Finding recipe with the use of filters and AI
Efficient meal prep recipe finding
Project description
Planning meals for the week and cooking and storing them all in advance is a growing trend to save time and money. However, many people who start meal prepping eventually discontinue it, common reasons for this include finding it too time-consuming and inconvenient. Food intake is non-negotiable to be part of the most important contributor to health. With a personal interest to meal prep and the presented behaviour of people discontinue to meal prep, I embarked on a journey to explore the reasons behind this discouragement, as well as finding ways that I can encourage the meal prep behaviour more consistently.
Research
To better understand what other existing meal prep platforms are doing to retain customers. Competitive analysis is conducted to access their strengths and weaknesses.
Consider the user interview’s insights, I am narrowing the focus and flipping the user problems into opportunities for my product. From there, I ask:
How might we use advanced features to generate recipes based on users’ preference?
Define
In order to narrow down the scope of solution, affinity map is used to visually organize users’ perspectives during user interviews. A persona presented below furtherpresent the user problem Buzz Prep focuses on solving.
Solution
With user’s preference, needs, and challenges to meal prep in mind. I conducted the card sorting exercise to users in order to better understand the information
architecture of the design. With all information to be considered, the product is first ideated by the sitemap framework, constructive user flow, and the
low-fidelity wireframes.
Low-fidelity wireframes
I focused on creating low-fidelity wireframes that showcase the use of various filters for user to customize their search results. The five filters include user’s available cooking time, preferred cuisine, dietary preferences, number of serving, and available cooking equipment. The use of AI to generate personalized recipe is similar to a chatbot design, where users can make conversation and be guided to describe other preferences that have not already been mentioned, or they can put special request to have their ideal meal be generated.
Hi-fidelity Prototype
Usability Testing
I conducted 5 usability testings next to evaluate Buzz Prep's ease of use and user experience with it. I observed if users can set personal preferences and apply various filters, as well as being aware of the AI chat usage, and other navigation points like the grocery list.
Result
With the positive note getting from the usability testings, I later also incorporated the users’ opinions and iterated on some of the filter’s interface for instance the button design changed to checkbox design for multiple selection purpose. In addition, I also adopted note to ease the in-app navigation, including added direct access to grocery list from the recipe page. Check out the final prototype below.
