February 5, 2023


The Joy Of Businnes

Yum! Brands’ secret Domo sauce: Jupyter Workspaces

Because the COVID era began and prevented people for a lengthy period of time from dining in at restaurants, people everywhere you go have increasingly relied on restaurant ordering and shipping and delivery applications to place food stuff on the desk for themselves and their family members.

To handle the shake-up in food-intake dynamics, Yum! Brands’ digital and engineering groups invested noticeably in the progress or enhancement of this sort of apps for our dining establishments, which include KFC, Pizza Hut, Taco Bell, and The Behavior Burger Grill.

For KFC-United States in particular, the principle of owning a restaurant buying app was rather new. To motivate KFC shoppers to obtain and use the application, we needed to assure that it was “relevant, simple, and distinctive”—or, Purple, as our preceding CEO, Greg Creed, liked to say.

But to genuinely make certain that it was Red, we wanted metrics. We wanted to know if the application was indeed making the process of buying fried hen much easier. Have been folks happy with the app? Were being there recurring styles among the prospects who loved the app (or didn’t appreciate the application)? Did specific application release versions perform improved than some others?

All those were between the questions we had to locate solutions to. Even though both Apple and Android present accessibility to buyer ratings and opinions, they do not offer a deep dive into what reviews indicate for a merchandise. So, we turned to Domo, and the tool that has turn into our top secret sauce: Jupyter Workspaces.

Jupyter Workspaces presents us the potential to entry and analyze this qualitative data. In my encounter with other small business intelligence platforms, text assessment has been minimal to phrase counts and word clouds.

Sample of a Domo/Jupyter Notebook project done on Doordash Testimonials

Jupyter Workspaces, on the other hand, normally takes text evaluation to the next level, making it possible for practitioners to mix Python’s highly developed Normal Language Processing (NLP) capabilities with datasets correct inside of Domo. It also permits Jupyter Notebooks to be scheduled as DataFlows to instantly refresh your facts. By employing Python and Domo in tandem, KFC can now do the next:

Python Domo
Import consumer critiques instantly from Apple and Android merchants and blend them into a solitary dataset Schedule the Jupyter Notebook to immediately refresh every day
Use Normal Language Processing types to discover the customer’s emotion toward the app in every overview Make a dataset that can be shared across the corporation
Extract vital metrics this sort of as when the critique was composed and the user’s star-level ranking Illustrate benefits and metrics in a charming way, working with company branding and interactive visuals

All of these capabilities lead to deriving insights for KFC’s cellular application workforce. Now, the staff can detect what works for prospects and what does not, and cultivate concepts for potential app improvements—which all goes to clearly show that when KFC consumers speak, we pay attention. And that, of training course, is vital to long-term brand name and solution success.