I set out to conduct an extensive and on-going research programme to support design and business decisions. By making more informed decisions we could learn and prosper faster than simply competing against luck.
I interviewed over 30 people from a cross-section of personas. These interviews would often be conducted with the sweetener of an included lunch of their choosing in the popular district of Shoreditch where competition in the hospitality sector is fierce.
It allowed me to get a glimpse into people's rationale for ordering food, taking into consideration the bias of the context.
I started with customer development interviews, trying to understand a users job to be done and continued utilising interview processes right up to usability testing on the production site.
I was able to employ various methods from my UX interview playbook including five whys, visual card sorting and monopoly value prioritisation.
Research narrowed in scope as we developed and refined our opportunity backlog.
This design thinking model helped us explore various ideas and then prioritise those which garnered the most traction in the interview invalidation process.
One of the card sorting methods I employed I call visual card sorting. I created various sketches of content that could be displayed about a dish, laminated and cut out each one.
Armed with a stack of cards I then asked users to pick up to 10 items to create a dish information card. After this, I asked the user to remove all but the essential items to see what was left.
I tallied up the results of both to help give me some quantifiable evidence to support the inclusion of some data points over others.
Once the first two weeks of user testing had passed I set about creating personas based on the people I had met. These helped me create useful stereotypes that reflected these users personalities, tastes, objectives and obstacles. I created industry-specific measures that helped me communicate how these users would differ from one another when making decisions.
These niche measures captured eating related traits like a person's group dynamic (were they leaders or followers), their appetite for risk (pun intended) and other behaviours identified as poignant to meal selection.
I also grouped the personas on the wall by the key attributes. This allowed us to see if we targeted a specific behaviour which personas would we serve best as a collective.
As I met with more users and conducted other research, I was able to backfill the persona attributes with data and research to support them. This helped us discover the addressable market and validate our assumptions with academic research papers.
Sketching is my go-to way of getting thoughts out of my head and onto screen/paper. They are usually very rough and usually only legible to me in the very first instance. But as an idea begins to take shape so too do my sketches.
I never claim to have the best drawing skills, I only need enough to convey meaning, first to myself and then to my peers. Once everyone has a shared understanding they typically get put to one side and we move on.
Not only is it quick to knock out ideas like this, but it also gives others the confidence to join in. You don't need to be fluent in drawing to sketch out ideas and it's important people don't feel obliged to draw a piece of art each time they want to communicate their idea.
Wireframing isn't a process I use all that often, with design systems I can go from sketch to prototype quickly without the need for these layout UIs.
But in this particular case, I didn't have a design system and I was working alongside a UI designer. I used a very stripped back minimal UI to hand over to the designer. It gave them the remit to design the visual treatments with as much creative freedom as possible all the way down to the microcopy.
Through extensive research, we uncovered some interesting findings that were relevant to our product development.
We observed users would describe a restaurant location in time not distance. We then started framing questions about distance to test the observation. I found that short distances would be described in feet then longer distances were almost exclusively measure in time to the destination.
The time was usually qualified by a mode of transport. We used this insight to show users times instead of distances in a bid to reduce their cognitive load.
Short distances were expressed in walks and longer distances in drives using an average speed calculation, later we could have considered using the navigation to get better estimates of time.
We discovered that users enjoyed salivating over big menus, while they caused friction and even anxiety in the decision-making process it was undeniable that people genuinely enjoyed imagining each dish.
Most users expressed the core job to be done was not eating but to spend time with friends or family. Eating is simply a backdrop for this activity.
We approached the problem of getting interesting meals to users with various unique ideas. One of which was the idea of creating suspense by introducing risk.
The concept came from restaurants in Japan that serve potentially lethal blowfish. People will flock to these eateries were expert chefs prepare the blowfish in such a way it is safe to consume.
Adding risk elevates the experience and provides a variable reward for the user who participates.
For businesses, it's an opportunity to promote a single deal against others in the locality and those can be high margin meal combinations.
Users were split on how they felt on reflection but at the same time demonstrated the kind of hooked behaviour we were looking for.