Pizza Hut: Delivering pizza where and when customers want it

About Pizza Hut

Founded in 1958, Pizza Hut is an American quick-serve restaurant chain and international franchise. Pizza Hut operates more than 16,700 restaurants in over 100 countries and is a subsidiary of Yum! Brands, Inc.

Industries: Travel & Hospitality
Location: United States

Pizza Hut U.S. uses Google Cloud Platform, including Google Kubernetes Engine, to transform its ecommerce infrastructure and speed response time for digital customer orders.

Google Cloud Results

  • Reduces average API response time to requests by 10x
  • Doubles the number of Kubernetes nodes in seconds
  • Identifies chokepoints to improve time-consuming processes

Automates and simplifies infrastructure management

Since its first location opened in 1958, Pizza Hut has grown to more than 16,700 quick-serve restaurants around the world. But much has changed since the late 1950s. Increasingly, consumers want their pizza, and other restaurant meals, delivered or prepared for takeout. In fact, 50 percent of restaurant dinners are now consumed at home, according to The NPD Group.

To stay competitive, Pizza Hut is laser-focused on serving customers where and when they want to eat. “We want our pizza to arrive fast and be hot when it gets there, and we want customers to have easy, reliable experiences,” says Chuck Rhoades, Senior Manager of Digital Infrastructure for Pizza Hut U.S.

Technology is playing an ever-important role in the meal delivery experience. Digital orders now represent 53 percent of all restaurant delivery orders, up from 33 percent in 2013, according to The NPD Group. Accordingly, Pizza Hut is enabling its customers to order their favorite menu items via the company’s website, mobile app, Facebook, and Twitter chatbots, and even by pushing a button on an extremely limited, fun series of high-top shoes known as Pie Tops.

Just over a year ago, when Chuck joined Pizza Hut, its ecommerce operations primarily ran off a traditional bare-metal data center infrastructure. But it was clear that a digital transformation was needed. To serve its customers where and when they want, while remaining one step ahead of competitors, Pizza Hut needed to be agile. It needed to quickly leverage new technologies to build and scale new customer-facing services.

A public cloud container architecture supporting microservices would provide Pizza Hut engineers the flexibility to innovate new features and services as quickly as possible. For example, microservices and containers give Pizza Hut the flexibility to build out robust internal and external APIs to scale. Internal APIs simplify the process of developing enhancements to digital ordering, while external APIs will enable Pizza Hut partners and third parties to add new functionality to their apps via Pizza Hut APIs.

When evaluating container orchestration frameworks, Chuck and his colleagues quickly concluded that Kubernetes was the way to go. “Kubernetes was becoming the dominant orchestration technology, and it fit all the use cases we would be likely to come up with in the future,” he says.

Google Kubernetes Engine in Google Cloud Platform (GCP) was the ideal Kubernetes choice for Pizza Hut to manage environments for deploying containerized applications. “Google engineers originally developed Kubernetes, so I knew I could trust Google to make Kubernetes work well in a public cloud infrastructure,” says Chuck.

“In just a few weeks of launching our pizza tracker, we could see where operations took too long to complete. We could break out operations and drop them into a Cloud Pub/Sub queue to be handled asynchronously. As a result, we reduced average response time to the API by 10x.”

Chuck Rhoades, Senior Manager, Digital Infrastructure, Pizza Hut U.S.

Along with Google Kubernetes Engine, the Pizza Hut Digital Infrastructure team is using Cloud Datastore for managing web and mobile app databases; Cloud Endpoints for deploying, developing, protecting, and managing internal APIs and the Apigee API platform for public, higher-volume APIs; Stackdriver Trace and Stackdriver Logging to monitor traffic patterns and errors; Cloud Load Balancing to distribute compute resources; and Cloud Pub/Sub for real-time stream analytics.

Scaling for Super Bowl-sized spikes

The Pizza Hut Digital Infrastructure team has been migrating many of its back-end services into a microservice-oriented environment that Google Kubernetes Engine enables. At a high level, the purpose is to enable Pizza Hut to better serve customers where and when they want to eat, by leveraging the flexibility of the Kubernetes microservices architecture.

Location services are one example. When customers go online to order takeout from Pizza Hut, they’re served up information about nearby Pizza Hut locations from which to choose. Customers who’d rather get their meals delivered to them can type in their address and are automatically given the closest Pizza Hut that delivers to their address.

The process of locating the Pizza Hut locations closest to a customer is accomplished through localization services that the ecommerce team developed in Google Kubernetes Engine. “We developed localization services in Google Kubernetes Engine because of its ability to use resources efficiently, respond to traffic spikes, and scale very quickly using pod autoscaling,” says Chuck.

“For example, every year during the Super Bowl, our traffic gets a huge spike. And because it’s such an important time to meet customer demands, we want our service to run effectively and reliably. With Google Kubernetes Engine, we can monitor how much traffic each service is getting and scale up the infrastructure quickly to meet rising requests, then scale down to save infrastructure costs once demand slows down,” Chuck says.

Pizza Hut leverages the Horizontal Pod Autoscaler (HPA) in Kubernetes to monitor traffic across each virtual machine instance and scale the number of pods to handle the surge. “In other cloud environments, we’d have to get and configure additional servers and make sure they’re working properly to handle expected traffic spikes,” says Chuck. “With Google Kubernetes Engine clusters, it’s just a matter of typing a command to increase the nodes in the pool. In a few seconds, I can double the number of nodes.”

“With features like encryption at rest for all data and the ability to build out service clusters in completely private IP environments, GCP makes it easy to help improve security.”

Chuck Rhoades, Senior Manager, Digital Infrastructure, Pizza Hut U.S.

Tracking pizzas and gaining visibility

In May 2017, Pizza Hut launched the Pizza Hut Delivery Tracker, a feature that lets customers track their pizza’s status via mobile app, website, text alerts, and, later this year, the Google Home virtual assistant. Cloud Functions and Dialogflow have been critical components to Pizza Hut’s Google Assistant efforts. “The zero-ops scalability of Cloud Functions is a perfect fit for the logic layer, and Dialogflow makes creating intuitive, highly engaging voice UIs very efficient and approachable,” Chuck says. GCP plays a critical part in enabling the service for customers, as well as in providing visibility into how quickly meals are being delivered.

Stackdriver Trace and Stackdriver Logging are core parts of the DNA in every service we build, such as our pizza tracker,” Chuck says. “Stackdriver Trace is an amazing tool for helping us identify points where a significant amount of processing time occurred and to diagnose what happened there. We’re using Stackdriver Logging heavily to watch for errors and unusual traffic patterns.”

In addition, Cloud Datastore and Cloud Endpoints give the Digital Infrastructure team a great deal of visibility into what’s happening overall with the pizza tracker service. “In just a few weeks of launching our pizza tracker, we could see where operations took too long to complete,” says Chuck. “We could break out operations and drop them into a Cloud Pub/Sub queue to be handled asynchronously. As a result, we reduced average response time to the API calls by 10x.”

Built for engineers

“GCP is extremely developer-friendly,” says Chuck. “Everything from the console and the documentation to the way the portals and tools are laid out makes it easy for us to experiment and start using services quickly. I don’t have to wade through 90 pages of marketing to find what I need. GCP puts it right where I’d expect to find it.”

At the same time, GCP enables developers to focus on developing features and services versus managing an infrastructure. “One of the big benefits of Google Kubernetes Engine is the upgrade process,” Chuck says, as an example. “It’s been flawless. All I have to do is push a button, wait a little while, and everything’s upgraded.”

“With GCP, we have an infrastructure that helps us solve problems as they appear and take advantage of opportunities as they arise. By enabling us to continue building more microservices in a containerized architecture, GCP will play a key role in everything we do moving forward.”

Chuck Rhoades, Senior Manager, Digital Infrastructure, Pizza Hut U.S.

Paying as you go, for what you use

The GCP flexible pricing model has been another big benefit for Pizza Hut. The Digital Infrastructure team only pays for the services and infrastructure it uses, without being locked into a long-term contract based on a prediction of what it will need in the future. “Some cloud providers make you predict the scale you’ll need three years down the road, in order to lock in a decent rate. But scale is so hard to predict,” says Chuck. “GCP has a great pricing calculator that helps me predict what we’ll need now and in the future. Plus, the pay-as-you-go pricing and sustained-use discounts make so much sense for us. I appreciate that the pricing adapts to how much we use the platform.”

Enhanced security and support

“The support we’ve received from Google has been outstanding,” says Chuck. For example, a Google Cloud engineer worked with the Pizza Hut team throughout the most recent Super Bowl, helping to monitor system performance and traffic to help ensure there were no disruptions of customer ordering experiences during Pizza Hut’s busiest time of year. “Google gives us a level of support that goes beyond what is expected,” Chuck says.

Enhanced security is another area where GCP shines, Chuck adds. “We have enough things to worry about, so it’s a huge benefit to know GCP is a highly secure platform,” he says. “With features like encryption at rest for all data and the ability to build out service clusters in completely private IP environments, GCP makes it easy to improve security.”

A key role going forward

The Pizza Hut Digital Infrastructure team’s move to GCP was, from the beginning, about building an infrastructure for the future. “With GCP, we have an infrastructure that helps us solve problems as they appear and take advantage of opportunities as they arise,” says Chuck. “By enabling us to continue building more microservices in a containerized architecture, GCP will play a key role in everything we do moving forward.”

About Pizza Hut

Founded in 1958, Pizza Hut is an American quick-serve restaurant chain and international franchise. Pizza Hut operates more than 16,700 restaurants in over 100 countries and is a subsidiary of Yum! Brands, Inc.

Industries: Travel & Hospitality
Location: United States