Discover the winners in our competitive UX measurement report
No internal design or development team? No problem! The user experience can still be managed and it can still come first with the right planning, tact and tools.
Recently we chatted to Maximilian Speicher, Ecommerce UX Manager at C&A about the intricacies of running lean user research and testing at the European ‘fast-fashion’ retailer C&A, as well as how the value of UX is communicated to upper management and how the current lockdown has affected the company’s digital business.
You can also read more from Maximilian, including his academic work, in his blog: maxspeicher.com
I studied computer science at university and it was around this time that I really became aware of user experience (or human computer interaction as it’s known in academia). I started working at a research institute at university and soon after this I started to specialize in user experience design.
I did a PhD in co-operation with one of the largest ecommerce providers in Germany that operated travel and hotel booking websites, and my topic was the automatic usability evaluation of search engines. I tried to find a human-centered design approach to automatically evaluating search engines and so moved away from software engineering to user experience.
After working with an augmented reality start-up and doing a post-doctorate at the University of Michigan, researching and teaching Interaction Design, I ended up in this user experience position with C&A. I saw it as an opportunity to build something from the ground up.
Start-up work is much more quick and dirty because you don’t really have a lot of money for everything. Things move a lot quicker, you really try to make more use of data, and try all the UX approaches you can. Often you don’t have time to do things very thoroughly for very good reasons, mostly time and money.
Now at a larger company, it’s pretty cool. I have the resources to do what I need to do, and also to do it more thoroughly. It’s still very exploratory but you also get to validate a lot more.
I see UX research as a pipeline where you put in ideas and you get insights out the other end. That’s basically what I wanted to implement here. I’m working in the ecommerce department and my vision is to create that process.
There’s a whole lot of different methods to use – contextual inquiry, A/B testing, in-house user research, field studies etc. I need to focus on a few that work for our industry so I’ve split the pipeline very roughly into quantitative and qualitative.
We look at what the research question is and then select the appropriate method. We do some remote studies that are rather quantitative, but 90% of the quantitative research is A/B testing. And the majority of qualitative research runs via UserZoom.
Very often the research we do generates new ideas and new research questions that we can then feed back into the pipeline. I would say that most of what I intended is working well for us so far.
“I see UX research as a pipeline where you put in ideas and you get insights out the other end.”
There was A/B testing going on already, though mainly through an external agency and looked after by product managers, not by dedicated UX managers. In general the volume of testing and UX work was much smaller.
My team-lead knows of the value of UX because she’s a very UX-savvy person herself. I joined C&A because she felt there was a need for someone to take care of UX. We have now grown into a real UX team with several UX managers, embedded into a larger product management team.
We want to involve the rest of the ecommerce department in our research and work, with everyone contributing ideas and being aware of the UX process. People submit their ideas because they know they gain valuable insights from it.
The amount of input and number of ideas generated is one measure we use to communicate success and convince people of the value of UX.
The other thing is regular benchmarks – we benchmark ourselves against other shops and look at quantitative measures using metrics like SUS and NPS. This way you can see that all the UX work you’re doing actually has an impact.
“People submit their ideas because they know they gain valuable insights from it.”
UX plays quite a big role there. People just know that it’s valuable to test as much as possible because most software development and software design is based on hypotheses.
Some assume that it might be better to put a button here or there, and to give it a certain color but it’s always better to test these kinds of things. All of these involved parties, the stakeholders and the agencies have started to frequently ask for tests and research. UX plays a huge role there because everybody understands the value that we can add.
It also comes down to proper communication tools and having regular meetings where you can talk about the latest research results, the current backlog of ideas, and prioritize and select what should be tested.
Luckily, we don’t have a lot of disagreements really. People trust the results that we deliver which is a good thing. We always try to just solve things with facts, hard numbers and insights.
They’re different stages of the process. A/B tests always happen on the live website and look at everything that can be applied quickly and easily. That would include things like changing the color of a button or the sort-order on a category page. We need huge numbers to find something statistically significant so they need to be tested on live customers.
The UX research comes earlier on in the process when you have designs that have not been implemented yet, such as testing a completely new design for a category page (of course this can still happen with live features on a website).
In this case, an agency might deliver a few different designs for a new category page which can be shown to users or potential customers to see how they feel about these different designs. Once you have picked a design in this UX research phase and it is implemented on the website, you can still test little tweaks and bits and pieces using A/B tests.
It’s also qualitative versus quantitative because A/B tests can never give you qualitative insights. They only tell you what’s happening, but not why it’s happening. So this is why we often do more qualitative follow-up studies to A/B tests.
We’re doing things more quickly now.
I think we can also extend this beyond ecommerce because I feel that user experience is not only for digital, but for a fashion retailer it can be more useful in general. You can do a lot of things that apply to your offline stores from user experience practices.
We’re all sitting at home now and we have to collaborate remotely. Apart from that, nothing has changed. We’re not losing any speed and are still running the same number of studies. The only difference really is we’re not in the same office building anymore.
It comes down to proper communication platforms, efficient meetings and just communicating on a very regular basis. If you keep that up, working remotely isn’t really a problem.
I don’t know anything about the perception yet. The obvious effect right now for a traditional retailer is that we had to close all of our offline stores in Germany, so online is our only channel right now.
That makes us a very important part of the company now, and we’ve only started looking at customer feedback from last week but customers seem to be really happy that they still have an online store where they can buy clothes and that is also being reflected in our NPS score.
It has been extremely useful. Before I started at C&A there was no regular remote online testing so qualitative insights only happened during rare in-house user labs.
In-house tests can be extremely costly and you have to put a lot of time and effort in there. So if you have questions that you can just as well test remotely, then this is a much quicker process. I should also explicitly thank one of my colleagues, Johanna, since she handles the vast majority of remote research. We absolutely couldn’t do what we’re doing without her.
Remote research has allowed us to answer more qualitative questions much more quickly and earlier in the process. We now get quicker qualitative insights that you don’t need traditional user labs for.
It’s also proven to be very complimentary to A/B testing. A/B tests, whether successful or not, can trigger a whole lot of different research questions so we can do some follow up studies just to quickly evaluate the qualitative side and get a more complete picture of the whole thing.
“Remote research has allowed us to answer more qualitative questions much more quickly and earlier in the process. We now get quicker qualitative insights that you don’t need traditional user labs for.”
Artificial intelligence. Not only automated evaluation of design, but also automating the design itself so website designs that are made by machines. This I believe, is a huge thing to look for.
One reason for this is that artificial intelligence is super interesting. So you want to work with it and then see what you can use it for.
Second of all, user experience research and design is still pretty expensive for companies. It’s a huge investment and if you can automate things there in terms of evaluation and design you would want to look into that.
For A/B testing it’s pretty easy. You can have an uplift in a very specific KPI and you can just compute the additional revenue it will most probably generate.
Through the number of ideas, we get the number of research questions that people want us to answer and we’re also doing very regular presentations for the whole department where we describe our approaches and show them what we’ve done.
We always ask people what will be done with the results so that we can prove that we’re not just doing research for the sake of it or just because this is an interesting question.
We want our results to be actionable and to be able to pinpoint a very specific, concrete thing that you do based on the results we deliver.
We can also compare the user experience of our online shops to a number of competitors and track the user experience of our website over time. Then we can actually see if we’re improving.
Onsite customer feedback is another way, and using SUS and NPS to track customer satisfaction.