand the market, maybe high level strategically, like great, great leaders should. but they don't, they don't actually. Get enough signal on individual customer verticals anymore to really know what's going on.

But you've lived in a world where the company has been built on the back of this founder's genius idea, right? Once upon a time. And so you don't realize as an org that top-down is no longer the right thing for you, because "top" isn't getting the input that it needs to be successful. And that's like a... it's a pretty dangerous point. You can see it where the, basically the founders intuition is no longer aligned with what the market needs. And as soon as that happens, you need to like pivot. As soon as you realize as a leader that might be happening either you feel yourself not really understanding the problems anymore, or you see the founders or other leaders, not no longer understanding them. You need to A) define the role of leadership and it needs to be only about strategy, right? It can't be about individual decisions. And then you need to really foster I think a bottom up environment.

If you want it to live in top down, then you need to make sure that top down. Top has a signal that it needs to make. Right, right. Decision. I don't know a ton about that. Like I personally like philosophically, I like bottom up. So that's what I spent my time thinking about doing, I haven't thought nearly as much about how you can make top down work, but there's lots of orgs that have this kind of command from the top. A framework that are very successful. Right. So I don't, I don't mean that it's not impossible. It's just outside of my wheelhouse.

But I think again, you know, the mistake is when you think you're in one, but you need to be in the other, or you're trying to do both at once. And that just, that leads to bad things.

One of the things that I tell any VP at a fast growing place is like, "when are you going to be aware that your leadership no longer knows enough to make the right decisions? And how are you gonna manage that? And how are you going to like message it clearly to the team so that they get it that now it's their responsibility?" It's not going to be a founder like reading a planning doc or an OKR saying this isn't the right thing anymore. Like, it's on your shoulders. It's on YOU like tech lead this on YOU PM. It's on YOU designer. Right? Cause if people don't internalize that, then you end up in that world where no one's clear about who's accountable and how decisions are made.

Jerry Li: Yeah, I think the notion of not doing both and there's a transition point, being aware of that just provides a framework also language for people to have a conversation.

Jean-Denis Greze: Yeah, one thing that you said it was powerful, Jerry's the idea that ahead of time, you kind of know that you know, the second time around that these things have to happen. Right. it's kind of this idea that there's like a map. as companies go through stages, the kind of challenges and things that have to happen. And that map doesn't exist. Like I haven't seen a management book, but that tells you, like, "as you grow here, how the things break here are the decisions that are important to make..."

but I've seen the failures, right? We've all seen VPs get hired. And then six months later, like being, let go. And I think sometimes it's because they weren't the right person, but I think often it's because leadership is not ready to hand the reigns or they don't even know what that means and what that looks like.

How to develop new strengths, capabilities or “spikes” in your engineering organization

Okay so I think for the examples, hopefully those three are kind of enough to explain what we mean by spike, which is like, there's these things that are kind of counter you're better off just picking one, really optimizing for it. Yes. That will give you a lot of blind spots and you're going to have to manage the blind spots. but I think there's these three principles, these three things that if you're a great leader, you need to make time to do.

the first one, you need to understand the problems that you're trying to solve. Okay. That's obvious. We all do that. The second one is you need to understand what kind of organization and organizational spikes you need to solve those problems. So in like management land, these are called capabilities. What are the key capabilities that you need? Right. But you need, to understand those two things totally in the abstract of what your team is today. Don't don't ask, can my team do X? No, no, no, no. You say if I need to do X, what capabilities would I have in an ideal world? Okay. And third thing you're going to ask yourself is if I have those capabilities, what are going to be my blind spots? What are all the things that I'm going to be really bad at?

That's the theoretical exercise. I think you have to do it, but you have to get to that place where you're not just observing the blind spots, For you. You have them in the back of your mind. It's tough. It's like, you're kind of guessing it may feel like bullshit, but I really think it's worth it.

Once you have all that, then you need to decide how do you transform your current organization towards the one that has the strengths, that the problem ahead of you, requires? And what I'm going to talk about are examples of Plaid where we maybe didn't have the strength that we needed and kind of how we dealt with it either by creating the spike or by finding an opportunity to have multiple kind of strengths that in the normal world would, would fight against each other, but that we could get to work together.

I think like at the end of the day, There's lots of parts of management, like leadership. Like you got to recruit great people. You gotta make sure people are happy. You gotta have business goals, but at a really high level you're designing an organization that can do those things. And organizations are about knowing the problems, building the strengths you need and finding a way to develop them and understanding the weaknesses that you have and either mitigating them or making sure that those weaknesses are not going to stop you from being successful. Because if they are, then that means you're optimizing for the wrong spikes and then you need to redesign the thing.

Jerry Li: So the initial process of identifying the, the capacity and figure out a blind spot is essentially building a map. where there's everything we need to pay attention to, and then find a way to get to the destination.

Jean-Denis Greze: I think that's right. Yep. People already do it. Generally the, the dimension where most leaders think about this is around talent management. they're like, Hey, I need to build a data science team. I need a red team and security. There's like specific problems like that. And they realize they don't have the capabilities. They're like, nobody knows machine learning on the team. Let's hire some. and that's great around talent. I don't know, like I appreciate.

But I think people need to think about it about every part of org design. Should we be a matrixed org, a vertical org or horizontal org. You're seeing all the time. You should understand, how you map say Go to markets to product and how that is, or isn't the right thing.

Like at Plaid right now, we have a vertical organization for products that's per vertical, like looking at new developers, looking at people who use plaid for moving money around, using people who use plaid for lending purposes, looking at how consumers use plaid. We have these vertical teams, right? And they all kind of touch the same surface areas and we've not built good platforms right beneath them. Because that's something platforms are something that horizontal teams build. Not that vertical teams build.

I will wake up once a month and I'm like, is it the time to change it? And I know what I have to do to change it. It's not that hard. Right. I'm going to build a new group. It's going to be called product platform. And it's going to build a product platform. Right. But I'm not doing it now. Cause I just don't need to, it's not what I need for the business to be successful. But I know it, I know I have to do it.

And the reason this is really important is because I'm not going to wake up six months after I was supposed to build it and be like, Oh, F... I should have built that six months ago. Right. I'm going to get it at the right time. Cause I ask myself once a month. So I can be off by like 29 days, but I'm not going to be off by more than 29 days. That's an exaggeration because I probably still will get it wrong, but I don't think I will get it as wrong as if I wasn't thinking about it.

Jean-Denis's process to create space for questions, creativity, and problem solving

Jerry Li: Yup. And if you have time to cover this. I like to dig into what are your list of questions you check yourself with, on a regular basis? That can that seems really interesting.

Jean-Denis Greze: Ooh, like you mean as a, as a manager, like what is the long list of things?

Jerry Li: Yeah, just like the one you mentioned , every morning you ask yourself, "is now the time to build , a product platform team. So you are not too late. and knowing that, the question to ask ahead of time is it's value.

Jean-Denis Greze: my God. I'm gonna, I think I'm going to use a background process in my brain. In the next, like 20 minutes to come up with a full list, I'm more of a process oriented person than like a checklist person. So I'll tell you what I do. I set aside time every week to walk around... and yes, I walk around with a mask, without my phone... and I go walk around for at least 60 minutes and I try to make it 90 minutes. And during the walking around time I have a piece of paper and a pen, and I asked myself to think about things that are not working with the org. And, that is how I do what you're talking about.

And generally when I walk around and ask myself what's wrong with the org, what's not working, what would I design differently? Nothing happens. Like I don't come up with any good ideas. It feels like a waste of time, but I am walking outside, which is nice. And then once in a while, there's a good idea that comes about. and then the good ideas I have, like a, I have a document. I have like a a thing where I write my potential future OKR is right. And I'd like a list of things. So I just write the ideas in there and then I review them. And I think from the ideas I get the, the questions that you are talking about.

I mean, right now there's a bunch of stuff that's caused by, by COVID. But I think a historical question that I've always had is, whether when is the right moment to endorse distributed more. From a, from a talent perspective, like, and just, now suddenly it's like top of the stack, whereas before we'll talk about how we've organized, but before it was something that was kind of in the background that I was always, thinking about.

The other question that I ask myself, like all the time, all the time is how I can give more scope to the people under me so I can do less work. I think it's it's very easy to assume that people can only do the things that they're doing now. Ifeel like I have to force myself to like put the hypothesis of, can this person take on more and what's the real risk to them doing it?

People change, you know, it's really easy for me to look at one of my managers. That's going a lot. And remember the version of them like a year and a half ago.

Jerry Li: Yeah. And that takes a lot of initiative from you to realize and to find out there's opportunity. And then sort of have to challenge people and also align other projects or ownership so that person can be challenged in the right way. But do you set up a time like you do put it on a calendar and...

Jean-Denis Greze: yeah. So my calendar is a wreck One of my EA's, asks is that she find this time for me. I must say the last five, six weeks have not been conducive to much longterm thinking, because of everything that's happening. every engineer, every person in the world is experiencing some version of this, but yes, so, so the idea is there's always a walking time. And then the only thing I have to do is remember not to bring my phone. If I succeeded in not bringing my phone, then I get the time. If I don't do it during the week. sometimes I'll try to do it like on the weekend, but I try to have , some separation.

and actually this trick is not mine and, but I urge listeners to try it. And you don't even have, honestly, you don't have to do like 90 minutes, but you have to do more than, I think you have to do more than 30. Like, cause what I, this is the way it works for me. Like I start walking first 15 minutes I'm definitely thinking about... whatever the candidate I'm trying to close, or like the fire that's happening or like who I have to hold accountable for a sev or who knows? Right. There's like something that's in my

Jerry Li: You're not unplugged yet.

Jean-Denis Greze: Yeah. And then suddenly I get to a place where I'm kind of bored. And I remember that I have this like notepad and I have to write on it. And then I can't think of much. And then eventually I get to a place where I'm kind of free flowing with about an idea. It could be a product idea. It could be org design but it's something it's like the prompt for me is something that I'm not normally thinking about. Like that's what I try to force my mind to. And then yeah, something sometimes, you know, something good comes out of it. something that has real value,

like the last big question for me, but we've asked, we answered it like six months ago was, when we need to make a hard pivot towards like better developer tooling and more developer efficiency and not just having this one group kind of taking on this craft mindset, but having it go a little more broadly, That's been like a big question in my mind. I do think I waited too long for it, but it ended becoming the theme for the year. and getting to that decision, realizing it was the right one getting buy in from my directors, my managers, senior ICs, making sure that it was correct. You know, that took a really, that's a really long time and a lot of. A lot of work. And so I think after that one, I felt really good about this year. We're about to enter annual planning for next year. So I'm probably going to force myself to ask myself all the questions one more time.

Patrick Gallagher: Really powerful practices. Cause I think about in the context where I've heard some practices like this to expand your mind, Relax, your focus oftentimes comes with like artists or, you know, inventors from history. And so it's really cool to see it in this context and how you create space for these ideas to come up with and driving your thought process with really good questions and being really disciplined with asking yourself those questions over time

Jean-Denis Greze: I think Chamath, who used to run growth at Facebook and run social capital and if you follow him on Twitter, I believe, that he still does it, but he used to go to, he has like a house, I think in Tahoe or whatnot, he used to go there for, I don't know, three or four weeks. A year in one chunk. And the idea was to come back with the, like the annual plan for social capital for the next year before. just this idea that like he needed to be in a different space and really focus on this longterm thinking to get his brain there.

And it wasn't like 90 minutes, right. It needed to be a huge chunk of time. Right. Or you can see executive teams when they go on off sites where they go on three to four day off sites. And yes, it's for some team building and you need to build trust across leads that don't often interact, but it's also different environment, right? you can think about different things. We'll talk about that because I call this outlets later.

But, as a leader, you have to force yourself to have different perspectives on the same problems. and no one will remind you to do it. So this is the difference. Like if you're in a line manager, like your director should be reminding you to do that. They should be telling you, like you are being too tactical. You haven't brought any new, big ideas to me. But if you're at the top, no one else is reminding you, but it is a core part of the job. And I think it's the difference between like leadership and management. Like leadership has I always like to phrase it, like you only make a few decisions that are really important to the business. You make them rarely, but you need to make them, if you don't make them, you've lost the opportunity, right?

Like I'm not just managing things. Every once in a while. I have to make a hard call. And if it's the right hard call, Fantastic. And if it's the wrong hard call, the only way you find out is there's an ultimate reality where we were more successful. And that's what you're accountable to. You're accountable to creating the best reality that you can. And it's hard.

So I think a good CEO, a good board, right? Those are, those are the things that need to be asked out of the leadership team. That's like, that's kind of the role.

Walking around for 90 minutes helps, but that's like, it's like a cheat code, right? It's not the. The real substance is you make it part of what it means for you to be a leader, to have this kind of thinking.

Jerry Li: Those idea came seemingly like, serendipity, but still people need to manufactured the serendipitous, idea that came to you by setting the time so that you unplug yourself and stop the thinking. To me. I think those moments, epiphany moments come really, when you stop thinking, and then suddenly some idea emerged. If you keep plugged and thinking about all the things you have been thinking of all day long, then it's really hard.

Jean-Denis Greze: Yeah.

Patrick Gallagher: for anybody's in the Bay area, the author Jack London talks about chasing inspiration with a club and what I take away from that and what you shared, Jean-Denis, is that to do this and do it well, it's the consistency and the discipline of the practice that then helps get the result, which is the idea and the answer to those questions. And they may not happen every time, but it's the discipline and constant practice of that. That helps make it work.

Jean-Denis Greze: I agree. I wish I were creative every time. I think maybe great leaders are right. That's the other thing, maybe I'm just not that good, but it's, it's quite disappointing. It feels like a waste of time often. That's the worst part of it where you're like, Oh, I could have been sending emails or I could have, you know, I could have been like spending time with people. But you HAVE to make the time. The brain is not, you know, you don't control the brain. It will do what it needs to on its own. It's own pace.

Patrick Gallagher: How do you let go of the productivity guilt, where like you're sacrificing creative time for. The other sort of completing...

Jean-Denis Greze: now my answer is not a good one. My answer is I'll still get all the other stuff done. It's yeah. I mean, that's, you know, I have like an over overdeveloped sense of responsibility.

So I try to do them in the morning. Right. The reality is like, Before I'm tired. And before too much of the craziness has gone. And if I have something to do, it feels like, well, I have all afternoon to do the thing. You know what I mean? It's, there's like, I can lie to myself that I'll, it'll be okay. And I have time to do all the things.

Whereas, you know, if you, if you do this, like if you tell I'm going to do this at 6:00 PM, right. Then if you have anything that you needed to do, you're, you're just, you're going to sacrifice the walking time, because you know, you don't want to be up until 10:00 PM

Dive in
Building a Successfully "Spiky" Org (Part 2)
Jul 26th, 2020 Views 888
Building a Successfully "Spiky" Org (Part 2)
Jul 26th, 2020 Views 888
The New Requirements of Technology Leaders in the Generative AI World
By Noah Olberding • Oct 9th, 2023 Views 215
Building your data engineering org
Jul 19th, 2022 Views 1.6K