How might we help foodies, people that eat out over 3 times a week, improve their experience eating out?
As somebody who grew up around a restaurant, it's not surprising that I grew to love and appreciate the experience of eating out. So I began this project to sate my own curiosity to find out what problem I could help solve for people eating out.
My largest constraint was that there were already so many services for various different purposes in the food space like Yelp for discovery, MealPal for cheap meals, and OpenTable for reservations. It was a challenge to find a problem that has not already been solved for yet.
The project was a rollercoaster that had me have to step outside of my own perspective and really learn to listen to the voice of users to guide my design process.
Solve for the unmet needs of people that eat out frequently
UX Team of One
User Interviews, Personas, Affinity Diagram, Competitive Analysis, Card Sorting, User Flows, Sitemap, Wireframing, Prototyping
Designing to solve a problem users actually have, not one I thought they would have.
Honestly, when I discovered my hypothesis was completely wrong, all I could do was freeze and stare at the data.
I felt like a hitchhiker without a map.
Where did I go from here?
Just where did I go wrong with my research?
Well the answer as cliché as it sounds was listening to the user's voice
Listening to the voice of users led me to designing Pallate, a social food recommendation app.
When eating out at a new place, people often rely on reviews and pictures on Yelp or Google Maps. However, even with those tools, people were still uncertain whether or not the food will be as good as it appears in photos or reviews.
Palate is an app that helps urban workers and students more reliably discover new places to eat by creating a social platform where they can both ask for and give recommendations to friends rather than strangers
MY INITIAL HYPOTHESIS
"We believe that urban workers and students that eat out over 3 times a week may feel that they’re spending too much money on food. We can verify that when 60% of people interviewed express dissatisfaction with their food spending"
At the start of the research process, I originally began with the thought that with so many people eating out especially, they must be unhappy with how much money they spent on burrito bowls and sushi because I know my wallet definitely had problems with it.
I was surprised to learn my hypothesis was completely wrong since 61.4% of people were satisfied with how much they spent eating out.
I spend way too much!
I'm happy with how much I spent
Number of People Happy or Satisfied
After conducting 7 user interviews to hear about the stories, motivations, and process people went through when eating out, I took the learnings from those interviews to set up a survey which revealed to me that my hypothesis was completely wrong!
The results were jarring to me because I expected people to be more upset with how much money they spent eating out.
I'M NOT MY USER
After my initial shock passed, I realized that while I was hearing what users said, I wasn't listening whole heartedly.
When coming into the project, I thought my hypothesis would be verified and I had already started envisioning myself solving a problem in the personal finance and food space.
It was like a bucket of ice water had been poured on me, but it was incredibly humbling to me since I learned while I could relate to the users, I wasn't them.
So I took a step back and revisited the data by organizing it into an affinity diagram to better see what patterns appeared through the research.
SO WHO WAS OUR USER?
I felt like a hitchhiker without a map, but the map was right in front of me, the user's voice
Tommy, The Foodie
"When I get to try new foods, it's really exciting!"
Learn which restaurants are the best
Spice up everyday life through food
To bond with friends
Discover new places to eat
To share a meal with friends
Seeking new experiences through food
Unmet expectations when trying restaurants
Too much information on apps and websites
Reframing The Problem
How might we help Tommy more reliably have a positive experience when trying new places to eat?
Tommy was our primary persona and representation for Foodies so I redefined the problem that was being solved around what would best meet his needs and that of the secondary persona as well.
A major problem people like Tommy were facing when exploring was that they feared that the experience they had wouldn't live up to the hype of reviews and photos which prevented them from trying new places to eat at times.
TITLE OF THE CALLOUT BLOCK
Survey results came out of left field and showed that 92% of people said they used friends to find food
Design Driven by Data
50 out of 54 people said used friends to find places to eat so based off of that, I narrowed down the solution to a social recommendation app where Tommy could get recommendations from friends.
I had asked people in a survey about what ways they discovered new placed to eat and I expected apps like Yelp and Google Maps to be the most common answers, but users threw me for another curveball when ALMOST EVERY RESPONDENT mentioned friends and word of mouth as how they found new eats.
This oddly surprising detail led to me reshaping my design towards the idea of allowing friends to recommend food to other friends because the data had shown that there was overwhelmingly more trust involved there than in random reviews from strangers.
Designing a Social Recommendation App
What users taught me along the way
People were had trouble answering the question "What do you feel like eating?"
In the first iteration of the design, I had the impression that users were looking for a certain type of food like pizza or tacos, but when I tested the design, often people felt like they felt like the reason why they were here was to get a recommendation on what to eat, not just the best places.
I explored adding an AI component that would give compatibility rating based off of recommendations.
To attempt to provide users the freedom to explore when searching for new food, I implemented a discovery feature that would show restaurants similar to Yelp or Four Square only with an AI component that would give a compatibility rating based off of what friends recommended to the user.
The voice of users, steered me back to focusing on the main unique selling point:
Friends recommending food to friends to create a more trustworthy experience eating out.
When testing, users found the AI aspect rather confusing due to their impressions of the app being about sending recommendations to friends.
So based off the 5 usability tests I conducted, I scrapped the AI aspect and Yelp-like discovery to focus more on connecting friends together so users like Tommy could find a more trustable way to try new restaurants.
Furthermore, I generalized how users asked for food into categories like: breakfast lunch, dinner, and dessert because users had trouble picking a particular food they wanted to try in tests.
It's been an incredibly humbling experience to work on Pallate because the failures helped me learn to set aside my ego in design and follow the voice of users as a guiding light.
I ran into two major roadblocks on this project: my hypothesis being proven completely wrong by research and my initial prototype of Pallate being too unfocused on the unique selling point of the solution.
Throughout the project, I felt like I knew where I was heading only to be surprised at the direction that users pushed me towards.
The key to dealing with these roadblocks was to set aside what ego and investment I had in "my idea" and think on a larger scale on what would help users.
Users had not problems navigating the app in the most recent usability tests, but users expressed a possible impatience when waiting for recommendations.
In tests when users thought aloud, they felt that they would want the recommendations quickly because they assumed that they were hungry.
While we cannot know how fast friends respond to recommend places to friends without launching the product, it definitely could be an issue if notifications are turned off in the app, which definitely is likely.
So the main next steps are to implement a way for people to share general recommendations in their own friend groups on Pallate.