The Production-First Mindset

DraftKings's Idan Fridman - Survival Instinct In Startups

January 02, 2022 Liran Haimovitch Episode 23
The Production-First Mindset
DraftKings's Idan Fridman - Survival Instinct In Startups
Show Notes Transcript

Rookout CTO Liran Haimovitch sits down with Idan Fridman, Director of Engineering at DraftKings.  They discuss his experience having his company be acquired, working with product stakeholders and finding common ground, building the culture at Blue Ribbon vs at DraftKings, and building the first satellite team.

Painless Cloud-Native Debugging
Rookout is a disruptive developer solution for Cloud-Native debugging and live data collection.

SPEAKERS
Liran Haimovitch, Idan Fridman

Liran Haimovitch  00:02
Welcome to The Production-First Mindset. A podcast where we discuss the world of building code, from the lab all the way to production. We explore the tactics, methodologies, and metrics used to drive real customer value by the engineering leaders actually doing it. I'm your host, Liran Haimovitch, CTO and Co-Founder of Rookout. I'm joined today by our guest, Idan Fridman. Idan is the Director of Engineering in DraftKings, one of the most exciting virtual sports companies out there. Thank you for joining us, and welcome to the show.

Idan Fridman  00:41
Hi, Liran. Thank you. I'm happy to be here.

Liran Haimovitch  00:44
Idan, can you tell us a little bit about yourself and your role at DraftKings?

Idan Fridman  00:48
Yeah. So I myself have about like 15 years in the industry, then different roles. I started as a hardcore developer. Moving forward, I've done more technical roles like architect and after that I moved to leadership, being a team leader. Small companies, and the big ones. In the last six, seven years I'm doing specialty startups. And well, the last one we made it get acquired by DraftKings. It's very exciting for us to be part of that.

Liran Haimovitch  01:24
So you ended up in DraftKings after they acquired your previous startups?

Idan Fridman  01:29
Yeah.

Liran Haimovitch  01:29
Can you share with us a little bit about what was that startup all about? What were you doing?

Idan Fridman  01:34
Yeah, so Blue Ribbon, as I call it, is a  startup where we are building gaming tools for different operators in the i-gaming industry. What we essentially do is our tools -once you integrate with us - they bring you increased value in the retention and engagement of your players. The first tool that we have our focus on is jackpots. Jackpot is thinking about like a progressive lottery pot where people contribute and they can win different prizes on different amounts. So essentially, you can take it and put it on anything, you can put it and connected to an actual game, you can put it and connect it into a virtual game. You can even connect it to supermarket if you want. Once you send some triggers. You can choose which amount you want to contribute and then there is like a big lottery but there is a winner who can win it all.

Liran Haimovitch  02:30
And I'm guessing you work with a lot of gaming companies as you're building this jackpot product.

Idan Fridman  02:36
Yeah, we're working mostly with gaming companies. But what's really interesting here is that we were acquired by DraftKings. They have different verticals. One of the major verticals of DraftKings is DFS, which is a Daily Fantasy Sports where you can compete against other players, virtually, and with leaderboards. You can rank different players on real games, and the scores that you get are based on the actual results on the real games. Once you can connect the jackpots to this party, then you can win different prizes based on leaderboards, and other metrics that you can pre-configure.

Liran Haimovitch  03:14
And now Blue Ribbon is a part of DraftKings. What is it like?

Idan Fridman  03:19
So this is very interesting, because now that we formed a company, we're about 20 people. 15-20 people now is part of 3,000 people. That's a different challenge, the scale, the impact, you can't really compare between them in regards to the excitement and the adventures that we've been through. So the challenges are pretty much different. And now we're speaking about, as I said, scale, and every change that you do, every new feature that you add can create lots of impacts among the DraftKings players. Yeah.

Liran Haimovitch  03:51
I've been working a lot with gaming companies or clientele. And I've noticed that many of them, I think all of them have multinational engineering teams, which is something you don't see across all segments. So I'm kind of wondering what's your take on that? How is it like in Draftkings?

Idan Fridman  04:07
Yeah, so multinational teams, I think today it's even less a challenge as it used to be because the COVID and everybody works from home, so once you work remotely it doesn't really matter if you're physically close to each other or to one another in that manner. In regards to culture and other differences, then once you understand the other teams and their culture and the way that they think and you're taking it under consideration, and their holidays, their mentality. Once you all get to understand each other and communicate with each other, then I think it's getting much more effective. And just, you know, having some routines that you think you can apply for your remote teams and then to your local teams.

Liran Haimovitch  04:54
Which locations do you currently work with? I mean, where does DraftKings have engineering teams that you work with on a daily basis?

Idan Fridman  05:01
So the daily, actually the daily basis because the Blue Ribbon team has been converted to DraftKings team. So most of them locally but we have teams in Boston which are in the United States, North America, and also in Eastern Europe, Ukraine, Kyiv, and in Bulgaria. And the time zones are challenging. But still, when you know to work on synchronously that makes things easier for us. Yeah.

Liran Haimovitch  05:24
And today you provide services to those other teams?

Idan Fridman  05:27
Yes. Today, as I said, Blue Ribbon is now a so-called Gamification platform team. What we do since we bring solutions to all verticals in DraftKings, so we give those services to different teams across the globe for DraftKings. Yes.

Liran Haimovitch  05:42
So what kind of services, what kind of gamification services do you offer?

Idan Fridman  05:45
So now, besides the jackpot solution, we are extending ourselves to more gamification tools, where there are event-driven triggers or we're also expanding our gamification, I would say product line to work with the leaderboards. Now, we are also thinking about bringing social elements in. This is still ongoing. But mostly what we've done the last year is to engage our current platform, working with different types of jackpot configurations, to other verticals.

Liran Haimovitch  06:17
And how does it work? I mean, in your case, you're not only integrating your team into the company,  you're also trying to integrate your service into every different product that DraftKing's offering. So what's that process Like?

Idan Fridman  06:30
Yeah, so this is very interesting, because now since you have different products that you want and need to integrate with, that brings a lot of variety, in your challenges, on day-to-day challenges. All of a sudden, the design is changed, your metrics are being changed. We are very metrics-driven, and we want to measure ourselves. So all of a sudden, the metrics that we use to measure ourselves when we were just blue ribbon are now getting different. Okay. I think it is very interesting to work with different product stakeholders and to find some common ground for all of them. So we can develop one thing or maybe multiple things, but still to have some common base where we can scale and extend ourselves.

Liran Haimovitch  07:13
Can you give a few examples of the metrics you've been tracking as Blue Ribbon, versus the metrics you're tracking as a gamification platform?

Idan Fridman  07:20
Yeah, so one very famous metric is engagement. Okay, we want to know once you put our tool and engage them into the game, how many new players for example, join the game since using our tool, which is very, very important. For example, in DraftKings, then you want to look at retention, because there are a lot of players and you want to see how those players are now behaving once they have games with Blue Ribbon tools. So those are user retention as well. But now it's very, very important for us to see how the existed players.

Liran Haimovitch  07:53
And in the past, which metrics were you tracking as blue ribbon?

Idan Fridman  07:57
So engagement was one of them, and DAU which is like the daily active users. How many users are, let's say in everyday are actually active, using our games that uses blue ribbon tools.

Liran Haimovitch  08:08
Now, I know you're a big believer in culture, and how do you build engineering teams to deliver the best possible software? By creating the right kind of culture? What was it like to build the culture at Blue Ribbon? And how is it different today at DraftKings?

Idan Fridman  08:24
This is a great question. First of all, I think the fundamental of every culture is communication. I think this is the basic and most mandatory thing that you need to be established in development teams because once there is a communication you gain other things like transparency for example. In practice, transparency can come in place where you're not going to meet the deadline, or your team are not going to meet a deadline or when you know I see a new startup and you're going to be out of budget so you want to communicate this and be transparent with your team as well. So once we have this communication, then everybody is aligned and feels safe to do things, to dare to do things and break things. That brings really great motivation to do great and big, I would say take risks. Okay, so for me to build this culture, as I said best is on the communication and afterward is the learning curve, which is super important. We want to focus on creating new features and creating new things but still maintain the current system in a way that it won't block you in the way. In a startup you need to move fast. So the culture is a little bit different. Not because you cannot move fast in a big company don't get me wrong. In a small company, it's just that you have a more survival aspect, the survival aspect is very, very huge, like, if you do a mistake, you know you might not be there anymore, right? So these survival instincts put you in spots where you want to push hard and make sure that you deliver value and show your investments and everything else. In DraftKings which is a big company like any other, the survival instinct is, as I said, it's getting weaker. On one hand, but on the other hand, you have a huge impact. All of a sudden, you have millions of players and users, you don't want to do mistakes over there as well. You want to make sure that you are safe and you do the things the right way, not to lose them. So as I said, you need to move fast on both ways in regards to culture, and to dare things but you pay attention to different things on both angles.

Liran Haimovitch  10:27
Now, you mentioned moving fast, and I know that moving fast--  Personally for me, the most important thing is feedback, being able to know if you're doing the right thing, or if you're making a mistake, and kind of stopping it before it gets too bad. So, how do you see feedback in the software engineering lifecycle?

Idan Fridman  10:45
Feedback, I think it's one of very crucial end-step in every dev cycle, life cycle. The sooner you get the feedback, the better you're going to get. This is our seat. And let's take an actual example. Right? You're taking a branch as a developer, you want to push your code and you want to make sure that what you've done is right. Once you release, you want to have the feedback as soon as possible. So you know if you were right or not, and that's right for startup, and that's right also for a big company. You want to get that feedback from users, whether it's another department in the organization, or whether it's a player. Of course, there are techniques how to get it, right? We have different technologies to apply that, whether it's a continuous delivery, or AB test, and etc. But as I said, feedback is the first milestone for you to understand if you're going the right way.

Liran Haimovitch  11:35
I'm sure you like many others agree that metrics is the easiest way to get feedback. You've already mentioned metrics. And metrics is usually the easiest way to know if you're heading in the right direction, or in the wrong direction. What kind of metrics are you using at DraftKings, at the gamification platform to know for every release, for every change? Is it improving the system? Or is it making things worse?

Idan Fridman  11:59
So, to that question, what's really nice that we can leverage in DraftKings or a big company compared to a startup is that you already have some workload, you already have some, I would say, action, right? You have players, their playing, you have traffic. It's very easy to compare between one state of the traffic to another, when you're putting a new tool of blue ribbon for example, gamification, or if you change anything. So back to the actual metrics. As I said, the retention and engagement are very, very important. Another metric would be in session time, how long a player is actually playing during a game that was deployed with a gamification tool or without it? session times is very important. Yeah. And also, in the end of the day, the wager itself. What would be the amount of the payments that have been paid on different product lines with gamification and without?

Liran Haimovitch  13:00
And what impact do you see on games at the gamification platform at DraftKings?

Idan Fridman  13:05
So our pretty new, we acquired about seven months ago, we do see some metrics, I am not able to expose them right now. But it's looking very, very good. We weren't very surprised because we know the impact that we had on different customers that we had. It's very important to say that on Blue Ribbon, when we were in startup, our focus was on b to b. And now our focus is in b to c. The end user for us in the end of the day is the player, which is a little bit different. So we are polishing our KPIs according to that as well. On the go.

Liran Haimovitch  13:39
Now without going into specific numbers, how is the scale changed? From the Blue Ribbon platform to now, the DraftKings gamification platform?

Idan Fridman  13:48
Yeah, so in regards to, let's take transactions per second, I would say that business transaction per second in Blue Ribbon would be… for a couple of 100. So all of a sudden, it's couple of 10,000. It's huge, not from day one. We are getting there slowly. But once we understand and polish where we want to put the gamification tools all over the verticals of draftkings, then we know which traffic and how much of it we want to open. So currently, right now, the impact is huge, but we're still careful, and understand on where and when we want to open those funnels.

Liran Haimovitch  14:22
Now I know many engineering leaders are worried about cloud costs, whenever they scale, whenever they start seeing more and more transactions. I mean, if you get to 100 export transactions, are you going to be paying Amazon 100x? And can you afford that? What's your view on that?

Idan Fridman  14:36
Yeah, so I think to move fast, you need to have good skills in the cloud, and of course, everything around that. I don't want to-- I mean, everybody knows that right? I don't want to detail a lot of it. But regarding to cost, I think you want to make sure and pay attention that everything that is in control. You can do it by putting alerts, you can make sure that you measure it on a monthly basis. What was your accounting compared to the previous one. I think that in cloud, one of the easier things, but also can be a pain that in a matter of clicks, you can create machines, you can create environments. But once you scale, if you didn't design it right and you didn't pay attention, it can get a huge budget. But again, if the business is correlated with the cloud costs, that's fine, right? I mean, we have more traffic, we need to pay more resources, and that's okay. But if it's getting out of control, then that can be a huge nightmare, because then you need to pay more resources to fix the cloud architecture that you've built before.

Liran Haimovitch  15:35
But sometimes, even if you do make more money, and I mean, you're still profitable, then you’re all of a sudden saying this, “I'm paying, I don't know, a million dollars a year, I can save those million dollars a year.” And the first often becomes a second priority later on.

Idan Fridman  15:48
You know what they say, you need to spend some to get something else. There is no way how you can actually progress. But eventually, you need to put yourself ceremonies for cost budget control. From day one, we were aware of this in Blue Ribbon from day one, again, because in every startup, you know, moneys is God, and you want to make sure you have a cash flow. And yes, cloud modification and services and everything, all the third party tools that we're using, they all were payment consumers. We want to make sure they're all in control. So we just created different ceremonies on a monthly basis or quarterly basis to make sure that whatever we do, we actually use it. You know, it's not a surprise that you have some machines, stop somewhere or not stop somewhere, nobody used them. And they're just there because it's easy for them to be there. But once your work through it, then things get a bit more in control.

Liran Haimovitch  16:47
So what kind of ceremonies did you put into place at Blue Ribbon to get things right from the get-go?

Idan Fridman  16:52
From the cost perspective? So since we were based on AWS, for us, it's very easy to use CloudHealth, for example, which is a tool that reflects to you exactly how much do you pay for different perspective in your Cloud account? For example, EC2 is the machines. How much of them are on-demand or not on-demand? How much do you pay for F3, how much you pay for your queue providing AWS, SQS, SNS, and etc. So what you essentially start to do is putting alerts. And when you put alerts on thresholds, then that makes you more aware, hey, I just got an alert that I crossed my, I don't know 5000 limit, for example. You get the alert, and then you pay attention to it. But you need to understand first what is available and what you can do, whether it's using Spot Instances or reserved, and you need to understand what the services are? And then just put metrics like anything else.

Liran Haimovitch  17:46
Makes sense. I would love to ask you another question that I asked all of my guests. What's the single bug that you remember the most?

Idan Fridman  17:53
The single bug? Oh, I think if you speak about cost-effective, I think one of the-- I don't know if it's exactly a technical bug, but we had a huge payment debt for Amazon because of the SQS. And what we didn't pay attention to is that when you use a technology like that, you need to make sure that what you pay is clear. And we found ourselves-- we had an issue that we triggered the API of AWS, a lot of it because we wanted to optimize the latency issues, and we kept on triggering these APIs. And all of a sudden we got a huge debt on Amazon. We found out that Amazon also collects things using API triggers and not just the service itself. And we found a huge bug over there, where we triggered those API too much, the cost reflected eventually.

Liran Haimovitch  18:51
I'm sure one of the reasons DraftKings acquired you even was to build a big Israeli R&D center. So I'm sure you're hiring, and I'm wondering kind of what's the key characteristics you're looking for in candidates? And why should people come work for you?

Idan Fridman  19:06
So yeah, hiring is a major point for all of us, right? Nowadays, here locally in Israel. And I'm looking for-- it's not really related whether I'm in DraftKings or a small startup. I think this is a guideline that I'm taking for myself during my career. I like to be around people who take responsibility, curving to learn more, because eventually you come to work and you have different challenges on different aspects in your technology. But the ones that really want to learn and take responsibility end to end. Those are the candidates or the employees, that I want to surround myself with because you can learn from them as well. They can dare, they can take, they want to take risks, and they want to do more than just what they have to do, as I said, more than what is expected. So those are the people, whether it's a developer, whether it's in testing engineer or product, it doesn't matter. we want people who take responsibility and can think out of the box. This is very important for me. Not if he's a Java developer, not just a developer, because all those skills you can learn, right? But the soft skills, the ones that push themselves to learn more and take more responsibility, and taking end to end context. For me, this is really, really important. That's what I would look in. And again, it doesn't matter. If I'm a big or small company, eventually you have your own group, when you execute your challenges with the team that you have. And for the second question, why would people come to us? First of all, as you said, this is the first satellite thing that we are building for DraftKings here in Israel. I think that what makes us unique is the fact that we are working with sports. We are living sports, we are breathing sports and everything around this. The focus is on American sport, actually. So this is really interesting. We don't know if anyone else does this here in Israel. And this is really-- besides the fact that it's challenging because sports is something very dynamic and moves fast and a lot of opportunities to squeeze our product lines. It is also fun. It's very fun and you can talk about it. You can laugh about it and eventually bet about it. So that's cool.

Liran Haimovitch  21:16
Sounds awesome. Can't wait to join DraftKings.

Idan Fridman  21:19
Thanks.

Liran Haimovitch  21:20
Thanks. Great having you.

Idan Fridman  21:23
Thanks a lot Liran, I enjoyed being here.

Liran Haimovitch  21:31
So that's a wrap on another episode of The Production-First Mindset. Please remember to like, subscribe, and share this podcast. Let us know what you think of the show and reach out to me on LinkedIn or Twitter at @ProductionFirst. Thanks again for joining us.