mycs is a Berlin-based company that sells high-end customizable furniture directly to European consumers. Our core product is the Furniture Configurator, where visitors can create and tweak designs to buy custom shelves, tables and wardrobes.
As the first UX hire, I kickstarted the user research process to find out about our customers, their attitudes on technology and purchasing behavior. I then led a full redesign of the Furniture Configurator web app, improving sales and engagement metrics. I am currently responsible for further usability improvements, designing interaction flows for new furniture types and adapting it to smaller form factors.
Who are the users?
When I arrived at mycs, there was little consensus on who our current users were. Stakeholders had different ideas about the age composition, preferred browsing devices or their attitudes towards technology. I set out to create characters based on data (personas) so the nascent Product Team and the company as a whole could relate and empathize with customers easily. To this end, I analyzed Google Analytics data and conducted semi-structured interviews with the direct sales team. There were a few key insights, including:
- most of our users were around age 45
- most of our users live with a partner and make decisions together
These findings made us adjust the company’s initial design assumption that the people who want our products would generally be looking for the best price possible (they prefer to be assured of our quality). We also came to expect longer times for visitor conversion, as we learned that purchasing furniture can be a tricky household matter.
What are the users struggling with?
Knowing that most people are not very good at performing complex computer tasks, we sought to uncover the usability issues affecting our configurator. To this end, we recruited a few participants for think-aloud tests.
With our initial configurator (see picture above), visitors:
- spent considerable time staring at the screen without interacting
- expressed confusion when changing a value immediately altered another (e.g. changing the width would change the number of columns)
- saw the icons on the right (size, color, boards, fronts and overview) as a toolbox, and frequently switched back and forth between these tools
Together with the Product Manager and the company’s co-founder, we brainstormed a configurator design from scratch based on what we knew were user needs and what we assummed were the causes of the problems observed in the tests. For example:
Observation: the user spends considerable time staring at the screen without interacting.
Assumption: he doesn’t know where to start.
- What if it had a conversational interface?
- What if the system made some changes for him?
- What if the configurator was made of sequential steps?
- What if we employed coachmarks for walking the user through his first visit?
I led several rounds of brainstorming, sketching proposals, making clickable prototypes and testing these.
Sometimes I wanted to simulate an interaction more complex than InVision prototypes would allow, so I quickly built demo iOS apps to test them out.
After a few weeks of iterating and working closely with the Engineering team, we shipped brand new configurators for four different furniture types, keeping an eye on weekly analytics. The redesign was a success, resulting in improved sales and better engagement (users made 34% more changes and saved 116% more products to buy later).
We are still struggling with a high bounce rate: some visitors are leaving the app without making a single change. One hypothesis is that people end up in the configurator before they are ready, so we are making revisions to our customer journey to better prime our valuable visitors.
You can visit the site below. Be sure to browse our tables and wardrobes, too.