Project Overview

The Project

This design is part of a larger group study, aimed at designing a recommender system for grocery store apps, which combines user shopping history, seasonal, and celebratory context to provide users with personalized recommendations for upcoming shopping trips. Alone, I designed the application in 2 days to provide the interactive element to fulfil the algorithm. More information on the research itself is provided below. This is a complementary UI element to an already defined solution, thus, this is my take on a solution-first approach to UX design.

The Problem

Many grocery store recommenders lack personalised recommendations that withstand the change of weather and seasonal occasions.

The Goal

Create an interactive element which is easy to use and seek personalised recommendations.

Responsibilities

• Conducted a literature review to identify gaps in current grocery store recommender systems as a part of a group research project.
• Modeled a recommendation algorithm which implemented seasonal context on grocery store recommendations
• Participated in semi-structured interviews with grocery store app users
• Participated in creating user personas
• Participated in Ideation
• Create app maps in Miro
• Create user flow diagrams in Miro
• Create low-mid and high-fidelity prototypes in Figma

The Problem & The User

Literature Review

This project was based on a group project literature review, which adressed the following research questions. We use this research as a basis for understanding the user for this solution. These general research questions helped us gather insights into user needs, preferences, and expectations, which will be valuable in designing a user-friendly and effective UI element for the grocery store recommender algorithm.

1. How do users currently make decisions about what products to purchase when using a grocery store app?
2. What role do personalized product recommendations play in influencing users' purchasing decisions?
3. Are users generally satisfied with the recommendations they receive from grocery store apps, or do they feel there is room for improvement?
4. What types of information or data do users expect the app to consider when providing personalized recommendations (e.g., shopping history, seasonal preferences, dietary restrictions)?
5. What are the key factors that would make users trust and rely on personalized recommendations generated by the app's recommender system?

From this research, we found that users might:

• Benefit from pre-generated grocery lists based on recommendations
• Many users do purchase recommended items
• Many users benefit from recommendations to repurchase items, although some systems lack these recommendations
• Some users find recommendations irrelevent to their current context in regards to weather or seasonal events
• Users need recommendations that are based on their needs to rely on a system.

Pain Points

<aside> 🌡️ Lack of Recommendations based on seasonal context - Many recommender systems lack integration of the weather or periodic celebrations to provide personalized recommendations.

</aside>

<aside> 🎄 Lack of user input in recommendations - Users want control over which recommendations they receive, and want transparecy as to why.

</aside>

<aside> 🥛 Lack of replenishment recommendations - Not all users need new things! Some just need a reminder to repurchase milk than ran out.

</aside>

User Persona

Due to severe time restraints, and the solution-first approach of this project, I experimented and used a method of inputing the details of our system into a chat Generative Pre-Trained Transformer (Chat-GPT-3.5) to create a detailed user persona using existing internet data to determine who would you use it, how they would want to use it, and the challenges they may face.

Sarah (34, F)

Sarah is a busy marketing manager living in an urban area where the weather can be unpredictable. She often finds herself struggling with her grocery shopping, particularly when it comes to keeping track of items that need replenishing and planning for special occasions or holidays.

Goals:

1. Efficient Shopping: Sarah wants to complete her grocery shopping quickly and efficiently to save time for her work and personal life.
2. Item Replenishment: She aims to effortlessly replenish essential items in her household without having to think too much about what's running low.
3. Weather Consideration: Given the unpredictable weather in her area, she desires an app that can adapt recommendations based on the climate.

Challenges:

1. Limited Time: Sarah's demanding job often leaves her with little time for grocery shopping, and she finds the process time-consuming.
2. Forgetting Items: She tends to forget items that need replenishing until they run out.
3. Holiday Stress: Planning for holidays or seasonal changes can be stressful, especially with her busy schedule.
4. Weather Variability: The unpredictable weather adds an extra layer of complexity to her shopping, as it impacts her food choices.

Needs:

1. Automated Replenishment: Sarah needs a grocery store app that can automatically detect and remind her about items that need replenishing.
2. Holiday and Seasonal Suggestions: She would benefit from recommendations for holiday-related items and seasonal changes in her area.
3. Weather Integration: An app that takes into account the local weather forecast to suggest appropriate items and recipes would be helpful.
4. Time-Saving Features: Sarah needs features that make her shopping more efficient and reduce the mental load of planning.

Problem Statement

I need a grocery store delivery system that can recommend products accurately based on my preferences, dietary restrictions, and real-time inventory updates to streamline my weekly shopping

User Journey Map

Here's a user journey map for Sarah.

Information Architecture

High-Fidelity Wireframes

Since the digital wireframes, I made changes to better reflect the requirements of the user flow as defined by our project. For example:

• Additional on-boarding screen - gender and finished screen
• Detailed Item screens
• Recommendation based buttons

Overall, the high-fidelity frames were designed to reflect simplicity and intuitive design.