Yummly
Drink Pairing feature
Role: UX Designer
When: October 2016
Duration: Two Weeks
What: Drink Pairing Feature Addition to Existing Recipe App
Objective
Yummly is an application that allows users to find recipes with ease. Users begin by filling out taste profiles and dietary restrictions, and then browse through a curated list of recipes. The app allows users to choose a wide variety of parameters, such as vegetarian meals, allergen-free recipes, or preferences for sweet or savory dishes over sour. One pain point Yummly has not addressed is how users will eventually find a drink pairing for the dish they choose to create.
Mission
As Yummly’s objective is to make finding recipes easier for users, I aimed to add a feature that would simplify drink pairing as well.
Market
Yummly users ages 25-35 who regularly try to pair drinks with their meals.
Problem
Users are often creating new dishes that they are unfamiliar with, or they are not sure how to pair drinks. Yummly users need a way to find drink pairings for dishes they choose from the app.
Solution
Create a feature on the app that would allow users to receive suggested drink pairings for chosen dishes.
Research Methods
Research Takeaways
Through my research, I validated my biggest assumption – that users try to get drink recommendations – and learned users tend to purchase what they know in the most cost-effective way.
User behaviors did not come without pain points, though. The top complaints remained a lack of time and options, as well as getting recommendations that were too costly for them.
Personas
Using the insights from my research I was able to develop two personas of which to base my designs off of
Armed with research and the flow of the Yummly app, I tackled the design. I began usability testing with the paper prototype and was quickly met with a number of pain points.
I had assumed that users would focus on their drink choice, but not only did they want to be able to select a meal type in addition to a drink, but they wanted the entire feature to be meal-based. Users also asked for more specific recommendations.
I created wireframes that addressed the issues from the paper prototype and continued testing and iterating. After going through a number of iterations, I reached what I had defined as my MVP (minimum viable product).
The main flow can be seen below, and the clickable prototype can be viewed here:
Earlier Iterations
The first iteration included a chat feature that was removed due to a lack of use.
The second iteration brought users to a non-specific drink guide
After a number of iterations and rounds of testing, I came to the final version of the main flow. Users had a clear path from recipe to drink recommendations that addressed the needs that had come up during testing. Users were provided with multiple options, various price points, and specific drink pairing recommendations, allowing them to confidently find the right chardonnay to serve with their fish.
