In 24 hours, our team created "Rescheduling Detection" to help managers of large restaurants save time when making schedules.
Team Members
Software Engineers: Tony Wu, Sarah Reeve
Data Scientist: Jay Kim
UX Designers: Eeman Fatima, Sum Liu, Vivian Cheung
Project type
24 hour Hackathon
UX/UI, Research, Ideating, Wireframing, Prototyping, Presentation Deck and Speaker
Figma, FigJam, Sous Chef Design System, React.js, SASS
Project Overview
Our client, 7shifts is an all-in-one team management platform for restaurants. To date, they've helped over 40000 restaurants save costs on scheduling, managing tips and overtime pay. They also provide tools for all stages of the employee lifecycle, such as hiring, paying, and training them. As 7shifts grow and attract larger restaurants to their platform, they are looking ahead to scale scheduling effectively for these businesses.
The design challenge

How might we help restaurant managers schedules shifts more efficiently?

The Solution
Our solution will highlight shifts that are likely to be swapped out using past schedule data, including an employee's attendance, which days they are prone to swap shifts, weather data, holidays and more.
Our Process
The tight time constraints meant that we had to be strategic in our approach. Our team of UX Designers, Software Engineers and one Data Scientist often split off to do their own research and execute tasks, before converging again to make sure we were aligned at all times. This was the basic roadmap of our process:
24 hours remaining: EMPATHIZING WITH THE USER

What are Restaurant Managers' Pain Points?

The very first step was to understand the clients' ask, their product, and their customers. With a whiteboard, we discussed and wrote out the Problem(s), the User, the Pain Points, Current User Journey, and Constraints.

I kicked things off with a whiteboard discussion to get everyone on the same page.
Realizing we had a lot of gaps in our knowledge, we split off to conduct secondary research. We also downloaded the app and created schedules from the manager's POV to understand their current experience.


The number of times a location republishes their schedule.


Of a manager's time spent scheduling.
1. From 7shifts 2. Source
We also learned about 7shifts' Shift Pool feature from someone who had used the app before at her job. Employees have the option to offer up or trade their shifts with others. The caveat was that managers always had to approve the shift change.
21 Hours Remaining: DEFINING THE PROBLEM

A Less-involved Manager

We discovered that shift swaps were taking up a lot of the Manager's time, because they had to manually approve every change, and republish the schedule each time. As larger restaurants use 7shifts, it's important that we nip this issue in the bud.
The bigger the restaurant = the more shift swaps, and the messier things get for the manager.

Meet our Proto-Persona

Due to time constraints, we couldn't conduct user interviews with restaurant managers which would've informed a validated persona. Therefore, we created Nolan, a simplified persona based on our assumptions and existing data.

Predict and Pre-approve Backups

After deciding on our problem, our team started throwing out solutions. And even after dot-voting, the chosen solution didn't satisfy all the user's needs. We were closing in on midnight and there was not much for the software engineers to start on.

In the end, we decided to combine a few different ideas, each satisfying a different need.
Our SOlution

Predict shifts that are likely to be swapped, using past schedule data, then give the manager the choice to preapprove suitable backup employees.

17 HOURS REMAINING: Prototyping (and Coding)

Rescheduling Detection

The UX team first created the task flow for this feature, which allowed the developers to have a clear idea what to build, and they can also tell us whether this was a viable flow for the time we had left.

Modal Explorations

We decided to build on the modal that 7shifts already uses for their scheduling flow. This allowed us to wireframe directly in high-fidelity. Here were the different layouts explored...

The Rescheduling Detection Flow (In Hi-Fi)

6 Hours Remaining: Testing and final Touches

The Manager's New Journey

In the last remaining hours of our hackathon, the team went back and forth on whether this solution was user-friendly and satisfied the user's needs. I decided to draw a storyboard illustrating how our solution can save time for the manager.
1. Employees received published scehdule.
2. Tony puts his shift up for grabs.
3. Amy grabs his shift.
4. Since Amy is pre-approved, the trade is automatically completed.


Our talented Software Engineers, Tony and Sarah, delivered a seamless MVP that matched our hifi prototype. Our final MVP has tooltips to inform the user how the system detects shift swaps and how it suggests dependable backfills to the manager.
Key Learnings

1. Tooltips can be a crutch.

At the time, we needed a quick and easy way to explain our feature. Implementing a tooltip required the user to do more reading and comprehension. Looking back, I wish there was more time to simplify the solution so that it didn't require tooltips.

2. Think more broadly.

When we were given the assignment to improve scheduling, we didn't have time to consider other parts of the entire team management lifecycle, such as onboarding, hiring, training, etc. All of these other stages affect scheduling one way or another, which meant we shouldn't have boxed ourselves into the scheduling part of the cycle.

3. Have fun and get some sleep!

Our solution admittedly could use lots of improvements, but ultimately, our team had tons of fun collaborating and creating something we all believed in. Not bad for our first 24hr sprint!

Thanks for reading!