#8 Cristian Constantin Olarasu – Director, Central Product Management at Nike on aligning expectations around AI projects

Consumer internet junkie with a passion for platforms and experience design. Forbes 30under 30 | 2013. Now tackling global growth at Nike. Previously attacked the enterprise at Lore IO. 10+ years of experience in growing technology companies. Background in new product initiatives, software product design and development. Figuring out gaps between prod/eng and business teams. Deep interest in big-data technologies and AI/machine learning.

#8 Cristian Constantin Olarasu – Director, Central Product Management at Nike on aligning expectations around AI projects Get Your AI On!

Ciprian Borodescu: I’m here with Cristian Constantin Olarasu – Director at Central Product Management at Nike. I’m super excited, and it’s an honor to have you on this podcast. Thank you so much for being here.

Cristian Constantin Olarasu: Hey, thank you for inviting me. I’m happy to chat.

Ciprian Borodescu: You have over 10 years of experience in growing technology companies and your deepest interest is figuring out gaps between product engineering and business teams, especially in big data and AI companies. I want to start by inviting you to briefly talk about your journey from founder entrepreneur in the early days, to the product management role you have today at Nike.

Cristian Constantin Olarasu: I think it’s a long journey. Since I was younger, I kind of started mingling with technology – I think we called them technology projects. And we ended up doing a bunch with all sorts of people in all sorts of teams and some of those projects became companies and some of those companies became more successful than others. The last thing we worked on was a system that helps teenagers maintain their weight. So, we ended up partnering with Stanford, and we built this obesity company that eventually got acquired by Weight Watchers. And then, one of my previous investors introduced me to this person that was working at Walmart labs and we started working in the Big Data space figuring out how to answer ‘why’ questions for enterprises and build systems that can answer them automatically. So, at that point, we started growing the company and I think it was 2019, we went through a pretty long series of M&A discussions after which we had some product innovation, and we decided to raise another round and continue. And for me, that was just just a mental break from what I imagined the path forward to be and I literally took a break. In that period of time, someone reached out from Nike that they had this, they were calling it ‘the growth’ team and they were looking to have this little team inside Nike that was doing literally great things. And after talking to them for a while, I kind of decided to join – being my first job ever if I can say that.

Ciprian Borodescu: Alright. So I want to push you a little bit. You’re originally from Romania, and I know that you were also a founder and entrepreneur. And I would like to hear a little bit about that part of your journey.

Cristian Constantin Olarasu: It’s kind of random. I think I started when I was younger – kind of skipped school. I started when I was younger, we were doing some tourism project that was extremely badly managed and from that one… You know, back then when you knew stuff about computers, people would reach with two types of requests. The first one would be “Hey, can you fix my PC?” So kind of IT services. And the second one would be, “Hey, can you design my logo and my website?” Because the idea of investment was kind of foreign to us back then and we didn’t really know how to generate revenue streams and grow a company, so we ended up doing all sorts of services around what we were set to do. And to some of those services, we interacted with some of the branding agencies around and we figured out that there isn’t… When they’re analyzing or doing a campaign for a client, there’s not much attribution or impact to the financial economical sense of the business. They were talking about results in terms of brand points and clicks, and nothing really correlated with the business. So we were thinking back then, like, there must be a way to quantify this value that’s being generated that’s not necessarily financial, economic. And we started with the brand value. We ended up being funded on that company, which was a small disaster also. But we ended up learning a lot. So through that process, we got in touch with Nestle, with their baby division, and we bumped into a different problem – totally unrelated – where we figured out that parents have a lot of questions when it comes to raising children between zero and three years old, and those questions are usually answered by pediatricians or online content, but it’s pretty specific for each child, and once you understand the specifics, you would be able to build a system that can help those parents in the system. So, we ended up building that. By that point, we had an idea of what a company is and what it means to raise money, and what we wanted to do, so we eventually launched that product in San Francisco. We went through an accelerator in Canada, then we met a couple of great people in the Valley and we ended up pivoting; we ran out of money, we got some money back, and some support, but eventually we started that obesity company. And after that, I just told the story.

Ciprian Borodescu: Nice. And if there’s one or two lessons from that experience, what would those be?

Cristian Constantin Olarasu: Oh, my God.

Ciprian Borodescu: I promise this is the last question about this topic.

Cristian Constantin Olarasu: No, no, no, no, no. Oh, my God, I’m thinking through it as you’re asking. It’s very hard to explain the context that was present at that time. I think it was 2011 or so. Like, I sometimes explain to people that when I finished high school, I didn’t really know how to explain what a job, for example, is. Or college. Like, it just didn’t click for me back then – the fact that the company hires people and what’s an economy. Like, all those concepts were kind of foreign and they just didn’t make sense because of the education, the context. So, everything that happened since I was 19, till I was like, pretty late, like 26-27, I think it was kind of a combo of luck, misfortune, and inertia, without a lot of thought put into the next step, or a lot of intent. Everything was kind of chaotic. So, I don’t think I’m in a position to give advice. I think the only thing that kept it alive was that we were genuinely interested in the things that we were working on, so that kind of guided us. Some structure would have been great. But every time the scope was not to solve some problems we were going after, it kind of went to flames. So maybe that’s one piece of advice, to just work on the things that keep you up at night and forget about everything around it. They’ll just figure it out by themselves.

Ciprian Borodescu: Yeah, and recently, I started studying stoic philosophy and in stoicism, you can say that by focusing on the things that you can control. Even today, even with a great lockdown, I think that it’s a good personal philosophy to have. You know, there are so many things that are out of our control, and focusing on what we can control is a good way to move forward, I guess.

Cristian Constantin Olarasu: Yeah, definitely. I wish I knew this when I was younger.

Ciprian Borodescu: Well, yeah, me too.

Cristian Constantin Olarasu: But yes, good point.

Ciprian Borodescu: Good. So, let’s get back to today. What are some of the AI initiatives you’ve been involved with – at Nike or in general – that you can share with us?

Cristian Constantin Olarasu: So, maybe first give a bit of context of how our team operates and some context about Nike, in general. Nike is a wholesale company and that means business-to-business – they had business-to-business relationships over the last 47 years or so. And, as you’ve seen in the news, a couple of years ago, Nike decided to go direct-to-consumer, and they started the direct-to-consumer office, which has been extremely successful, growing from zero to over $10 billion in four and a half years or so. Now, to do this, part of the work that they’ve done, one of the teams that really pushed the envelope is this team that acts as a center of excellence for product. And I think it’s a model that has been employed in other companies pretty successfully before, where you have this some sort of a SWAT team that goes after high-growth initiatives, and at the same time, acts as a center of excellence for product and bringing a product model to a large enterprise where decisions are being done to solve real problems that impact the business directly. And you have this dysfunction of product management that’s being structured. Now, what AI initiatives are happening? I don’t think I can speak a lot about it, but I think, in general for companies that are going from retail, and they’re stepping into eCommerce, and they’re going from selling and marketing one-to-many, to selling and marketing one-to-one at scale, they have to go through this, I don’t know, let’s call it digital transformation for lack of better words, where they truly need to understand what’s sticking for their customers, and what are the needs of their customers, and how to engage those users and how to retain them. They kind of have to transition to thinking a bit more of like a consumer technology company, as opposed to a traditional retailer. And, in that sense, I think anything related to personalization, anything related to operational efficiencies and automation will be high on the list.

Ciprian Borodescu: Okay. And I know that you’ve been involved with a lot of product teams over the years. What are some of the most important roles you believe such team should consist of? Especially if it’s an AI company at the core. And maybe, you know, you’ve been a founder and entrepreneur – maybe talk a bit about these teams as relative to the stage of the company, from startup to scale-up company, to maybe enterprise the size of Nike.

Cristian Constantin Olarasu: So let’s first talk about a different thing. I think it’s useful to talk about the maturity of the technology and product organizations and those companies. Let’s say on one end, you’ll have traditional companies and on the other end, you’ll have the Amazons of the world. And in between, you’ll have things like Walmart or Target or something like that. Now, depending on where you are on this scale, you’re going to approach problems related to, let’s call it data and automation and AI research differently because, most probably, it won’t be core to your activity; or maybe it will. So, the chances of being very different between the three buckets are pretty high. Now, assuming that the AI department is not really set up yet, it’s a very different approach in my experience than traditional product management on how to build those products. Because, instead of going after features and wireframes, and customer discussions, you’re building things with one of the biggest scopes in collecting data and getting more input for your system. And at the same time, internally, inside the organization, it’s very hard. Like, everything around AI is expensive, from hiring to educating everyone, to getting expertise on board to iterating, and teaching the organization that there’s no big launch moment, it’s an optimization problem, and you have to work for it and you have to change your mindset as an organization. So, that is built gradually, in my experience. And the best way to build it is you kind of want to start small, but with something that has a pretty large business impact, and something that’s feasible also. Even for that thought, you have to bring a lot of people together inside the organization – you need data science, you need data engineering because most of the work is in data engineering before people can mingle with the data. You need your researchers; in some cases, you need design involved, you need the product teams involved, you need the domain experts that will tell you and educate you on the business feasibility, you will need the technical experts that will educate you on technical feasibility, and so on and so forth. And then you’ll have to pitch it to executives or senior leaders and you kind of have to link it to the AI strategy if the company has one or you have to just build a strategy at the same time, so your project is successful. I think it varies a lot. But to summarize it, I think, in most organizations, it requires a lot of executive leadership education. It’s almost impossible to start with a big bang, so you have to figure out a way to start small, but pick something that can either be scaled or has a pretty big business value. Otherwise, no one will listen to you or your project or your strategy. And the last one is the mental model in product management kind of flips a bit. And that needs, again, iteration and learning.

Ciprian Borodescu: Okay, so basically, I think one way to start or to dip your toe into the AI realm is to implement kind of like a proof of concept or a pilot. And probably you’ve seen that a lot. What are some of the challenges you’ve noticed when taking an AI project from the proof of concept stage to production? What needs to happen?

Cristian Constantin Olarasu: There’s a lot of misaligned expectations. That’s my experience. You either have a lack of communication around the project and what it’s going to achieve or you’re in the middle of a domain expertise problem that the technical team doesn’t really understand, so they’re going off of a different outcome, or the outcome of the project is not fully aligned – so again, misalignment in expectations. Then the second one, I’ve seen a lot where people are planning for the first iteration, but they’re not planning for a continuous cycle of iterations which probably won’t work, to begin with, and then you have to make it work. So, you’ll definitely need multiple iterations.

Ciprian Borodescu: So, in other words, a proof of concept doesn’t end with a binary result – 0,1. It’s a continuous process, right?

Cristian Constantin Olarasu: Exactly, exactly. And that affects the planning process in those companies, which is again, impacted for sure. Because you can’t expect to plan the same for your scrums, or in your waterfall development in some cases. There’s a lot of strategy decks that are talking about a lot of things that are not necessarily feasible and didn’t go through a real feasibility review on the other side, where the business problem is kind of clearly defined, but, again, there’s a misalignment in expectations of what AI can actually solve or what it can’t solve.

Ciprian Borodescu: Yeah, and I think this is where an AI product manager can have a real impact in explaining everything and all the expectations – set the expectations to the leadership of the company. And this is one of the questions that we also got during Techstars, when I was presenting myself and Alexandra as CEO, and CTO and people were asking, “Okay, and who is the product manager?” And obviously, as a very small early-stage startup, we didn’t have a CPO, or the product management was whatever came out of my conversations with Alex. So, based on your experience, when do you think is the best time to actually hire a CPO?

Cristian Constantin Olarasu: Well, I think the function of product can be independent of role or background. And we’ve seen that very successful. I think it’s useful to have experience and so on and so forth, but the question is not necessarily when do you want to hire a CPO, but when do you need a product function in your company? And what does it do? Like, what the heck do those product people do? I think the easy way to explain is you kind of need a product person from the start – that’s my opinion – even if a technical lead is acting as a product lead, also; that can work especially for smaller companies. On larger companies, there’s a lot more work on alignment. But you want the product function to be fully accountable for solving the right business problem and nailing the right expected outcomes for the thing that you’re going after. We keep on giving the example with the needs that the PMs can solve. And we’re telling people inside the organization that you need to think about a need that will be there, it’s big enough and won’t change over time. Like, for example, safety in a car is a good example where safety was solved initially by brakes and the seatbelt, but slowly it evolves. And there’s all sorts of other tools that are solving for safety. And now, in recent years, we have AI or machine learning that help solve for safety. But the problem with safety in cars it’s an enduring need. They’ll stay there for a while. So, that’s one function of the PM – identifying those enduring needs and being fully accountable for the expected outcomes that have been put in front. Now, for smaller companies is useful – depends on the product, of course – but regardless of whether the product is an interface, an API, or anything else, having someone to think through all the angles that come from the product is extremely valuable. Now, a CPO is a formalized leadership role in product and I think either the company is at the point where they can hire a CPO – so that’s a bit at a later stage – or someone grows into the role, which I think it happens in a lot of the examples I have around.

Ciprian Borodescu: Alright. I think we covered this one – what are some of the do’s and don’ts of building an AI product. If you want to add anything to this point?

Cristian Constantin Olarasu: No. Just don’t assume people understand.

Ciprian Borodescu: Okay, that’s a good one. Not necessarily linked to AI products, right? It’s a general rule.

Cristian Constantin Olarasu: Yeah, I think in technology in general, especially when you’re going into those larger organizations, either internally or externally, they just don’t assume people understand. Moreover, rather it’s better to assume people don’t understand; they have an expectation from you to explain how things work and why are they valuable. My experience is that you always want to lead with the value, as opposed to the technology solution.

Ciprian Borodescu: Lead with the value. That’s a good one. Yeah.

Cristian Constantin Olarasu: And with AI products it’s very, very easy to create false expectations. Again, that’s very, very common, in my experience, so making sure you’re aligning with people that the output will be, I don’t know, 85% accurate, that’s good enough for the business. I think it’s a useful thing to do upfront.

Ciprian Borodescu: Alright. There’s something I would like to discuss with you. So, it seems that there are two schools of thought when it comes to communicating about a company’s AI capability. Either package it as a solution to a problem and don’t communicate at all that there’s an AI involved behind the scenes. That’s one extreme. The other extreme is yell from the rooftops that everything that the company does is AI or machine learning. Where do you stand on this spectrum and why?

Cristian Constantin Olarasu: If you have a group of renowned researchers in this field and you just build a company to solve AI problems, maybe it’s great to just yell that you’re good at AI. In most other cases, again, I think it’s useful to lead with the value and then if you discovered some areas where machine learning or AI can help enhance your product, you can talk about it, but most often people are not talking about the fact that there’s some complex machine learning system behind the magic. And on the previous one, like, there was a first-hand experience on my end, where we figured out after we sold the company is that nobody gave a crap about our AI. So, it was useful to be a buzzword on the deck at that time, I think four or five years ago, but other than that, it’s unfortunate that that was the case, but that was the case. We were naive enough to think that if we build it, they will come because it was an enterprise sales process, and, due to our backgrounds, we were able to convince a bunch of enterprises, we got some contracts, but eventually when it was time to see usage, no one gave a crap about our AI.

Ciprian Borodescu: Got it. Okay, so this first part of the year was crazy.

Cristian Constantin Olarasu: Yeah, it was. It actually is, still.

Ciprian Borodescu: Yeah. What about the second half of this year? What are your plans?

Cristian Constantin Olarasu: I personally think it’s gonna be a bit crazier because we’re gonna see the actual economic impact of this whole shit show. So, in the last couple of days, a lot of social issues that were barely spoken of, have been put in front of a lot of people, so that will definitely have an impact too. So, you have two things that are kind of lingering there. One just started and the other one has been there for a while. I think it’s a good example to look at what some of the Asian countries’ progress was through this crisis. Like, I know for sure China is fully functional today and maybe that’s a positive indicator, but at the same time, the scale of the economic impact on their end was… You can’t really estimate on the other side, but my sense is that they weren’t hit that harsh.

Ciprian Borodescu: Yeah, yeah. And on a more positive side, or a positive note, if today or tomorrow, you could get out of the house and COVID all of a sudden disappears, what would be the first thing that you want to do, now, in full freedom, let’s say?

Cristian Constantin Olarasu: Oh my god, it’s super clear. I would take the next plane and go and hug my parents and tell them that everything is gonna be alright.

Ciprian Borodescu: That’s nice.

Cristian Constantin Olarasu: For sure, being away from the family for this long in this context, I don’t think it’s easy for anyone.

Ciprian Borodescu: That’s awesome. Alright, so, for the final special section of the podcast, lightning questions and answers – a series of fun, short questions that you have to answer really fast. Ready?

Cristian Constantin Olarasu: Okay, let’s go. Yeah, let’s go.

Ciprian Borodescu: Game of Thrones or Lord of the Rings?

Cristian Constantin Olarasu: Lord of the Rings.

Ciprian Borodescu: Star Wars or Star Trek?

Cristian Constantin Olarasu: Star Trek.

Ciprian Borodescu: Joe Exotic or Saul Goodman.

Cristian Constantin Olarasu: I’m not gonna answer this.

Ciprian Borodescu: Okay. Your favorite movie?

Cristian Constantin Olarasu: I thought about it. Let’s say Up.

Ciprian Borodescu: Okay. Cats or dogs.

Cristian Constantin Olarasu: Dogs. Definitely.

Ciprian Borodescu: Alright.

Cristian Constantin Olarasu: I think cats are great, don’t get me wrong. I just prefer dogs.

Ciprian Borodescu: Okay. Favorite startup?

Cristian Constantin Olarasu: UiPath.

Ciprian Borodescu: Okay, that’s no longer a startup, man.

Cristian Constantin Olarasu: Yeah, that’s fair. But a favorite new startup that I heard about – Alpaca Markets. I really like what they’re doing there. They have an API for operating exchange markets.

Ciprian Borodescu: Alright. The last book you read.

Cristian Constantin Olarasu: The coach. I think it’s called “The Trillion Dollar Coach” or something like that.

Ciprian Borodescu: Okay. I think I heard about it.

Cristian Constantin Olarasu: By Bill Campbell. Great book.

Ciprian Borodescu: Awesome. Okay, so the bonus question. I know you’re passionate about blockchain – and so, I have to ask. Let’s say you’ve been given the chance to have dinner with Satoshi. What would you order?

Cristian Constantin Olarasu: Wow. Okay. What would I order? I don’t think I’ll focus on the menu – or anyone would focus on the menu and what to order – if we would have dinner with Satoshi. 

Ciprian Borodescu: That was a trick question. Alright. Is there anything you would like to ask this guy?

Cristian Constantin Olarasu: No. Okay, let me give you some context. I think a lot of people want to have dinner with those people – like, disregard whether it’s Satoshi or Obama or whoever. When you actually ask the question, “What would you talk about?” For me, personally, it’s a gap. Like, I wouldn’t know what to ask Satoshi, for example.

Ciprian Borodescu: Yeah.

Cristian Constantin Olarasu: There’s no advanced plan or some sort of information that I can barely wait to ask him.

Ciprian Borodescu: Yeah. Alright, man. Cristian, it was a pleasure to have you, and thank you so much for sharing your wisdom with me and with us. How can people reach out to you for ideas and comments?

Cristian Constantin Olarasu: Just cristian.olarasu@gmail.com.

Ciprian Borodescu: Awesome. Thank you so much, Cristi.

Cristian Constantin Olarasu: You’re welcome. Thank you for having me.