Freddy’s and Domo use Amazon SageMaker Autopilot to Compare 5X Larger Datasets

Domo is an AWS Advanced Technology Partner

Executive Summary

Freddy’s Frozen Custard & Steakburgers, a fast-casual restaurant chain headquartered in Kansas, turned to data science to find a better way to evaluate the quality of its restaurants. Working with Domo, an Amazon Web Services (AWS) Partner, Freddy’s was able to tap into the value of its data and seek predictive insights to go one step further. Freddy’s leveraged Domo AutoML, powered by Amazon SageMaker Autopilot, to help onboard and deploy machine learning (ML) models to predict staffing fluctuations and save money.

Inefficient Staffing Schedules Hurt Restaurant Revenues

Early on, Freddy’s began working with Domo to sort out its data, which required an evaluation of 18 different datasets spanning 100 columns created for different Freddy’s locations across multiple points in time. The project was so time intensive, that Freddy’s IT team had difficulty justifying a large ML investment as a next step in the company’s data strategy. Still, the team had a hypothesis in mind, which if fulfilled through ML, could smooth out staffing fluctuations. Through their work with Domo, the IT team realized that locations with high quarterly sales were successful because they had enough staff to sell the food. As a result, predicting sales volumes had the potential to also predict the number of team members needed at each restaurant, avoiding the headache of being over or under staffed. “At the time, there wasn't a tool that was accessible and allowed us to take everything that we'd done with Domo and build on it,” said Sean Thompson, IT Director at Freddy’s.

SageMaker Autopilot Enables Fast ML across 5X Larger Datasets

Using Domo AutoML, powered by SageMaker Autopilot, Freddy’s IT team was able to derive business value from ML in just weeks—something that took months previously. With ML tools at the ready, the team used 5X larger datasets for more accurate predictions, ramped up quickly with ML modeling, and proved their hypothesis around staffing. “Because Freddy’s is getting started with their data science practice, SageMaker Autopilot allowed them to follow ML best practices on which ideas to pursue while they further develop that expertise,” said Ben Ainscough, Head of AI and Data Science Products at Domo. “The integration between Domo and SageMaker Autopilot lowered the barrier of entry for ML, so they can quickly explore a variety of predictive opportunities that could be transformational to their business.”

“I like having a nice big toolset with Domo and Amazon SageMaker Autopilot, so that if something comes up where we need to crunch a bunch of data and find relationships, we’re able to predict anything out.”

- Sean Thompson, IT Director, Freddy’s Frozen Custard & Steakburgers

Accelerated Time to Insights by 2X

Domo worked with Freddy’s to combine data on staffing schedules and sales, then cleaned and processed the dataset with Domo’s Magic ETL that features drag-and-drop tools. Once the data was loaded into SageMaker Autopilot, the feature automatically created the best ML model, and Freddy’s was able to take it into production with just a few clicks so the team could start making accurate predictions. The Domo interface visualized predictions for easier analysis, and as Freddy’s data analysts were already familiar with the Domo platform, they were able to accelerate time to insights by 2X.

Smoothed Out Wavy Staffing Fluctuations and Boosted Sales

Based on early ML modeling results, the IT team has seen its wavy staffing fluctuations smooth out. “We were able to take data on how many people worked in a restaurant on any given week going back the last five years,” Thompson explained. “When we graphed it out, you could see waves going up and down. Using Autopilot, we can see that line out into the future and so far, it's looking good.” Thanks to predictive insights from ML modeling, Freddy’s has seen increased foot traffic in its restaurants and achieved double-digit same-store sales growth, despite the effects of COVID-19.

“Domo integration with Amazon SageMaker Autopilot helps our customers automatically create the best ML models and get to market faster.”

- Catherine Wong, Chief Product Officer and EVP of Engineering, Domo

Laid the Groundwork to Answer Tough Questions

With a data science strategy powered by Domo, and the automatic creation of ML models with SageMaker Autopilot, Freddy’s now has a firm footing to delve into new predictive analytics. Specifically, Thompson envisions using ML to support research and development and run A/B testing to determine the best indicators of restaurant quality. “I don't know what’s going to land in my lap next,” Thompson said. “And that’s why I like having a nice big toolset with Domo and SageMaker Autopilot, so that if something comes up where we need to crunch a bunch of data and find relationships, we’re able to predict anything out.”

“The integration between Domo and SageMaker lowered the barrier of entry for ML so Freddy’s can quickly explore a variety of predictive opportunities faster than before.”

- Ben Ainscough, Head of AI and Data Science Product, Domo


About Freddy’s Frozen Custard & Steakburgers

Freddy’s Frozen Custard & Steakburgers is a fast-casual restaurant chain based in Kansas with nearly 400 locations nationwide.
Freddy’s had begun a data science journey with Domo, but needed a fast, cost-effective way to create ML models to make accurate predictions.
Freddy’s and Domo leveraged Amazon SageMaker Autopilot to quickly deploy ML models that more accurately predicted staffing levels and accelerated time to insights by 2X.
  • Identified optimal staffing levels
  • Used 5X larger datasets for accurate predictions
  • Accelerated time to insights by 2X
  • Increased foot traffic in restaurants YoY
  • Achieved double-digit company sales growth

About Domo

Domo is a cloud-native platform that combines iPaaS capabilities for data integration, visualizations, and analytics for real-time and predictive insights. Domo also provides a foundation to compose your own apps to modernize business processes and take action on those insights, while minimizing the burden on IT.

Published December 2020