H&M, the Swedish multinational clothing-retail company, has recently seen enormous growth in its ecommerce business as it continues its strategy to integrate the offline and online experience.
This integration will likely accelerate in 2020, as H&M, and many other retailers, discover brand new audiences who have previously only experienced their brick and mortar locations.
So how does H&M ensure their digital experience is as user-focused as possible? What are the practicalities of running effective user research and testing at a global, multichannel retailer?
We sat down with H&M’s Product Design Lead, Stefan Twerdochlib to discuss these matters. But first a little personal background…
Stefan Twerdochlib has a name that literally means hard-bread in Ukrainian. He’s a Canadian now working and living in Stockholm, Sweden and he has two cats appropriately named Jotunheimr (Norse land of the giants) and Yggdrasil (Norse tree of life).
Stefan has a passion for design, racing bikes and drawing sumo wrestlers, which apparently led him into a career in UX design where he now leads H&M’s check-out journey. In his spare time, he’s training to compete in his first triathlon this year and enjoys sipping on a nice tequila.
In this recent interview with Stefan, we talk about one of the above things. See if you can guess which one?
After college, my first job was web design for a life insurance company and it was about six years into my career before I began to learn about UX. I’ve worked with some designers that prefer form over function but I was always about function over form. I knew I was designing for other people so it made sense that they should understand those designs and be able to use them.
Now I’m here at H&M where I lead the checkout team, and building our design system, which is a bit of a beast itself. It’s a team of five right now and we handle the web, app, physical checkouts and payment teams.
When I started a year and a half ago, UX was just a wireframe factory, and UX was a consideration that came into play relatively late in the design process.
We had long pre-study methods and processes which took months, but I wanted to speed it up and bring in something like the Google Ventures design sprints where UX was a consideration from the beginning.
For me, the best approach is not just tell people about things but to show them and bring them along for the ride. So we’ve involved people in the process – for some it’s been their first time actually witnessing usability testing.
We’re going through a big transformation, changing our way of working, and UX became one of the key pillars.
What I try to do is to test and ask questions at every stage – does this flow make sense? Does the sketch make sense? We don’t just wait until we have some sort of visual design or a finalized wireframe.
‘Show your working’ – Stefan’s own notes on the H&M checkout process
When testing, it’s important to question everything and avoid any type of bias. It’s important to test, not just validate your hypothesis, but also to try and disprove it because if it still stands up after this you can then in turn prove that it’s correct.
This helps to avoid confirmation bias, and makes it easier to prove the value to stakeholders.
This is still in process. We are always testing and trying new ways of working, especially when it comes to cross collaboration and bringing in other areas in our process.
In terms of product development, we’ve been doing these design sprints and getting UX insight so we’ve been able to influence development from the start.
So we’re helping to set the product strategy and understand what features are being focused on and having that conversation early, and also drawing up what the designs will look like.
The new checkout experience was a labor of love. We were able to redesign the checkout and the one thing that I learned from this is that good design is viral. The business is now behind it because they love it.
Redesigning H&M’s checkout
I think my major challenge from a design perspective is just a matter of getting the new checkout out and being able to measure and iterate and understand where to go forward from this point.
That’s the short term. I try not to think too long term, but I think it’s essential to have a good organizational structure and to understand how UX can deliver and help every team within the business.
I like to get them involved in the process, sharing designs and UserZoom projects. I can involve them in things like writing questions for tests, as one of my key pillars is that UX should be a team sport.
I think the greatest success story right now is just actually having a UX solution integrated. We knew we needed to do more testing, but we just didn’t know that we could have the different types of testing. It’s now become a part of everybody’s daily routine.
I’d rather not have presentations and show numbers, saying “this is why you need UX, and this is why it’s important.” I’d rather demonstrate practically why it’s important and get people involved so they understand.
There’s a great slide that I always use in presentations about the ROI of design work, which says that, if a mistake will cost you $1 to fix at the design stage or $100 to fix it in code later, what figure would you pick? Wouldn’t you rather fix it in the design stage?
It’s about understanding that there’s a business that you’re supporting, so we need to make sure that we invest in the right things that actually generate some value for the business.
To me, I don’t see any difference between evidence based design and data-driven design. It’s just using the information you have at hand to be able to put the best design forward.
If you look at A/B testing that gives you some quantitative numbers. But there are other things that may impact the test, so you have to look at it more holistically.
There was a culture before of A/B testing everything, which produces a lot of variants, sometimes too many. This doesn’t always help, so now we look at the research platform first before we get to the A/B stage.
This allows us to be a little bit more confident in what we’re testing. So we have the strongest variants which can help us make the right decisions going forward. I find it interesting when we are able, together with A/B testing, to run usability tests with the variation links. Then we can also understand the whys behind the numbers.
I think for some designers, it’s been a big eye opener because they’ve never used any type of remote testing or usability testing. We can be more confident in the designs as we progress.
It can also mean fewer disagreements, as now we have a platform where we can test the ideas we have and move forward. This helps to change the conversation, helps us to be more collaborative and to understand each other’s points of view.
Right now for example, I’m hitting refresh on a test that has a different flow proposal from our tech team.
This also applies to what we’re trying to do in the checkout. So instead of us debating it and saying, well, no, this is what it should be, we can test the proposal and see what we can iterate from there.
It’s not about us trying to push an argument and get our way, we can just use testing to find out and maybe even discover that what we’re thinking is totally wrong.
It helps you be a little bit more humble, and I think the nice thing as well is it helps you detach a little bit from the process. Designers can get attached to projects that they’ve been putting a lot of work into, but this helps to get a different view.
I think design operations is the next big thing. I think we need teams that understand the work of designers in order to help facilitate the things they want to do.
I would like to see it go back to Nielsen Norman’s original definition of what UX is – understanding the full experience, not just the digital experience.
We have user experience designers, not just app or web designers, and understanding all the digital and non-digital touch points is important to the experience as a whole.