Host Intro (Gene Kim)

Thank you, Alfredo and Olympia.

About one year ago, I saw Jakob Knutsson, Product Area Lead Engineer at H&M, present at a conference, and I was so intrigued by his presentation that I reached out to him. In our correspondence, he shared some amazing things that made me immediately ask whether he would be able to share his story here at DevOps Enterprise, because I know their story will inspire this community.

H&M Group is a leading global fashion company. According to their annual shareholder letter, in 2020, they operated nearly 5,000 stores in 74 markets around the globe, with net sales of 187 billion Swedish krona, which is about $22 billion US.

I am so delighted that Jakob is co-presenting with Daniel Claesson, H&M Group's Chief Product Officer, who is responsible for overseeing the strategy, development, and operations of all digital and tech products, enhancing the shopping experience for customers to H&M Group's global family of brands.

Here is Jakob and Daniel.

Opening Video (H&M Group colleagues)

Yes. I'll just bring out me.

My name is Pahvara Sam.

I am Fredrik.

I am Misba.

Arpita Nagpal. I'm a product owner in new Business Tech.

It's very important for me to be creative. That's my way of unwinding.

When I dance, I just forget the outside world.

I am a machine learning engineer. We are looking at how the customers behave as a group, or even as an individual. It's magic with math.

I am a software engineer. When you think too long on a problem, maybe you get stuck. When you do something else for a moment, then you solve the problem much faster. If it doesn't turn out the way I want it, undo and go again.

I am a product owner at Business Tech. H&M, of course, as a global company, gives me great opportunity to interact with other stakeholders, different countries, cultures.

We make sure that it's secure, frictionless, and fast and easy to pay.

What are the cool things that we can do? What are the realms of possibility? We want to build that together with all these new people that will join us.

Jakob Knutsson

Last year, I was invited by Gene Kim to present H&M's cloud and DevOps journey over the last years and how we're doing that at scale. Today, we will give you the other side of that story from a business perspective, as well as where we're heading in the future. And we're going to tell you why tech transformation is needed, how we got started, and also implemented organizational changes, and how that has allowed us to serve our customers and colleagues better.

I'm really glad to be joined today also by our Chief Product Officer, Daniel Claesson, here to tell his side of the story and to break down the importance of incorporating tech into the DNA of our business. Tech is fundamental in everything we do.

We're now living in the age of software. So are we a software company? No. I would still argue that we are a fashion and design company, but that we must master software at scale.

And if we go back a couple of years, we were a classical enterprise IT department, heavily siloed, built for stability and expansion. This served us really well for a long time, but we were slow and there were a lot of handovers and bureaucracy. We actually started our first online shop in 1998, and that was built around mainframe. And already back then, they had a tight interconnection with our business. With a focus on globalization, they could quickly enter new online markets, but as they grew, even changing the color of a button took a lot of time.

So they went from a few agile teams to scaling that to eventually do waterfall. Around this time, we also saw a grassroots movement that were exploring new ways of working, new technologies that had emerged, and challenged different time-consuming processes. There were smaller agile islands popping up in different parts of the organization. There were these teams around the company that acted as change agents within our buying and production and within our online, and then later also within AI and analytics and cloud as examples.

What were our blockers? Where did those time-consuming handovers take place? And what were the constraints? One example is that onboarding a new development team and getting a dev environment up and running could take up to two to four weeks on premise, whereas in the cloud, we have that fully automated and could provision a dev environment and a landing zone with integrated dev tooling in under 10 minutes so that the team can start creating value and test their ideas instantaneously, fully democratizing infrastructure and providing 100% self-service capabilities.

We also pushed teams moving to the cloud to DevOps practices and to handle their own cloud resources, pushing the now known ideal from The Unicorn Project, locality and simplicity, so that any team could independently develop, test, and deploy value to the customer.

Cloud gives us a faster pace, a better global scale, value for money, built right of course, and access to value-added services and innovation, so we really can focus on our core. These teams could grow together and inspire each other, and everyone had a lean and DevOps mindset. We fostered a culture of automation, new ways of working, and pushing responsibility left, and teams really started to prove value. We also recognized that the number of deploys each team can do is a true competitive advantage. We were also having a lot of fun and had leadership that believed in us, true to our values of having an entrepreneurial spirit, keeping it simple, and believing in people. And this was the start of something new, and the feeling was exhilarating. This wasn't one person, one team. It was a movement.

In driving our digital business model, we've come up with several key tech principles. We want to be customer-centric and become local everywhere, online and off, and at every stage of our colleagues, customers, and partners' lives.

Open: we want to move away from complex legacy integrations and use APIs and events all the way. It should be easy to communicate with H&M in a secure way.

And agile: being able to release on demand, create value faster, and continuously surprise and delight our customers.

And also intelligent and data-enabled. We have a lot of information and events flowing through our platforms, and we want to leverage that data to get insight, take data-driven decisions, and enrich our business cases with AI and analytics.

And finally, global. We want to be able to provide a great customer experience to serve our customers in each unique market.

At this time, we also started doing a lot of internal events and knowledge sharing, and we hosted forums, hack days, problem-solving sessions, and training. We set up a developer portal to centralize the information, started to build up shared reusable assets, reference architectures, reference implementations, and started to form new golden patterns.

But the community also lives in different channels. You can ask questions, and people from every part of the organization jump in to answer. One of the most popular channels is called Absolutely No Idea, and it's amazing to see how fast things can get answered and resolved. This is our engineering community that we now call TechSpot. And the best part is, it's community driven: democratized events and knowledge sharing, making sure everyone can schedule events and reach the entire community. And as you can see, we have 159 questions posted in the last 30 days that got over 729 replies. Talk about empowered community and shortening lead times, pinging in the right team for your problem or collaboration.

We really started to see a culture shift. The agile islands, they grew. Whole divisions started scaling and doing big room plannings. The teams were building up speed and enabling innovation, and we also set up enabling teams to onboard others. Key development areas had been identified and anchored, so we were on a good path. The teams could go to work. But you can imagine where we're coming from. We don't turn around in a day, but now we had full focus.

We always had great values, and these are combined really well with agile ways of working and a DevOps mindset. For instance, an example in change in our way of working resulted in moving from 18 to 24 months feature deployment to three weeks. We still have some work to do. What got us here won't keep us here. It's not enough.

All these small steps led up to a bold decision to update our cloud strategy, saying that all of H&M Group should run entirely in the cloud, supported by an organization that can build effective and secure cloud solutions using DevOps practices.

This later grew into strategic tech transformation moves for our whole group, empowering our teams to have clear direction and objectives. And having clear strategies and objectives helps a lot. DevOps, agile, decoupled architecture, AI, and cloud are foundational building blocks for us. The vision is there, the foundation is there, culture is there. Our focus areas now are breaking down platforms, monoliths, and systems of record. And joining me now is our Chief Product Officer, Daniel Claesson, who's going to tell his side of the story.

Daniel Claesson

Thanks a lot, Jakob, and thanks for having me. Really, really pleased to be here.

To understand our journey also from a business perspective, I think it's important to understand some context of where we are as a company. H&M Group is really a family of different brands, and we have been around for a while, with, of course, H&M as our biggest brand.

Looking at fashion, looking at our industry, looking at retail in general, of course, this is probably the most interesting time in history of being in retail. It started before the pandemic, but it has really, really accelerated. And of course, a lot of things are changing, but some things are also actually the same. If we look at our customers, a lot of the customer needs are actually still the same. The customer is still a person behind that actually wants to find what they are looking for. They want to look and feel good. And they want to have a great experience and experience a great service and a great interaction with the company.

The things that really have changed is that with the rise of digital and tech, we have gotten, as a company, a completely new toolbox that we can use in order to serve our customers. So with this full context, we have realized that we can't any longer make a difference between tech and business, because business is tech and tech is business.

And with this in mind, we actually took a step some years back in starting to explore: is there another way of organizing ourselves to really make sure that this comes through? And a year back, we established a new function that we then call Business Tech. This function is a consolidation of what before were our traditional IT department that Jakob talked about, our business development department, and our AI department. And we formed this new function into a true product organization with the purpose of really bringing all the competencies needed into one team that can solve the problems they need for the customers.

So with this, we are taking some more steps, and the only thing we know is that we see clear tendencies that the only way forward is to continue to bring in business and tech even closer to each other, because this brings a lot of value for all the different perspectives of our business.

Continuing with this, there is a thing for us, for these teams that we are now building. We are really trying to form the teams in a way that they can solve the problems for the customers and the business. And we are not just here to serve the business. With this, we are, of course, digitalizing our customer experience, we are digitalizing our colleagues' experience, and we are digitalizing our full operations.

And there has been, during this time, of course, a lot of different steps and, if you can call it, aha moments or important things that we really have realized. The first thing has been to really deeply realize that we can't any longer make this difference between business and technology. It really needs to go hand in hand and together.

And the other one is that there's a big mindset shift in this. As a business person, you need to understand way much more tech and what possibilities tech brings to the table. And as a tech person, you need to understand the business problems to a deeper extent. But when this happens in the teams, when we get the different competencies around the table to really solve one problem, magic clearly starts to happen.

But the mindset shift has been way much harder to achieve, and finding that way of working and that way of thinking, than reorganizing or doing other stuff.

So what's next? In our minds, there are two things that we really are thinking around and would love to have the help of this community to think around. The first one is: how do we continue to bring tech and business even closer together? Because that is only one way forward, and it's continuing that journey.

And the second thought we are thinking around is connected to omni-channel, and how to really utilize both logic, AI, data, and all the tech we have, but at the same time finding the balance, because we think it's a balance between tech and the human interaction. So would love to get the input from this community: what do you see as this balance between human and tech going forward?

With that, we would like to say big thank you for having us. It has been a true pleasure to being here together. And hopefully we will be back next year and tell you how far we've come on our journey then. So, thanks from us.

Closing

Thank you. And thank you, Jakob.

Yeah.