Las Vegas 2022

How Software Engineering Boosts the Next adidas Commerce Transformation

adidas: How Software Engineering Boosts the Next adidas Commerce Transformation


Fernando Cornago

Vice President — Digital Tech, adidas



All right, so about the next speaker. Four years ago, the team from Adidas attended this conference, um, in London, and they left inspired specifically by a presentation that Jason Cox from Disney gave. And I have been blown away by what, uh, the Adidas team has done since then. So, Fernando Cargo is now VP of Digital Tech at Adidas. Uh, and he's presented at, uh, DevOps Enterprise three times and describing the progression of their journey. So last year, Fernando presented on the pandemic response at Adidas, which included both cost control and a focus on improving their e-commerce capabilities, which suddenly became one of the most important channels, uh, for generating revenue. This year, he will present on how top business leadership has chosen e-commerce to be a critical part of the strategy where they will be choosing to not just compete, but as they say, to dominate. So this will, uh, this is a story that was some challenges and some setbacks. Uh, they told, uh, last year about some of those challenges. Um,


He was, uh, in 2021, he presented with his team describing the growing pains that they had as they geared up to deliver against a strategic goal of having 50% of all revenue come through e-commerce within five years. So at this time, Aidas has thousands of engineers working on creating these capabilities. He just shared with me that due to their efforts and the support of people across the AIDAS enterprise, Adidas e-commerce sales outgrew Nike for the first time in years. So I saw two days ago, Fernando, uh, sitting with Courtney Kissler, who used to be at Nike. And to me, it's just this exam, amazing example how in this community, how friendships and collaborations are made, and yet they still compete fiercely in the marketplace. So I'm so excited that Fernando flew here to share the story. Here's Fernando. Oh, a video. Then Fernando,


I cannot see the notes, by the way. I cannot see the, the, the slides, but it's, it's, it's okay. I can, I can look back. So, hello everyone. Thanks a lot, Jean and, and Margaret for bringing me again. Uh, it's a blast to be here in, in person with, with all of you. Uh, my name is, is Fernando Cargo, and, uh, sorry Jason. I have the best job in the world, uh, seven years changing people's lives through sports. And the first five and a half years I was very into the foundations growing the foundations of software engineering, our cloud strategy, our cloud strategy, our platform strategy and operating model, our engineering expansion. Uh, but I'm a company man, and I was missing a little bit the action. So a year and a half ago, I move, uh, closer to our markets, to our business, and I could not choose a different, different channel that our most growing channel and the most technical lead of the company.


That is our, our e-commerce. And it is very, very important, very transformational because we are moving a big ship of 73 years of history into a 50% direct to consumer, uh, into 2025. So I'm gonna talk today a little bit about business transformation and applying architecture and ways of working into business transformation. And for this, I bring in virtually my colleague. I'm responsible of half of my, of our 3000 users, uh, margarita. So, uh, we are already the top five six Mno brand e-commerce in the, in the world, right? So our e-commerce business is the size, more or less of Puma or Lululemon entirely. And we are in that lease with the likes of Apple, Sarah, or h and m, but we are more complex than them in three things, or as I will say, we are more fun than them in three dimensions. One is our regional expansion.


So we operate in 62 markets right now and targeting 88. And you can imagine that the consumer expectations as well as our capabilities in us is a little bit different than what we have in Pakistan, right? Or Peru. Our scale is, uh, I think that the most changing in the industry. So while we operate typically at a hundred times lower order of magnitude than, than Nike, than Nike, than than Amazon, on the first two, three minutes of every of our hype drops, when we launch a limited set of our most hype products to our members, we have a scale of a thousand times bigger than than Amazon during these two to five minutes. So imagine, imagine the, the load that is put in our systems. This scale is making that every hour that our ecosystem is down, we lose on average, on a yearly basis, 1 million euros per hour last year.


We know that Nike in December, bad luck when the, we, when the east coast of Amazon went down, they lost more money than we in the whole last year. And the other, the third dimension of complexity or fun is a number of produces that we merchandise. So we have double, more or less than Nike on a on some of our markets, but 13 times more than Sarah on h and m. And not only that, we have this across 25 key categories, 25 key sports, and for every consumer, super premium buyers and also super value buyers, family supers runners, and we need to be the best brand for each one of them. Uh, we do this, uh, for doing this. We deploy to production, not counting hot fixes 4,000 times a year with more than a thousand product and technology people, which is basically almost half of the, of the engineering team of Adidas.


Some non-negotiables to go there. One is our tech hub strategy and our hybrid location strategy. We make each one of our hubs own a technical capability globally. This increase absolutely our power of attraction in Spain, in India, in Colombia, anywhere, and remove this first, second class citizenship of headquarter versus tech hub, right? If you go to any of these hybrid tech hubs, because we can, we have a hybrid remote working policy, you don't differentiate whose product, whose tech, who data, who is nothing, right? Uh, of course we, we heard yesterday we heard a lot of talks about, about OKRs and how to distribute work. Our five year strategy on the game that will take us to 50% direct to consumer has clear business KPIs on a yearly basis. We land this in yearly planning. We land this into quarterly OKRs and even into daily black logs of the teams.


So we are able to follow on a strategy and also to pivot quickly as a result. And it's not very visible in the, in the don in the last, in the last image, our workforce, our engineering workforce, more than 95% of them feel, feel top value and top mission of what they do. That's incredible. Coming from an engineering workforce that used to work in the minus one floor, right? Aside from the business, what they are asking us with only a 75%, they are asking us to remove the dependencies to be faster and to even be closer to the, to the consumer. So very proud of the, of the results of the last year. So we are premiumizing our complete ecosystem for originals with also partnership with Balenciaga, with Gucci, et cetera. And we are, we have relaunched in a month the complete loyalty program for Adidas ADI Club, where you can now redeem your points not only by discounts, but also by access to exclusive products, by experiences with our top at leads or even by sustainability moments, uh, for the, with the brand. Very proud of the green number on the bottom where you can see that in Q one, our toughest month, our toughest quarter with regards to inventory due to lockdowns in Taiwan last year, we were able to react better than any other brand with a task force changing completely the experience of all our channels in the lowest availability of product ever.


So this is the digital momentum and what, uh, what Gene commented is, uh, for the first time in years, and sorry, kni, we are outgrowing, outgrowing Nike in the, in the last quarter in the Q one, that is a little bit sifted than, than our time and not growing by far our competition, but these numbers are far away from our expectations. So we are not growing as we want it, and it's due to the perfect storm, uh, that we live in our macroeconomic macro political situation, right? Not only our inventory issues, but also data privacy restrictions everywhere, especially in apac, when the west part of the world recover from Covid, there's lockdowns in, in China all over the place is our biggest market, by the way. We are the stupid, uh, at the most stupid, uh, race in the, in the world. And after the two toughest years, we start killing each other by a piece of landlord for a, a little bit of gas in Russia.


And this creates like the economical situation that we all have right now. And if you zoom in, this is creating that. Now we are in a channel, in a situation in the company where we are set for a bigger growth that we have. So we need to control our operating overheads, especially, especially on the business side. So what is technology strategy? What should drive, what, what should we be optimized for us, as you advance said yesterday, right? For me, clearly for speed. So we need to be fast sometimes for us being the second is being already late and staying fast. A perfect example is TikTok in China, for us, it was a channel livestream. Sales didn't exist a year ago. We were, the first one in the market is now 30% of our China sales, which basically is 10% of our sales over the globe.


So more than 400 million euros that we will not be getting if our architecture will have not been ready to sell in TikTok in a week. So how in this situation, we are gonna dabble our revenue. We thought more people working on that. And it's what my, my, our chief digital officer said that what brought us here will not bring us to the next level. And I always use force wording, not only because he was, uh, jumping with Adidas, it's because he was not an athlete, he was for me, an innovator. He was in 1968, it was the first athlete using technology and it was a new technology. It was the first year where the people were landing on farm instead of concrete to think and work differently. And this is what we as technologies need to do, be aware of what is there and be ready to use them.


He was not the fetus, but he won the gold medal in Mexico, 68, and they prove that he was not the fetus is that he never competed again, right? And this is when what I tell my team and the, the thinking that I put my team, and this is the slide that every one of my team members is having below the pillow and is we need to stop thinking about technology and start thinking about game changing business capabilities that allow us to grow, to, uh, engage with our consumers without effort from our business users. And is each capability needs to be data driven and smarter than us. It cannot be that manual slotting of content or manual taxonomy is working better at automatic one, right? So it cannot be that our search is not learning about the next clicks of the user then operating in a hundred markets.


The time where we put the capability in the market and we spend nine months to put it globally are over. So if something is working tomorrow needs to be everywhere. And the last one is capability, adoption or automation. We need to stop, stop thinking about what the user only want and is how can we automate this with our user so they don't need to do any manual action anymore. And, uh, which capabilities are we talking about? And I'm talking about five of them just in the, in the, in the consumer steps of, uh, of the journey, right? And it's 70% is communications. So we try to automate, we target to automate 70% of our newsletter and behavioral driven communications to bring more users to our ecosystem. Second, with future of content and, uh, and personalization. Only 40% of the content that we create globally is used in the markets.


And the adves time of a campaign being live in a market is four days. Why a family sper cannot see the back to school campaign for a month. Why a running sper? Only a running supper should see the campaign about Adidas, about bras and size and fit. That doesn't make any sense, right? So this, this is gonna be the biggest disruption, uh, improving search and recommendations. So our customized recommendations for our most frequent and premium users are converting 900 times, 900% more, sorry, than the normal recommendations for our users. Last but not least is the checkout experience. So it looks commodity, right? So you buy and you click, we are only converting 20% of the people that create the basket. So imagine in a physical store, you go with your shops and only one out of five is buying. If we convert 35% of them, our targets for 25 are done, are done.


So basically we see that now with one, one click payments with Apple buy, apple pay with Google and biasing for these payments, we can reach up to 50% conversion on the lower funnel. And last but not least, our internal capability for buying trading, we saw the target target team on Monday is very important. We cannot afford to have a product. Our, our, our digital ecosystem is our biggest, most important and our most expensive store in the world. We cannot afford to have produce without inventory, produce not converting or campaigns not converting. So I'm gonna talk about what we are doing architecturally and ways of working so architecturally and thanks to our architecture team to come with this idea, we have come with a very easy concept that is our capability diameter. So a business capability is composed about a consumer experience for our consumers, about a lot of base backing capabilities provided by different teams.


And it's at the end creating an experience for a business user. Ideally automated. Imagine green segments. You have the experience, you have backing capabilities on our application, ecosystem, our social ecosystem on our omnicell and our supply chain. And then a market can enable or disable it depending on, on the strategy. So we are moving from an architecture of headless custom experiences and out of the box admin based on tool set into an intelligent, what we call ecosystem. We need to put the name on it. And now we are, we are creating a, a serving the company to put the name where we have an intelligent ecosystem. First of all, it's global through d s l technology. Same code with a t a different configurations in the markets that the markets can touch is decentralized. So the user, the the team in charge of a business capability for a user, sorry, is in charge of the business capability in the micro frontend admin environment.


It is, it is evolutionary. So the way that you create content right now is not how you create content in three years from now. So now you create content for the position one tomorrow you create content for the runners the day after tomorrow you create content for the women runners and the content is gonna be there until it's converting more than 2%. So this is the mindset and this is how through data flows and real time data architecture, we can create smarter experiences. And last but not least, how is this for our ways of working? So thanks to us moving into microservices and micro frontend from the last three years, we are now able to change our operating model for our business transformation. So our tech transformation was enabled by us thinking frontend and backends. While our business transformation is on us thinking consumer steps of the journey.


So we have teams and this is really optimizing for speed and getting closer to the consumer. Every engineer is aligned no matter if he's working on front end and backend with a business metric, with a business goal, reaching a consumer, converting a consumer, serving the consumer, et cetera. And just for free and just by chance this alignment is also aligned with our business users so we can really accelerate the transformation together with them. Uh, just to finish, I brought virtually, uh, my friend Margarita. So she's responsible of all the business operations of e-commerce as well as consumer service. And she's gonna talk about how this, this experience in our most complex sale of the year was in August. It was our hub day hype day and is when during 32 hours across 60 countries, we launched continuously drops and raffles of our most wanted products, reaching 5 million requests per second in our systems, 85% of them being bought for reselling afterwards. We want to, uh, allocate the products in a fair way to our most loyal members with 50 dimensions, uh, around allocation and serving out of 2000 nodes, distributions and the retail stores, et cetera, in order to provide and deliver the best, the best the first shoe in less than an hour to the first consumer. Uh, all this in a complete new ecosystem because there was not software in the world that is giving us the scale that we needed with these numbers. So let's, let's hear. Margarita. Hi,


My name is Margarita. I'm the key business operations for eita. I'm based in na. I lead teams who oversee business performance across IT, fulfillment and consumer services. In essence, we are running our virtual store, which is also our number one store. And our ultimate goal is to make sure that we offer our consumers a full range of products that we have available in our disease and that we offer the best consumer experience before you order and after you order. And of course this is easier said than done, right? This is also where Fernando and tech teams are really our strong partners. And I will give you an example how this August we executed easy day. It was our biggest one day sale, which generated 210 million in under 32 hours and it was executed globally. Now, the biggest part of workload for the operations teams used to be in managing inventory.


Inventory is replicated across several systems, which means the teams had to manually reconcile inventory, snapshots, doing it sometimes in exam. And after the drop, they had to manage the backlog of orders. And unfortunately we had to cancel quite a number of orders 'cause we oversaw, which means that we showed the stock, which we no longer had. It's really a negative consumer experience. Imagine there is a shoe that you really want, you sign up for the drop, you get selected, your spirits elevated, you pay, you wait for the shoe for a week, and after a week or 10 days, you receive an email that unfortunately you order was cut. Imagine that. So with the new technology and the new tooling, inventory replication is automatic. It's done once. There is no manual work involved. And there were zero discrepancies. My team woke up the next day after the drop, after the easy day, they came to the command center and they were like, look, we are looking strange. There is no backlog to manage. Are you sure things are right? This is a great message to receive.


Uh, it's a pleasure to hear a business user talking, talking about this just to close, right? Uh, and Jean asked me, what will you tell Fernando 2017? So four things, right? So first of all, enjoy what you do and follow your passion, even if it's costing a lot of gray, gray hair. Uh, second on tech for me, focus on value and speed. The rest is coming. So the biggest efficiencies you are creating, if you focus on value and speed, impossible is nothing. So don't, don't feel threatened, uh, about big software vendors with 10 times more engineers than you. Because for your battles where you wanna be the best you have, something that they don't have is passion and focus. And last but not least, uh, an improvement point for me is be aware of everyone's culture and everyone's fear to change. Try to bring them into the journey. Do whatever it is. So they, they, they feel that they are part of it, but understand and accept that there's gonna be one person wanting to be the last man standing in a data center with three machines, right? So thanks a lot.