Today on Next Gen Builders, Francois speaks with Edik Mitelman, SVP & General Manager of AppsFlyer Privacy Cloud, the trusted open platform for ecosystem collaboration and innovation with a clear mission: Enhancing end-users experience, while preserving their privacy.
Their conversation covers the challenges and strategies for building a startup within a larger organization, and why you would even do it in the first place. Edik highlights how a smaller, resource-limited team can align themselves much closer with their users, ensuring each line of code they produce serves a clear purpose.
Francois & Edik also touch on the role Product Managers should be serving within organizations and how they compare and contrast across various company sizes. One glaring similarity in Edik’s mind? The importance of maintaining a beginner’s mindset. In his words “ experience is important…but we as product managers must assume that we don't know and we need to test our hypotheses.”
Finally, Edik shares his thoughts around risk and failure. He stresses that more organizations should be comfortable with potential failure in order to foster an environment full of experimentation. Through each iteration, your team will gain learned experiences that will lay the groundwork for your next big success.
—
AppsFlyer Update
Since we recorded this episode, the journey Edik discusses of building what's essentially a startup within AppsFlyer is bearing fruit, with AppsFlyer's Data Collaboration Platform being used across several industries -- including quick commerce, consumer packaged goods, and financial services. The results for early adopters are impressive: marketing campaigns for brands using the product have seen up to 400% return on their ad spend and significant performance uplifts for their businesses. The partnerships embody what we talked about regarding the advantages of building a startup within an established company — leveraging AppsFlyer's resources, credibility and relationships while maintaining the innovative spirit and customer focus that drives true product-market fit. It's a real-world example of transforming concept into practice, process into product, and experience into measurable success.
—
Guest Bio
Edik Mitelman is the General Manager of Privacy Cloud at AppsFlyer, an ecosystem collaboration and innovation environment which preserves end-user privacy. Edik is an experienced executive with previous leadership roles at Autodesk, Wochit, and Conduit. He also runs the Venture Creation and Product Innovation program as part of the GMBA at Reichman University.
—
Guest Quote
”As product managers, we must do what customers need. So if you sold it and you don't care anymore, you will never care for that customer any longer, until maybe the renewal time comes up. I want my PMs to constantly worry about their customers, and to constantly deliver value throughout the year. So that people use our products more and more.” – Edik Mitelman
—
Time Stamps
00:00 Episode Start
01:45 When Edik left engineering for product
03:50 Maintaining a beginner's mindset
05:20 How the role of product varies between organizations
08:25 Building multiple startups within a single orga nization
10:15 Edik's work on AppsFlyer Privacy Cloud
13:45 The advantages that come from a startup mentality
17:05 Measuring success
20:55 Ensuring your team is rowing together
24:40 Failure breeds success
32:45 Are GMs for everyone?
—
Links
0:00:00.1 Francois Ajenstat: If you build a feature that nobody uses, did it really matter? Did you really ship anything?
0:00:04.8 Edik Mitelman: And especially in zero to one, when every day matters, every line of code matters. You do not have the privilege to build just because you think it's cool or because it's a cool new technology, or because the CEO sent you an email asking you to find a way to integrate AI. You can't. Every single thing you do, must be tied to a customer need.
[music]
0:00:38.3 Francois Ajenstat: This is Next Gen Builders. The show for the growth and product leaders of tomorrow. Building a startup within a larger organization, is that even possible? And if it is, how do you make it happen? Cutting through bureaucracy, fighting for resources, betting big on a long shot, having failure after failure that lead to success and growth. Well, our guest today thinks it's definitely possible, and he's already doing it. Edik Mitelman, is the SVP and General Manager of Privacy Cloud at AppsFlyer. Welcome to the podcast, Edik.
0:01:17.5 Edik Mitelman: Thank you, Francois. It's so great to be here. I'm a fan, I've been listening to many episodes and now is a small dream come true to be here. So I'm excited.
0:01:26.2 Francois Ajenstat: You're too kind. Well, thank you. Well Edik, would love to learn a little bit about your background. You've been in a lot of different companies. You've had a wide variety of experiences. How did you get to your current role? How'd you get into product?
0:01:42.5 Edik Mitelman: Yeah, so this is, I think it's a pretty regular story for product managers. I started as a software engineer, started it in high school. I was a geek, I'm still very much a geek. Started coding, liked it, was pretty good at it. Did my bachelor's in computer science. And then at some point, I decided that it's boring. Like frankly, no offense to my engineering friends. The best of them are my friends still. But for me, I lost the passion. And during my work, I was already a software engineering manager. I met a lot of customers, and I met a lot of users. And I realized that this is what I'm enjoying the most. Talking to people, talking to customers, understanding their needs, and then going back home and actually building it. And back then, it was way, way, way early days of product management, there was no even such thing. I didn't know that such a profession exists. Product management was something from the factory days, but when I was looking for a job that gives me that, people told me, "Oh, you should try PM. And I was like, cool. I don't know what it is, but I'll try.
0:02:54.1 Edik Mitelman: And I went all the way back, started as a junior product manager, left my cozy little engineering career, moved to PM and never looked back since then. Went all the way through different PM positions, had my own startup as a VP of product, corporate, startups, small companies. And then, I got the opportunity to be a GM for the first time in Autodesk. I led the Autodesk Israeli site, and I managed basically the entire operation in terms of engineering, UX, UI, PM, everything but sales, which was not managed by the technical people in Autodesk. And then my second GM role, is what I do here at AppsFlyer. So this is kind of the path from engineer, product management, general management that I do.
0:03:42.9 Francois Ajenstat: That's amazing. If you could go back in time when you did that transition from engineering to product, what would you tell your younger self?
0:03:51.8 Edik Mitelman: Well, I would tell my younger self the same thing I tell myself every single day. You know nothing. One of the basic flaws of young product managers I see, is that they believe that their prior experience or something really helps. And of course experience is important. I'm not here to say that experience doesn't matter, but we as product managers must assume that we don't know. And we need to test our hypotheses, we need to run tests, we need to talk to customers every single day. And when I started to become a PM, I came from an engineering world where you know your stuff. And here it's like, oh well, you don't know. And if you think you know, you might be very well wrong. So this is a tip I give to all junior PMs and senior PMs, myself included, every time we start to do something new, let's do research, because we are not the users, we are not the customers, and we don't know a lot of the things.
0:04:48.8 Francois Ajenstat: That's so true. I love the expression having a beginner's mind, which is you're always curious, doesn't matter what you knew, you have to keep being open to new ideas and new perspectives and always being curious.
0:05:02.4 Edik Mitelman: Absolutely.
0:05:03.6 Francois Ajenstat: Now, you've done startups, you've done large company, Autodesk is pretty large. And now you're in a fast growing company. In the role of product, what would have been the big differences across each of those types of companies? Did you approach it differently or is it wildly different between the sizes?
0:05:21.0 Edik Mitelman: Yeah, I think you must approach it differently, just because the dynamics of the organizations are completely different. So the craft itself of product management, I would argue is pretty much the same. I teach product management now in one of the universities here and people always tell me, what's the difference between B2B and B2C? And I always surprised them in saying, really frankly, between the two of us, not much. You need to do discovery, you need to have goals, you need to have OKRs, you need to have processes, you need to work with engineering, blah blah, blah. The craft of being a product manager, is the same.
0:06:00.3 Edik Mitelman: What changes, are the stakeholders you work with, the amount of stakeholders, the different organizations across you, that you need to really take into account. Because we as the PMs, we are at the center of this nervous system and we must make sure that everything plays along together. So being in a startup where everyone fits into one room and you just talk to the people every day, is completely different than trying to influence and impact without authority. Somebody on the other side of the ocean that doesn't know who you are, and you need something from them. Either it's marketing or finance or somebody else in another department. So this is where I think the PM role really changes, based on the size of the company and the various departments there are.
0:06:50.4 Francois Ajenstat: And so, do you think the communication mechanisms have to change the approvals? What actually changes in how the work of product changes?
0:07:00.9 Edik Mitelman: If you need to push an idea through in a startup, you don't even ask. It's better say sorry than ask for permission. In corporate, it doesn't work that way. You need to go through the chain of command, you need to go up and down to reach somebody. You need to cross the aisle so many times, and everything takes time. So also it was very hard for me when I first joined Autodesk from startup land, I was shocked. Things that used to take me a day, would take a week just because, for no good reason, somebody's on vacation, somebody, blah blah. Even these minor things. So the way you manage the product, the cycles you do, the different communication you need to do, the amount of people you need to bring to alignment, becomes a big chunk of the day to day, and less so the work with the engineering team of yours where it's pretty much the same every day.
0:08:01.6 Francois Ajenstat: Yep, that's so true. Now I think that right now you have a really interesting role in a really fascinating company, where you're essentially building a startup inside of a very fast growing company. Can you tell us a little bit of how you got into that role, and what exactly are you building?
0:08:23.1 Edik Mitelman: Yeah, that's a great question. So first of all, all the credit is due to the AppsFlyer culture of growth and innovation. The company, AppsFlyer is probably, many of the listeners here should know, is a market leader in the mobile attribution space and the measurement space in general. And it's a 12 year old company, so not a startup anymore. I'm not sure we're even a scale up. I don't know what category where the borders are, but the core product...
0:08:56.2 Francois Ajenstat: You're an app, you're an app. Everything's app.
0:09:00.1 Edik Mitelman: On the app yes. The core product is massive. It serves thousands of customers, it serves thousands of partners. The entire marketing, mobile marketing ecosystem is dependent on it. It's like a very important cog in all of this that makes the marketing world work, and there is competition and it's all established and it generates a lot of revenue. And this is the whale. But the company realized that you need to innovate, and you need to grow and you need to invest in adjacent markets or adjacent product offerings or adjacent verticals, in order to break the ceilings of the core product. Because no matter how big the product is, it always have a TAM. I worked previously for AutoCAD, that's a $3 billion ARR product. Maybe even since I left, it became more. So that's a huge number, but it's still a TAM. And there is a limit to where you can get. So if you don't become a multi-portfolio company or a multi-product company, you always have this limit. The hockey stick cannot grow forever. And the company realized that, and they established a structure, where we have a few GMs like myself, who are brought in to the company from the outside.
0:10:26.2 Edik Mitelman: This was also I think key and started ideating, started innovating, started thinking, and we now have different units that are building different products which are more or less of course adjacent to the core product. Of course it must, must be correlated to the company vision and mission. This is one of my biggest tips, like if the company produces cars, do not start selling apples from this company. It will not work. It needs to be aligned with the vision, with the mission, with the strategy. But, it's a zero to one effort. And we have many of these. Some of these fail, as the nature of every startup or every idea. But if you hit on one or two of these or three of these for the long term, this is how you scale a huge company to be even bigger and even bigger. And we see examples of it, not just AppsFlyer, look at the Facebooks of the world, and the Googles of the world, you have to. So that's what AppsFlyer did. And I'm lucky to participate in this, and I'm actually running an organization that we call the Privacy Cloud. And it was created three plus years ago when the privacy shift in the marketing land became huge.
0:11:46.8 Edik Mitelman: Of course, Apple ATT changes on iOS were one of the catalyst, but not just that, GDPR before the CCPA in California, many different regulation changes, platform changes, created this mind shift of, well, the way we did performance marketing or marketing in general for the past 20 years online, probably doesn't cut it anymore. And this is where we started experimenting. We started looking at things. We now have a data cleanroom product, which when I started, data clean rooms and PETs, PETs, were something from a science class of your kids and cats and dogs. These days, everybody in the industry knows that PETs are privacy enhancing technologies. Everybody knows what a clean room is, even to the extent that it's even confusing how many definitions there are. But the industry went through a rapid change, rapid adoption of the things like, oh, everybody now knows what a data collaboration platform is, what a data cleanroom is. And really we are building a technology and a product on top of it, that enables any two parties, usually it will be an advertiser and the network, but any two parties to collaborate on their first party data, in a private and secure way without exposing anything to anyone that they didn't know before, and use the result of this collaboration to target, to activate, to measure and to really complete the entire workflow from the audience creation and segmentation, all the way toward measurement insights, rinse, repeat.
0:13:27.3 Edik Mitelman: And this is the data collaboration platform. This is the organization I'm running. And this industry in general, is in zero to one phase. And this is exciting and I love every single day of it.
0:13:38.4 Francois Ajenstat: You can tell from the passion that you're bringing to all this. Boy, this is fascinating. I have a million questions to come out of it. So I guess first is, how did you guys decide between incubating it as a startup within the company, versus running it from the core product? When do you split it out, versus keep it in the main?
0:14:01.6 Edik Mitelman: So I think, and this is a culture that I believe in and I'm lucky to be in a company that does that. It is very hard to do something from the core. The core by design, by essence, is slower. The core thinks about every single dollar of investment, with a predictability of ROI, and you're really moving in a specific direction, customer obsessed and improving the product and the value. To do an incubation of something new, and one might argue crazy, or more or less requires agility, boldness, you become a cost center. So if my CEO Oren, will listen to this, he will agree and he always tells me, "I am still a cost center for him." I'm not a revenue center. So there needs to be trust and there needs to be a belief that out of these few opportunities, something big will grow. And doing it from within the core, is very hard. You get bogged down with support tickets and customers and all of these things, which is very important to grow what pays all of our salaries. But you really kind of almost need to force this separation, for these incubations to run as a startup, to have their own accountability, responsibility and freedom in many cases.
0:15:33.8 Edik Mitelman: So I for instance, have my own goal to marketing. It's very small, but it's dedicated only to our data collaboration platform. So if I go to the big machine, the salespeople there, can sell seven different products and they carry huge quotas. That's what you get when you work for a multi-hundred million dollar ARR Enterprise B2B company. I cannot do that. I cannot guarantee them to meet their quotas. So why would they sell me? They wouldn't. And I understand them. I like the metaphor of comparing salespeople to slot machines in Vegas. A lot of bells and whistles, but before you put the first dollar in, you cannot play. So now with me having my own dedicated team, this is where I can run, I can do discounts, I can experiment with business model, I can be very creative. And I'm not competing on the time or the resources or the thoughts really, the share of mind of my colleagues. And this is very important.
0:16:41.5 Francois Ajenstat: Fascinating. Now, how do you think about the horizons for an ROI? Clearly, you're incubating something right now, so the success is not guaranteed, nor is it immediate. What's the window where you're saying, this was a good bet, we should keep on investing in it, versus saying we tried, we learned, we failed, we'll move on?
0:17:05.8 Edik Mitelman: Yeah, that's an amazing question. I'm still not sure I have the full answer to it, because I'm in the middle of it. But the way I personally look at it, is that of course, ARR-wise there is this classic zero to one, one to 10, 10 to 100, and then you're off to the races. But this is not enough. You also need to have a time horizon. But what is most important in my mind, is to find product market fit. Now, product market fit is this vague, magical term that every PM out there is looking for, and nobody knows how to define it's magic. But you really need to have some metrics, both early indicators and lagging indicators, some KPIs, you have to adjust it to your business of, okay, this correlates to product market fit. And then, if you find that with your early adopters, if you find that with the first few, then you can safely, well, to the extent safety is in startups, but you can come to the board, you can come to the management and say, "Look guys, there is oil here, here why?" Not hand waving, not I believe, I have a dream.
0:18:15.6 Edik Mitelman: No, that's not enough. That's enough for the first stage, not for the second one. You need to show product market fit. Whether it's first few deals, whether it's some customers that have design partnership with you and they have quotes, something tangible that shows, yes, people need it, now we just need to work very, very hard to deliver, to beat competitors, to be better, to grow. But product market fit is the very critical stage at which point I would personally decide, it doesn't work or it works, now, let's scale it.
0:18:48.6 Francois Ajenstat: And think about that product market fit. In some cases, you're B2B, and so the size of the deals, it's not necessarily about number of customers, it might be size of customers. How do you think about usage? And are there specific metrics that you look for, that tells you you're shooting in the right direction?
0:19:07.6 Edik Mitelman: Yeah. Usage, I think should be the basis even of the business model and definitely the product model. Because I personally think that long gone are the days where you just sell licenses, and then you don't care. And the reason for that, is not even the ARR of it, the SaaS model of it, is that then you're not customer obsessed.
0:19:31.6 Edik Mitelman: And we as product managers, we must be, we must do what the customers need. So if you sold it and you don't care anymore, you will never care for that customer any longer until maybe the renewal time comes up. I want my PMs to constantly worry about their customers, and to constantly deliver value throughout the year, so the people use our products more and more. And then of course, let's not be cute, stickiness creates renewals, creates NRR, good for business. But it starts from product value. People need to use the product more and more and more. Either it's more seats in the account, more accounts, it doesn't matter how you get it. It also depends on the product. But you must see a strong graph that shows you, that usage is not plateaued. It's not like somebody came in once a month, clicked the button maybe, and they pay for something they forgot they have. No, that's a bad model.
0:20:30.0 Francois Ajenstat: I love everything you just said. And do you have any techniques for how you infuse that mindset into your teams? 'Cause it is different than maybe in a scaled company, how do you create that perspective that every team member, whether they're an product, engineering, sales, go to market, whatever, have that same view of the world?
0:20:56.2 Edik Mitelman: It is very hard, frankly. And I tried it numerous times on numerous teams and I failed a lot of times, because it takes a mindset change. That's the key. Note that I am a big fan of OKRs, but OKRs are a tool. They are not the goal. They're not the golden solution, silver bullet for everything. People need to adopt the thinking, that we're all doing business here. Engineers need to understand customers, marketers, designers, sellers, everybody, the cleaner in the office. Literally every single person on my team, has to meet customers. This is like an obligation. An excuse of I am just a backend engineer, doesn't cut it. These people are great engineers, I love them all. But they're not suitable for that product stage. For the core business, where there are hundreds of engineers and you might do some, that might be fine. You cannot put everybody in front of people. But for a few tens or maybe up to 100 people company, it must be just a basic rule. And then, you install the OKRs. Every OKR, talks about customers and talks about business outcomes. No outputs, no, we will deliver that. Who cares? Somebody will use that, is what cares, is what matters.
0:22:27.9 Edik Mitelman: And then you see engineers, running after the PMs and like, "Hey, did somebody use my feature?" And this is the culture I'm trying to build. Because they shouldn't care about, we shipped it on time, there are no bugs. Yeah, I bonus. No, if nobody used it, nobody gets bonuses.
[laughter]
0:22:47.3 Francois Ajenstat: It's a great expression. You also have this view of, if you hear a tree or if a tree falls in the forest...
0:22:56.2 Edik Mitelman: And nobody hears about it, nobody cares. Yeah.
0:22:57.9 Francois Ajenstat: Yeah, exactly. It's the same thing. If you build a feature that nobody uses, did it really matter? Did you really ship anything?
0:23:03.8 Edik Mitelman: And especially in zero to one, when every day matters, every line of code matters. You cannot, you do not have the privilege to build just because you think it's cool, or because it's a cool new technology, or because the CEO sent you an email, and asking you to find a way to integrate AI. You can't. Every single thing you do, must be tied to a customer need. A customer's explicitly asking the sales team that, or something that you actually know people are waiting. That's the only way to break through the product market fit glass ceiling. It's the only way to really be efficient and effective, especially if you're a startup fighting an incumbent, which might be much bigger than you. So if you will be complacent and will work regularly, they will beat you, because they have more people. You work at the same pace. Math is easy. So you must be much faster, much more efficient, and much more focused only on the most important things to your customers. And that's how you can beat incumbents, who tend to be a little bit more complacent.
0:24:14.0 Francois Ajenstat: Now, one of the things I hear in this, which I've heard objections from others, is well, if everything has to have an OKR, and I'm rewarded based on usage, does that mean there's a disincentive to make big bold bets or try things that might fail? How do you balance off the trying, experimenting with this drive for results?
0:24:39.7 Edik Mitelman: So, I'll tell you a true story. I installed OKRs in my organization here at AppsFlyer, three years ago. And we are of course, part of the large planning meetings of the entire company. The entire product and R&D organization comes in, all the stakeholders come in, and we talk about quarterly plans. My team, was the only one at the beginning, who had failed, failed, failed, failed. Everybody else, and again, I'm not blaming them. That's how the core business works. It's like green, green, green, green. Well, maybe here are some red, here we were delayed, here, but we were failing. And that is the core of a good progress in my mind. If you are successful at everything you did, oh, you did something terribly wrong, or you're just underachieving.
0:25:33.2 Edik Mitelman: So at first, it was very hard for my people because they felt weird. You come into this planning with the entire management team, everybody, and looking at you like, oh, these are the failures. But actually now, three years went by and now I think it's part of the overall culture, that failing is amazing, as long as you learned something from it and now you know what to do better.
0:25:58.6 Edik Mitelman: Failure is unacceptable. If like, what are we doing next? I don't know, that's unacceptable. But if we failed, because we tried something and we say, "Okay, we learned A, B and C, now our plan next quarter is to do T, Z, Y." Because of that, great. That's how you discover things, that's how you innovate. If you know ahead of time, that you will succeed, that's not a success in my book. That's extremely easy. That's like putting your money under your couch and saying, "I know how much money is here." That's not really the way to grow your assets. We all put money, based on our risk appetite. Some will put it in S&P, and some would put it on one weird stock. That's risk appetite. But you need to risk something to get a reward. Otherwise, if you're always successful, always in the green, you're not taking risks.
0:26:55.1 Francois Ajenstat: So, you're really celebrating a learning culture. Failure is success, if learning comes out of it?
0:27:04.3 Edik Mitelman: I want to take this quote of you and put it on the loop in my team. And again, to tell you that everybody agrees with that, and it's so easy, no. From my experience, it takes between a year and a year and a half, for a new team to get into this mindset. And some people will leave. Also, managers need to take this into account, if you really want to be serious about it, you need to accept attrition of talented people. Not bad. Like not the bottom 10. That's nobody cares about. But talented people will leave you, because they are operators, and they want to be in a corporate environment where successes are celebrated, and not failures and everything else. And that's okay. We as managers must be okay with super duper talented people, no matter the profession, who leave us because the culture is not for them. That requires managers to also change or say to, "I know, but he's such a good engineer." Yes, but if he doesn't care about business, if all he wants is to write cool code and don't touch me, he is not the best fit for this organization and we need to let him go.
0:28:14.0 Francois Ajenstat: I know it's a hard thing, but it's also a very important thing to instill in orgs. Can you make it concrete and maybe give an example, where you guys tried a thing and it just failed. And as a result of that failure, it actually drove a success afterwards?
0:28:31.4 Edik Mitelman: Yes, we tried for a long period of time. In startup days, it was very long to connect what we do to the core business in a more complete way. And I was banging my head, I was trying to get to the same ICP, to the same Persona. And just to say, this is kind of more of AppsFlyer for you guys. And that made sense for everybody. Nobody kind of challenged it. It was natural. It was like, "Yeah, we have this and now we're just expanding the borders of this." And then I just said, okay, it doesn't work. There's not enough compelling event here, it's an upsell. What is it exactly? It's a tough thing for the salesperson to pitch. It was weird and it didn't really work. And I spent a lot, a few good months on trying to make it work. But the result of it, was that, I had to look around. And I wasn't forced to look around beforehand, but I now had to. And suddenly, I realized where the puck will be, and the puck will be in the retail and commerce media networks land, which was a completely new term within AppsFlyer and outside the world. And I said, this looks interesting.
0:30:00.5 Edik Mitelman: Actually, I was forced by the failure to go and look and find these new things. And that's how I realized RMNs are the future. I bet on it. We started building towards RMNs, and now this is a $160 billion business. Now you have the Chase banks and the PayPals of the world and the United Airlines and everybody and their uncle, becomes a commerce media network. And we have a solution for it, and we are in this game. And it all happened because I failed to build an extension of AppsFlyer, and I was forced to take three, four, five more steps away from the comfort zone.
0:30:39.7 Francois Ajenstat: That's fascinating. The thing I always find interesting in here, is there's learning in terms of being able to explore spaces and look at an adjacency and maybe there was nothing there. Sometimes when you're learning and you're creating a new pattern, you're discovering the pattern, maybe you got it wrong, but it's reusable, 'cause now it's actually accelerating the next thing you wanna go do. And I think it also, there's something interesting in there as well in terms of when you build a new business, what you heard at the beginning, versus what actually happens in practice, may be a little bit different. And so, you'd learn to ask better questions.
[laughter]
0:31:21.4 Edik Mitelman: Yeah. I think the culture and the having a team, that went through this process a couple of times, makes the time n+1, 3X easier, 5X easier. So as long as the team is okay with, "We might fail." And that's great. Nobody will get fired or whatever. We are free to experiment, we are free to go nuts. We can bring up crazy ideas as long as we keep learning and iterating, code can be deleted. There's no such thing as sunk cost biases. Then this team, will find it. Because they get better at it. For the first time, it's very hard. For the second time, it's easier, for the fifth time, it becomes part of what we do. So hopefully, you don't need to do five pivots. Hopefully, you are kind of your instincts and the data is better. But two, three is normal. And then you find it, but the team is already in place. They know how to tackle a problem, they jump on it, and they tear it to pieces. And you do this first jump much faster than to start over and over again.
0:32:33.6 Francois Ajenstat: Yeah. Creating the habit, and you have those learned experiences and you're creating process. You're laying the road down as you're driving on it. I have something I wanna come back to. You're a GM in a large business. Not every organization follows a GM model. Now clearly, it seems to be working for AppsFlyer to create different GMs. But I guess my question is, to GM or not to GM, is GM a good thing, is it the mindset that everybody should have or are there pitfalls, recommendations to think about before embarking on that journey?
0:33:13.7 Edik Mitelman: I wouldn't be as arrogant to claim that I have an idea. This is very, very situational, even for the same company. There are cons and pros to independence, because at some point if you're too siloed, you are not maximizing the full potential of the company. Every GM, is striving to achieve their local maximum, to use a math term, sorry for not the math bachelor's here. But then, you're not sure that the company reaches the global maximum. And we're not talking enough, and I'm saying it as a fact, we the GMs at AppsFlyer, we are not talking enough. Everyone is so busy driving their own thing, which is important, that we sometimes lose the commonalities. And we sometimes could have done something better together. On the other hand, if everything is kind of top-down and the more classic structure, VP product, VP R&D, VP marketing, everything is stifled, very hard to innovate. You need to find bottom-up kind of little corners of people who decided to do it overnight in a hackathon. So there's no good or bad, and there is no, you have to do this and never do this. You might start with a GM model and you create a very clear criteria of what is incubation, graduation.
0:34:43.3 Edik Mitelman: And once it graduates and reaches these KPIs, it moves to the core business. And the core business, by the way, at AppsFlyer, is managed by the VP structure. The core doesn't have a GM, because most of the people work on it. So this is structured in the classical way of doing things, the incubations have the GMs. So it's a mixed bag. It requires a lot of management, it requires a lot of communication between the different parts. It's not easy, it's not a silver bullet, and in some cases, it doesn't work, and it's fine. But I would recommend companies who are thinking to become multi-portfolio or multi-product, try it out. It might not fit your culture, it might not fit your people, but give it a try. And if it doesn't work, to consolidate is always much easier than to break up. So give it a try, and if you see that it doesn't go anywhere, consolidate again. The same VPs get more headcount, everybody is happy and try to work your way out of it with one kind of unit.
0:35:55.0 Francois Ajenstat: Again, I love everything you say, 'cause you still bring a learning mindset to everything. To building, to organizational structure, to business. It's so critical.
0:36:04.2 Edik Mitelman: Yep.
0:36:05.4 Francois Ajenstat: My view of it, is to have a mindset, you don't necessarily have to have the structure. So what does a GM do? They think about the business, they think about the customers, they think about the right product and the resource allocation. It's the same mindset you have to have. You can't, not think about customers, not think about all the elements that go into it. Even if you're in a large or small. The mindset is important, the structure I think is secondary. It's optimizing for the best way of doing the work.
0:36:35.1 Edik Mitelman: I couldn't agree more. I think it's the ideal situation to work in an organization, that doesn't lose that with scale. I think this structure, is just an enforcement of that thinking. If that doesn't work, kind of let's call it organically, you put a GM, and thus you just enforce that type of thinking because that person is 100% measured on this. So this is also maybe a good way of, if it doesn't work from the people in the organization, put somebody whose entire job is to do that, he or she will try to propagate that thinking in a small group of people, and then maybe you don't need a GM anymore, because you now have a great bunch of executives or leaders who are wired that way. Plug them back in to be your change agents in the corporate, and it might happen. So that's also a great way of how to take this thing, cultivate it small, and then use it as a trigger to change the bigger company.
0:37:45.2 Francois Ajenstat: Edik, this was such an amazing conversation. I think we can go on for hours and hours. And I'm glad that you're teaching this also in schools and building the next set of leaders. Really excited to see what you're doing at AppsFlyer and how you're going to create even more success for the company.
0:38:05.3 Edik Mitelman: Yeah, thank you so much for having me. It was a great conversation and really, we just touched the tip of the iceberg here, so I hope that we can continue talking about that. And again, I encourage everyone to join the conversation. I'm passionate about these things. I love to learn from anything and anyone, because again, I always believe I am Jon Snow, I know nothing. So please talk to me, reach out and let's have these discussions. Whether it's about how the data collaboration platform of AppsFlyer can help your business, or product management, best skills, whatever.
0:38:44.4 Francois Ajenstat: And I'll say it again, and we've said it a number of times, success equals learning.
0:38:50.6 Edik Mitelman: Absolutely.
0:38:51.8 Francois Ajenstat: That is key. Well, Edik, thank you so much for joining and thank you all for listening to Next Gen Builders, and look out for your next episode wherever you get your podcasts. And please, don't forget to subscribe.
[music]