Design, Growth & Evolution of Product Organizations - with Casey Winters

BOUNDARYLESS CONVERSATIONS PODCAST - SEASON 4 EP #3

 placeholder
BOUNDARYLESS CONVERSATIONS PODCAST - SEASON 4 EP #3

Design, Growth & Evolution of Product Organizations - with Casey Winters

Casey Winters explains how the growth model is a peculiar aspect of each company and how reducing friction is key so that growth can flow more strongly. We also discuss how the responsibilities to drive network effects often moves from local teams to more centralized functions over time, and finally how leaders should be thinking about autonomy when managing growth and product teams.

Podcast Notes

Casey Winters is a growth advisor and operator that helps companies scale. Casey is a legend of product growth and marketplaces, and, among many other things, he’s an advisor and operator that helps companies deal with the problem of scale. Most recently, Casey was the Chief Product Officer at Eventbrite: before that he had tremendous experience working for and advising companies such as Grubhub, Pinterest, Airbnb, Canva, Thumbtack, Reddit, Hipcamp, Faire and many others. Casey also is a partner at Reforge, one of the leading growth programs you can attend.

Besides some real world experiences from his career, in this episode, Casey shares insights about how it is possible for a company to leverage on its organizational model as a lever for growth. 

Casey also explains how the growth model is a peculiar aspect of each company and how to reduce friction is key so that it can flow more strongly. We also discuss how the responsibilities to drive network effects often moves from local teams to more centralized functions over time, and finally how leaders should be thinking about autonomy when managing growth and product teams.

 

Key highlights of the conversation:

  • The Organizational structure can change according to strategy especially at the early stage of a business
  • Growth model: what is and what structures you need around that.
  • OKRs and creative spaces: what is the balance a leader should put in a team?
  • How can marketing and growth teams work together?
  • Brand vs performance: how a startup should think.

 

To find out more about Casey’s work:

 

 

Other references and mentions:

 

Casey’s suggested breadcrumbs (things listeners should check out):

Recorded on 17 October 2022.

 

Get in touch with Boundaryless:

Find out more about the show and the research at Boundaryless at https://boundaryless.io/resources/podcast/

 

Music 

Music from Liosound / Walter Mobilio. Find his portfolio here: https://blss.io/Podcast-Music

Transcript

Simone Cicero:
Hello, everybody. Welcome back to the Boundaryless Conversations Podcast where we meet with pioneers, with thinkers, with doers, with entrepreneurs. And we speak about the future of business models, organizations, markets and society in this rapidly changing world. I’m Simone Cicero, and today I’m joined by my not so usual co-host Manfredi Sassoli de Bianchi. Ciao, Manfredi.

Manfredi Sassoli:
Ciao, everyone. Glad to be here and join the conversation.

Simone Cicero:
And we definitely have a very special guest, Casey Winters. Casey is a legend of product, growth, marketplaces, many other things; now advisor and operator that helps companies dealing with the problem of scale. Most recently, Casey was the Chief Product Officer at Eventbrite. But before that Casey had a tremendous experience with advising companies such as Airbnb, Canva, Thumbtack, Reddit, Pocket, Hipcamp, Fair and many others. And is also partner at Reforge, that, as you know, runs one of the most interesting growth programs around. So, Casey, thank you so much for coming here. Hello.

Casey Winters:
Yeah, thanks so much for having me.

Simone Cicero:
Let’s jump into the meat of the conversation, and I’m sure that we’ll have a lot of food for thought tonight. So, when we reached out to you in preparation of this conversation, we shared with you that we wanted to ask you to talk about something that you not so usually talk about on many of the podcasts where you’ve been featured, which is, I would say, a little bit of focus on organizational structures. And all these interests started – to talk about this with you – by reading your not so recent piece called the types of product team organizational structures.

Which is a massively interesting post that you shared, if I remember where something like a year and a half ago, where you essentially covered the different ways that people can organize in a growing organization, or at least in an organization that is focused on growth; there is any pattern that you have been seeing relatively often, and especially also to double click on a very interesting point that you brought up in the preparatory conversation, where you said irrespective of the model you choose, you have to be able to understand your shortcomings and act on them and react, let’s say, to the issues that your organization brings up as you move forward.

Casey Winters:
When you think about the product organization, right, which is normally: think of product managers, as the core of that, though different structures can be slightly different. You know, I manage the product design team and the research team, the growth marketing team, in addition to product management. But generally, the most important thing you’re trying to get right at most technology companies, is to make sure that the teams that work most closely together have the exact same org structure, even if they report into different leaders. So, in that case, it means engineering, product management and design, and maybe some closely related functions should have basically the exact same org structure.

Obviously, engineering is going to tend to be a little bit bigger, but in terms of the team structure, and then the pillar structure, and then themes, however, you want to organize up to your executive functions, they should be the same, and they should have the same types of goals. And a product manager, a designer, an engineer should basically be sharing the same goals instead of be trying to work on very different things. I think that’s a concept that’s become fairly dominant inside of technology companies more recently. I think what’s also starting to emerge is an understanding that there’s different types of product work that you can work on inside a software organization. And there’s different talents. People are generally better at one than the other, right?

So, you could be working on totally new products like trying to do that zero to one new product innovation, you could be strengthening the core product with additional features and making sure the core remains healthy. You can think of that as feature development or a core product work. There’s the growth area, which I lead at Pinterest, and that’s about trying to connect people to the existing value of the product. So, you’re not building anything new, you’re trying to reduce friction that’s preventing people from seeing the value, or just get more people to it, like acquiring more customers. And then there’s things like technical scaling, which is helping the company be able to support more and more users and helping the rest of the engineers, designers at the company be able to continue to build quickly as you become larger and larger, right? So, those are like some core structures that I think most companies in some shape or form now have.

Beyond that when you look at how teams organize around their goals, you start to see some major differences, and you start to see some mix and match as well. So, one of the structures is just aligning people towards the type of work they’re doing like you just mentioned. You also see structured by strategic initiative for the year, right? So, we have three strategic initiatives. And then therefore, we have three pillars of our company. And every engineer, designer, PM is organized around that. Particularly in marketplaces, where more complex structures, you’ll see top level organization structured by like, which customer you’re supporting, right. You know, at Eventbrite, for example, right, you have, this is the creator teams, and these are the consumer teams because we work with both event creators, if you sell tickets and then the consumers who buy them.

Or you might have an organizational structure by the value propositions of the company, right. So, at Pinterest, we were structured around discovering, saving and doing as the three different value propositions. And then we had core product teams that were associated with that. And what you also see is that there’s like a mix and matching of these, right. So, at Pinterest, the core product team was organized by value proposition. But then we also had a growth team, which I was on, and we had an ads team. And those were combined with the value proposition teams to create the overall structure. So, you have these different structures, none of them is necessarily a dominant structure. And I think the important thing to understand about any sort of structure is that the structure is going to have flaws. And those flaws are by default, going to be showing up in the product somehow.

If you have one team working on the new user experience, and one team taking over after that, generally, the spot in between is going to be like a little shaky. There’s a term for this. It’s called Conway’s Law. Basically, you ship your org chart. So, no matter what structure you put in place is going to have some issues with it. And those issues are going to show themselves in the end product by default, which is just as important as figuring out how to structure the team is what processes go along with it to help mitigate the flaws that might show up based on your org structure. And we could talk more about some of those examples.

Simone Cicero:
That’s really interesting. And I think for us it’s great that you mentioned Conway’s Law, because as you can imagine we are a bit like organizational nerds. We like to talk about the Conway’s Law all the time, right? I was thinking that maybe the fact that you ship your org chart and what you ship is so important in a product organization. Maybe we have to think about the org structure more as a tool in product organization. So, for example, we go through a moment in the life of the organization, for example, where we have to focus more on acquiring supply, and then maybe it’s important to organize the supply side part of the organization. So, like a team that looks into that. So, is it true that in your experience, that maybe the organizational structure can change more according to the strategy when you deal with a product organization? Or is it something that is just my fascination?

Casey Winters:
I think that’s definitely true, and I would say it’s probably even more true, earlier stage of a business you’re in, right? So, to take that example, right, where you’re focusing more on supply, that’s generally a need of marketplaces earlier on in their career. And then later on as they develop, they can tend to lean more into demand. And one is you can completely restructure based on that. Or you can reallocate resources toward a more stable structure between supply and demand. And you know, you do that reallocation whenever makes the most sense. But I think you’re absolutely right that org structure is a tool. And the best org structure is not just the best org structure, it’s the best org structure for a moment in time for the current strategy and size of the company. And then when you say double in size, which startups will frequently do very quickly, it almost always is never either the right org structure, or the processes that we’ve put around that org structure are no longer scaling, and those need to be revisited. Maybe it’s both, but it’s usually at least one of the two.

Simone Cicero:
Gets me back to when Steve Jobs said you first design the user experience and then you develop the technology backwards, which is a bit like looking at the product experience first and then developing the organizational structure backwards. Which I think is definitely a perspective that the organizational design world doesn’t embrace very often, I must say. And it’s totally in line with our kind of the things that we preach at boundaryless where we always say you need to embrace an ecosystem driven force, shaping the forcing function to your organization you had to design for the ecosystem.

So, I was thinking, in your experience, how much the type of product organization you’re building influences the choices. So, is there any correlation between, for example, let’s say you’re building a certain type of marketplace, let’s say for example, a location intensive marketplace or a global one or maybe something where the relationships are more high value, high AOV, versus high frequency. Or do the property of the network, for example, when you design a marketplace, or the particular kind of perspective or software as a service company you’re building, of course they do. But in what ways do they influence the organizational structure that you have to think about as a leader in such an organization?

Casey Winters:
I think they have a massive impact, right? Obviously, in a SaaS business with only one type of customer, you’re not going to structure your teams around customer, because there’s only one. Everyone would be on, like the exact same grouping, and it would be really hard to manage the work effectively. So, definitely, right. Like one of the things I’ve talked about in the past is, how local is your network effect in some of these network effect businesses, right? So, take some companies, they have extremely local network effects. I work for this company called Ritual in Toronto and they are basically like the Starbucks loyalty app for every other independent restaurant and coffee shop, where the network effect is basically how far you’re willing to walk from where you work, just not that far. Compare that to say, like an Etsy or an Airbnb and those network effects are much more global, right? It’s wherever you can ship or wherever you can travel.

The local network effects, I would say, over the last 10 years have biased a bit more towards decentralizing their org structure and having a lot more management happen at the ground level at the particular location that you’re trying to build initially, some form of liquidity, right? Enough drivers, and enough riders in the case of Uber and Lyft, enough restaurants and the people who order from them in the case of GrubHub or DoorDash. And that has a lot of challenges once you actually find liquidity, because while the effort to get that initial liquidity in a marketplace can be quite local challenges, how you get your first 1,000 people to order food per se, how you get like your first 50 restaurants. At scale, at least what you hope is a lot of those markets look like in terms of what they need. Is it going to be performance marketing that’s driving growth, is it going to be incentivized referrals that’s driving growth, whatever the case may be. If you’re decentralizing all of that ownership, then you’re not building these competitive advantages with scale, because you have expertise being built that can help every local market.

So, I think there’s companies like Uber and Lyft, that went hard in the paint on building out local GM structures and ownership, they really had trouble in the middle phases of building their company in that the expertise inside of like headquarters, for lack of a better word had exceeded what the local GMs understood being kind of like general athletes. But these people that had less expertise were telling the experts what to do. And many times shifting strategy too often or doing things that were generally suboptimal for the marketplace’s health, right. And now, those companies are past the scale where they run into those problems. The local GMs have a lot less ownership. But one could make the argument Uber and Lyft would have not even gotten to the scale where this would have been a problem if they didn’t start with those very local decentralized structures, right. So, obviously, Uber and Lyft is a case where the optimal structure changed pretty dramatically over time. At a SaaS business targeting the same segment, you might not see that structure need to change very much at all, or it might be more micro optimizations.

Manfredi Sassoli:
Actually, I remember talking to a senior growth person at Uber, and they were telling me how they were shifting from a more kind of localized growth team to a more central one. And I remember also studying like a case study on Procter and Gamble, and they kind of, they’ve been huge for a very long time, right. And every five years, they shift from kind of more local to more central, simply because like things eventually always break. And then you start getting more value from either the local or the central aspect.

But apart from these kind of huge businesses, I was curious to ask you about kind of more kind of smaller businesses in the kind of scale up phase where maybe you don’t have these kind of massive economies of scale. And so when you say like, kind of a mid stage marketplace, you’re doing pretty well, in your local market, the single market could be the US, could be the UK, and you’re kind of structuring your company, at that stage, how much do you look at external factors versus internal factors? So, like, the people you have, the skills you can play with, the kind of talent you have at your disposal, or that you can hire?

Casey Winters:
Yeah, it’s a great question, and I think there’s definitely a trade off because I do agree with Steve Jobs, or Jeff Bezos that you really want to focus in on the customer and build the right experience for them. But I think practically as you’re building a startup, you need to be flexible towards the talent you have and the talent you can acquire. So, when I advise startups about structuring, like say a growth team and a product team or things like that, I’ll say something to the effect of like, hey here would be my plan A on how I want to structure it. But in finding someone like this, then switch it to this other model, right? Because talent is still scarce, especially at building at that startup phase, there aren’t that many people that can come in and have an exponential impact on how successful you are. And those people are not going to necessarily perfectly match the job description that you built in your head, or that you wrote down.

So, when you actually find someone that’s got 70% of what you need, you tend to pull the trigger on that hire as fast as you can. But you might change the scope of the role, you might change the org structure based on that. So, I tend to be very flexible in practice on the org structure, especially when the team largely isn’t there, that needs to be built particularly in that scale up phase as you’re talking about. Because I know talent is scarce and that once I get one of the pieces in place that will help me fill out the rest of the missing pieces. I guess another way to think about it, right, is you have certain skill sets that you need to bring into the organization. And depending on how many skill sets you get in the first hire, you change the job description for the second and the third.

At the start, this can be quite difficult because it can feel like a three body problem of like, well, I can’t figure out any role to hire, because depending on what I get from that role I’m going to hire something totally different. And I understand that point of view. So, I think that’s why you want to have a strong plan of what you want to build. And then as you need talent that could come in and help you, don’t be too rigid to that plan because talent coming into the day tomorrow that gives you 70% of what you need and a slightly different configuration is much better than the perfect hire a year from now, right, where you might be running out of runway or especially with startups, right, you just don’t have that time to get everything perfect. And done is much better than perfect at that stage.

Manfredi Sassoli:
Definitely. So, thanks for the very practical advice. I think it would help many founders to think about kind of how to hire and think about like a 70% fit, the right personality is probably a great match for the role they’re hiring for and actually that 100% fit might never happen. It’s risky to look for that. Another concept I’m really interested to dive into is the growth model, and if you ever kind of structure teams around that, maybe you could also kind of summarize what a growth model is for some of the listeners who might not be familiar with the concept.

Casey Winters:
Yeah, sure. So, I think of my job as a chief product officer is to try to map the model of how the business grows, and then figure out the right allocation of resources to unconstrained growth, right? And it could be that unconstrained growth is done by building new features or by optimizing, or by in some cases, building totally new products, like we’ve done the last couple of years at Eventbrite. But what a growth model basically is your best approximation as to how the company holistically grows, typically, the way we build these is the concept of growth loops, which are if something happens, like a person does an activity on your product, like saving a pin, in the case of Pinterest, what happens downstream from that? Well, Pinterest gets smarter about which pins are interesting to recommend that to more users of the product, which then gets them more engaged and drives more ad impressions. And that’s like one element.

And then another element is like, oh, well, that helps us get smarter about which images to show to Google so that we rank higher and bring in more new users. So, you start modeling these different loops together, and it forms some pretty complicated diagrams. But the important part is that the diagram has a lot of related elements. In certain related elements, if you’re able to unconstrained them, like say, doubling your conversion rate of people that view a landing page to signing up, or having a 10% improvement in the number of people that sign up that activate into being a weekly active user. Some of these efforts will not massively impact your growth, because they’ll just make another part of the constraint in the growth model even more constrained. In some of them, if you’re able to unconstrained them, everything downstream starts working better, and you start getting this exponential improvement.

So, when you start to map the relationships between all these areas of the product, and understand how they impact acquisition, how they impact monetization, how they impact engagement, then you can start figuring out how to allocate resources. But you know, more pertinent to this conversation, how you structure around those constraints. Because ultimately, you want as many people as possible in an organization to be attacking the constraints of your business versus the areas that are not constraints. Of course, you’re going to need to have people maintain other parts of the business that aren’t exactly constrained. But if you stopped working on them, they would be, that’s super important too. But especially in these startups, scale up examples, the focus is definitely around driving growth, either in revenue or users.

So, being able to map exactly how the business runs, being able to translate that not into just like a qualitative view that you can explain to the organization. But a quantitative view of like, well, in Excel or Google Sheets, I changed this metric, how did it downstream impact growth of the overall business over the next six months, is really what I tried to put in place, whether I’m working with startups or large organizations, and I think it just helps focus decisions on what to work on, on how to structure in terms of okay, well, like what’s going to have the most impact on growing this business mid to long term? And isn’t that what we’re all here for?

Manfredi Sassoli:
At Pinterest, you could have developed this phenomenal growth loop that carried the growth of the company for value very long time, right. And I’m wondering, like, when did you realize that you could do that? How long did it take to realize that you’re onto something big and what kind of resources did you decide to allocate? Was it kind of front loaded or was it something that you learned as you went along and you kind of added incremental resources to it?

Casey Winters:
Yeah, definitely learned as we went, right. So, I was fortunate in that I had come into Pinterest, right after GrubHub, right, where we had gone from series A to IPO. And SEO was a very big part of that. So, I and I think the company definitely had a thesis, which is part of why they hired me that SEO could be an important channel for growth. So, a lot of my early work and understanding the business was trying to spot like, well, what is the SEO opportunity going to be for Pinterest, if we have all this great user generated content? When I joined, they had one engineer working on SEO full-time, then we had a couple people who were part time out of a coding bootcamp that joined in to help us on SEO. And I came in to work on that as the PM. And we pretty quickly were able to show that we could drive more traffic through Google.

So, traffic is great and all but traffic alone doesn’t move the growth model of Pinterest. People need to sign up and become engaged pinners, which then helps us recommend great content, and then ultimately drive a bunch of impressions that we would much later on sell to people who want to advertise. Okay. We’re driving more traffic, that’s not very interesting, because the traffic is just viewing great content on Pinterest and then leaving and then we’re not really seeing them again. So, one of the things that we pivoted to is trying to improve conversion from those pages. And conversion was technically another team’s responsibility. But they were focused on conversion from the homepage and from viral sharing flows.

So, my team just went ahead and did it. And we did like a two day hacky experiment where people can view some content for free, then once they indicate that they’re interested in the content, either they start scrolling a couple times, or they start clicking on images, we block them and ask them to sign up to see more. We thought we could have like a 10% improvement in conversion rate, and the first experiment had a 50% improvement in conversion rate. So, at that point in time, like every metric, just kind of like kinks straight up at the company. So, the head of product is like Casey, like, “What the hell did you do?” And I thought I might get fired, right? Like people thought it was bad user experience, something that we did was kind of hacky. But the reality is, like, we showed that we could really inflect the growth of the business for the first time. We went and raised more money based on that fact, and we put more resources on SEO.

But still even that wasn’t like a fully optimized loop because people coming in from Google had a very different intent from people that were coming in via their friends. They didn’t have any other friends on the platform, so discovery, which was mostly friend based at the time, was pretty weak, which meant they weren’t activating near as high. So, people were like, oh, yeah, this Google traffic’s like bad traffic, like people aren’t activating. It’s just like vanity metrics. So, I said no, I don’t think that’s true. But we haven’t optimized our onboarding of this user with a slightly different intent. So, at that point, we started working on the next phase of, you’d normally consider a funnel, not a loop, which is improving the activation experience, customizing it for people coming with very specific search based intent.

And by doing that, over a course of a year and a half, we were able to double the activation rate. And at that point, we have like a fully humming loop, right, where we show content to search engines, people see a free preview of it, they sign up, we activate them with more relevant content. And then they start saving more content, which gives us more of a quality signal of what we should be showing Google. But then a bunch of them get the extension and they start bringing new content into Pinterest, which allows us to send more content Google hasn’t seen. So, then that loop started just like really firing. Right? And at that point we’re like, okay, well, how can we unconstrain on getting more traffic? How can we constrain on improving conversion rate? How can we constrain on improving activation rate? But fundamentally, like it was like turning really quickly. And at that point, right, now the SEO team’s like north of 10 people. We have like multiple people on each platform, Android, web, and iOS on the conversion team, the activation teams swarms to like 15-20 people. So, as we started really showing the return, the organization was eager to give us more resources to optimize it because they saw the benefit,

Manfredi Sassoli:
Right. So, by the end, you’re probably a team close to like 50 people from what it sounds like.

Casey Winters:
Yeah. Well, the growth team part of it, which is not the entirety of it, because of course, that recommendation system was powered by the core product team. But the growth team, I think it swelled to 75 people by the time I left, and I mean, now I’m sure it’s like hundreds of people a few years later.

Simone Cicero:
It seems like these kind of structures need to be designed in a very, I would say in a way that is mirroring the product you’re creating. So, essentially, you have very clear objectives, right? What kind of balance you as a leader in this organization you bring to the teams in terms of the mix between OKRs and let’s say, more creative space for them to be fulfilled at work or finding meaning in what they do. So, what is the human development thesis behind working in such an organization in your experience? What are people fulfilled by?

Casey Winters:
Well, I think it depends on the experience of the team, not only in terms of kind of years of experience working, but also their experience with that specific problem. So, one of the things I did when I started advising Eventbrite is I built this concept of the autonomy spectrum. There’s a bunch of things that companies work on to ultimately build something. And a top end that’s like the vision, and maybe things like the target user, and then you start getting into things like your product development process, like which ideas — who’s generating ideas, who’s prioritizing ideas as well as like how Sprint’s are run right at the very, like more detailed level.

And I told the team, like, I’m drawing the line here. The things to the right of this line you own, the things to the left of this line, I own. And as I see that you’re developing more proficiency, I’m going to move this line of what I own versus you own back, and I’m going to give you more and more. And the intention is to get you to be as autonomous as possible, but with an appropriate amount of stretch, so that you can grow and not too much stretch so that you fail, right. Asking them to define the mission of the company is going to be too much of a stretch, right?

So, for a team, like Pinterest, after we had built that growth loop, right, we were giving a lot of autonomy. We set the objectives and key results. We set the missions of the team, and fundamentally, it’s not about the roadmap we’re building, it’s about achieving the goals. And we have a roadmap that we think will get us there, but like we’ll learn as we ship some things and adjust the roadmap accordingly, what’s more important is the objective, and the key results, which are company level metrics that we’re owning. And we had to do a lot of iteration on how to set key results effectively between experiment based and measuring seasonality, and all of those kinds of things. And we just got better and better at it over time, right. Whereas Eventbrite, when I joined, a lot of teams were working on something new. So, they needed something a bit more well defined until they were able to become the subject matter experts. And then they get more of that autonomy spectrum loaded on to them versus being delivered by a leader like myself, or design leader or the CEO of the company.

So, you want to have that balance based on what the team is ready for in terms of experience and as well as their experience with that specific problem. But eventually, I’m a big believer in autonomy, so I think you want to get them to write like, here’s a mission for the team. Here’s how rituals are done on the team, here’s — at Eventbrite, we call it the Desi model, right, like, who’s driving, like who’s approving, like who are you collaborating with, and who’s informed. So, mapping all of that out and just agreeing to it, right, and making sure that their managers of that team have agreed that, okay, that’s the way the team is going to run. So, things like that really help at scale. You know, when you’re early, you can just kind of like mind meld and people get it. But at scale, having that stuff written down and making sure it’s been aligned on and approved can be really helpful.

Simone Cicero:
As I understand a lot of structuring organizations like these is about setting the aims for the teams in terms of organizational leadership, am I right?

Casey Winters:
So, I think what you’re trying to make clear is what’s the vision and mission of the business, what are the goals of the company for this year, and maybe a couple of strategic themes. And that’s generally what you’ll always want leadership to handle. And then beyond that, as much as possible, you want the teams to be able to react to that, and use their expertise to say, hey, based on the understanding, all of that, here’s the way I can have the most impact on those goals and on the strategic theme. And if they’re confused as to how to have that impact then leadership has to step in and build more structure until they build up either the experience or the expertise in that area to be able to do more of that on their own.

Simone Cicero:
The picture of the organization that I get it’s culture wise, for example, and I would love if you can double click on this idea of culture. Seems like a very disciplined execution of these kinds of skills that now we are increasingly codifying, essentially growth, product design, the things that you teach at Reforge, essentially. So, that’s what kind of impression I get. But I think there would be a lot also of collaboration co-design. So, I’m also interested to see in terms of culture, if you see maybe the potential to develop more cultures where more creative entropy is allowed to the people working in these organizations, or is it something that instead these types of organizations really tend to avoid, tend to basically, I don’t want to say hire people that are good at executing a plan, but sort of?

Casey Winters:
Yeah, it’s a great question. I think you want to be in the middle between those two extremes, right. So, that kind of unbridled creativity can lead to teams like going and building something that doesn’t make sense for the strategy of the company at all. Right? They should just go be startup founders. And people that can only develop what’s executed on are never going to be able to build those key insights that as a leader, I might not see because I’m not as close to the problem or to the customer. So, I think you want people that are driven by the mission, that understand what the strategy is, and then can use their unique expertise to translate that into what’s the biggest opportunity to help our customers, which means in the long run, we help our business. And the concept we use to describe that at Pinterest was the concept of knitting. And knitting basically means that when people from different backgrounds work together on the same problems, they produce solutions that are greater than the sum of their parts. That’s design, that’s product management, that’s engineering, that’s research, that’s analytics, etc., right? So, core to how we operated with this concept of collaborating of different disciplines creates better outcomes for our customers. And I totally believe that to be true, and that is how we operate at Eventbrite as well.

So, you definitely want, not everything to be prescribed to the teams, because then you’re not really able to leverage the expertise or the approach that that designer might bring versus the engineer, versus the PM or the data person, etc. But you also need to work your way to that in a few different ways. One is the trust between all those team members, right? They might benefit by a little bit more structure as they get to know each other. And then they can more freely form collaborate on the best strategy and growth map for their particular area. Two, what are the needs of the business, right? When the pandemic hit and our revenue dropped to zero at Eventbrite, we needed to be very deliberate about what needed to be done so that company wouldn’t go out of business, right?

So, at that point in time, teams need to accept, like, I’m going to throw out my traditional planning process, I’m going to throw out my traditional prioritization framework, and I’m just going to do this thing immediately. I know some of it might even not be the best thing. But like, this is the way the company needs to operate at this very moment to survive. We can’t go do a three month brainstorm, because we won’t be around in three months if we do that, right? So, you know, all of these things have a lot of nuance in terms of the right approach at the time, the right people on the team, the strategy for the company, are you doing like a bet the company sort of approach on something. So, it’s hard to be extremely deliberate in all cases, as to what the right answer is, but I’m biased toward, especially in peacetime, that collaborative approach, and giving as much autonomy as the team is ready for, and then pushing them to take on more and more overtime.

Manfredi Sassoli:
Do you make kind of marketing product and growth work together? And then you know, in a marketing team, you might have like growth marketing, performance marketing, and then potentially brand marketing. What’s your experience? What have you seen?

Casey Winters:
Yeah, we have both of those teams at Eventbrite, right. And the first part is thinking about, well, we all share the same goals, we have the goals of the company in mind, right? Which means then we’re not organized around marketing versus growth versus product versus engineering. We’re using our different disciplines together to unlock those goals right? At Eventbrite, a product management and growth marketing set under my team. But then things like creative and content marketing and brand marketing sat under the marketing team. But those teams were aligned to the same top level goals of the company. So, how we operated those teams had shared OKRs and they worked together as a cross functional team that blended all those different talents together to ultimately unlock some very tactical things that needed to get done. Performance marketing that has an effective period on the supply side, SEO content that needs to be written to rank well for different terms to attract creators, as a couple of examples, right. Like, emails that effectively drive activation, right?

That structure is scaled pretty nicely for Eventbrite that those teams you know, work pretty well together, and can be effective at achieving these company level goals together. I think where things go wrong, is when, say, one functions, goals, like the marketing team are just entirely different from, say, the product teams. Then you get into a situation where the marketing team is like begging product for resources that don’t actually map the product strategy looks like, and then they’re not really getting anything, and then they can’t hit their goals. And we’ve certainly seen that, I’m sure you’ve seen it at a few companies. So, we really try to make sure upfront in the planning process that at the strategic theme level, we have the different functions aligned to it. So, there’ll be a marketing leader, a product leader, an engineering leader, etc, present in helping build that strategic theme or initiative or whatever you want to call it. And then the process of aligning teams around it is naturally going to include people from their teams to make sure it actually happens.

Manfredi Sassoli:
So, yeah, this concept is super interesting. And so I was wondering if we could kind of deep dive into that in the alignment of the different teams, particularly because sometimes I’ve seen different departments have fairly different kind of tempo, right? So, for example, the brand strategy might be, like, multi-year, right? So, you have to kind of build that awareness, that reputation and that’s kind of a long process. Performance marketing or CRM, it might be kind of hyper cyclical and very short-term and you can kind of maybe optimize each month or each week, even around the same KPI, sometimes even each day, while maybe product will have kind of maybe bi weekly sprints, but also tend to maybe focus on projects, which tend to have like a quarterly cadence, for example. How do you kind of overlay those needs and in order to create a good collaboration?

Casey Winters:
It’s a great question. And I certainly agree with your framing in that different parts of the strategy have different timelines in which they’re executed or realized from a results perspective, right? You’re right, brand marketing tends to have the longest, right? But brand marketing isn’t a completely independent exercise from product or performance marketing. Performance marketing, yes, while it’s going to be optimizing more regularly, and will be looking for hopefully, in year returns, it still needs to reflect the brand or where the brand is going. Right?

So, I think a mistake some companies make, right, is and this can happen in both directions, is if you’re a performance marketing team, and you just optimize entirely based on the metrics without thought as to what the long-term brand strategy is, you might get the best results short-term, but you might not be unlocking the long-term benefit that you could be unlocking through infusing some brand marketing into that. And then the flip side is doing a bunch of brand marketing without measuring any return and saying, oh, well, this is like a three plus year investment into brand, we’re going to be like Apple and Nike. And then you could just burn millions of dollars and not be able to show anything for it. And in general, this is where CMOs get in trouble with CEOs is that they make that investment.

So, the way I prefer to think about it is all investments into paid marketing are measured on performance. So, you need to be able to get that return, whether it’s on a television ad, or a Google Advert, or an Instagram advertisement, you need to be able to measure that and have confidence that it’s generating a direct response return. And then on top of that, you want to infuse it with as much of your brand as possible to be laying this foundation for the brand and then impact a more emotional connection with the product can have long-term, but you want to be getting all of that for free. And I think that’s the way that startups need to think about brand versus performance. Obviously, when you get to Nike and Apple scale, you can do like a media mix model. And you can measure things over years to really understand the return that dedicated brand versus performance has, and you’re going to end up doing all of it, right? But that’s for companies at massive scale.

If you’re going to be investing in that more of the startup or scale up phase, then I think you generally need to be confident that your performance based return meets your payback period targets, but then don’t optimize entirely for it. Make sure you’re infusing some of that brand element in it and getting that long-term return, quote unquote, for free as part of the process. So, I think that’s the way that if you use those sorts of things. And then I think it’s relatively similar the way that that works with product, right? Which is, if you have a key understanding of what your brand guidelines are and your brand key, then everything we build in product has to reflect that. And if we’re learning something in product or inside performance marketing that is challenging that, then let’s debate that right now. Right?

So, I’ll give you an example. At Pinterest, our call to action was pin it. And that meant that you saved it to your board. So, pinning it to a pin board. That analogy did not work so well outside the United States. And we didn’t even translate pin it to the local version in whatever other language. So, as we’re doing to use the research, we saw that a lot of people weren’t saving things, because they didn’t know what that button meant. They thought it might just be the logo, and we wanted to change it to the local word for save. And we got a lot of pushback from the brand team saying like, oh, this is our version of tweet, right? Like, you can’t change this. This is a key brand moment for us. We made the change anyway, 15% improvement in activation rate.

So, we had to really take a step back and say like, well, what would you say the goal of a brand team is, right? And it’s like to communicate the promise of the product experience and make sure people easily comprehend how the product can help them in their lives. Clearly describing this in an obscure way, is preventing comprehension. So, like, let’s stop trying to be clever for a moment and just do something that helps our customers. We had a similar version where people coming from Google didn’t know what pins were yet. And we would say, hey, these are 15 pins related to men’s fashion. Where if we changed it to images, it would get more clicks, right. And the brand team didn’t like that as well, because we weren’t using the brand term pins.

But again, it’s like, if we’re not even given the chance to build a relationship with our customers due to a comprehension issue, then these brand terms are not doing us any good. Let’s like, use a much deeper relationship to develop the brand by getting them to use the product first. And then we can infuse a lot more into that relationship, right? So, you can have this perversion, right, of brand goals, if you’re not having that debate between the product team, the performance marketing team versus the brand team. But also the flip side is not healthy, right? Where you’re just like, oh, yeah, we’re just going to ship every ad and it says get $100 free if you sign up for this app, because then you’re not going to give them $100, and you’re going to break promises, and that might get you the best click through rate and conversion short term, but your business will go out of — you’ll go out of business long-term, right, because you’re breaking promises with customers.

Manfredi Sassoli:
Well, I think increasing the brand relation — improving the brand relationship by increasing product usage is excellent. I really liked that. And then after that kind of boosting the kind of brand infusion is, yeah, it absolutely makes sense.

Simone Cicero:
Definitely, it makes sense with the rest, I would say, right, this kind of stronger customer orientation. Also during the creation of the brand, essentially, Casey in the preparations, we had this discussion about Web3, and you told me I’m not following this, I’m not that interested. I’m curious to ask you now that you are in this particular moment, you left your role at Eventbrite, you’re looking into new things, so maybe you have already something coming up. How are you looking into the future of your work? What motivates you in terms of, back into a maybe a new role or maybe doing something different?

Casey Winters:
Yeah, I recently left Eventbrite, I’m still working with them three days a week as an advisor and helping them search for the next chief product officer. And then I’m working a couple days a week at another startup called Whatnot as an advisor. And it was over three years at Eventbrite, and then two years advising them before I joined full-time. So, I’m really trying to take a step back and think about what I want to do next. I’m definitely really passionate about discovering and scaling the best practices of building companies. So, that is something I will continue to do in various contexts, right? Whether it be operating, advising, investing, building programs at Reforge that teach, or you know, just writing blog posts that should happen at more frequency, and maybe I’ll be able to get back to that.

So, what I do will always be in some orientation around that mission, and perhaps shift between full-time operating versus advising, investing, writing, etc., right. So, that’s always going to be the land that I play in, and I’ll always be working with startups in some capacity around that. In particular about Web3, no, it’s still not an interest and not an area I’m looking at helping startups in. And I think we talked a little bit about it earlier, right, in that Web3 still feels like it’s not starting with the customer and working backwards to the technology like Steve Jobs advises. It’s starting with the technology And then trying to find a way to find something valuable for the customer through that technology. I think it’s actually a little bit deeper than that. I would say, right now Web3, to me, feels just like a sophisticated way to hide a lack of product market fit through financialization or a gambling element. That’s not something that excites me.

Simone Cicero:
What do you think about, let’s say, the political element? So, the kind of institutional innovation that may come from these shared databases, trustless databases, what do you think in terms of enabling maybe different business models? What is possible with these new things that maybe before it wasn’t?

Casey Winters:
I am hopeful that we are able to come up with use cases that we can’t currently enable with current technologies. I think I’ve become more and more bearish on us realizing that because let’s be honest, this has been a technology that’s been around for over 10 years. And we’re still searching for viable use cases 10 years in. And the web had email relatively quickly, it had search relatively quickly, it had forums, user groups relatively quickly. Like those things were immediately better than the real world alternatives almost immediately when they arrived. And in the same amount of time, with this new technology of the blockchain, we just haven’t seen that. And some of it is due to speed, so if the speed can get better, maybe when we enable some more financial alternatives that are interesting, I’m definitely open to that. But I just think it’s been a while and nothing’s unlocked. And that makes me increasingly concerned that nothing will unlock over time compared to alternative technologies you can use.

Simone Cicero:
That’s a great reflection to maybe end up. But before we believe you, I would really look into some of your breadcrumbs. So, if you want to share something with our audience, a couple of things that you believe are worth sharing, and that’s the only requirement. What would you want to share with our listeners?

Casey Winters:
Outside of work, I think some of you might be familiar, I’m a pretty big music guy. So, I used to be a music blogger for five years. So, I’ll leave you with an album recommendation.

Simone Cicero:
That’s great.

Casey Winters:
So, this artist Kuedo, he created a really seminal album that a decade ago, called Severant, which is a mixture of a bunch of like 80s, science fiction soundtrack type work with more modern underground electronic music styles, like footwork or some hip hop production and things that were going on. So, that album’s called Severant. It’s amazing and they just reissued it. But then Kuedo has also released a new album called Infinite Window, which is more of a modern take on Severant, and that is also amazing. So, if you want some electronic music that’s slightly askew from what you might normally hear, check out either of those Kuedo albums.

Simone Cicero:
Thank you so much. I mean, that’s great to have a music breadcrumb. It’s the first one, so you will be…

Casey Winters:
A Pioneer.

Simone Cicero:
Yes, you’re pioneering, definitely in this space. And we’re happy because — so the next guests will feel like I can also speak about music. That’s great. So, thank you so much, Casey. I mean, I have no words to say how much this conversation was interesting and to the point to the leaders that maybe listening to this episode and finding themselves into the challenges of building these organizations, or maybe designing them for future developments. And I think that was really great. So, thank you so much for coming. I hope you enjoyed it.

Casey Winters:
I did. Thanks.

Simone Cicero:
Manfredi, you too. So, thank you so much.

Manfredi Sassoli:
I had a great time.

Simone Cicero:
When it comes to our listeners, check the show notes of this show with all the links to the many things that Casey spoke about on the website, boundaryless.io/resources/podcast, you will see the podcast episode over there. And I mean, listen to me. As always, remember to think boundaryless.