Use Amplitude and Snowflake to Drive Million-Dollar Revenue Increases in the QSR Space

Insights/Action/Outcome: RBI wanted to determine whether there was a connection between app startup speed and customer spend. Through Amplitude, they saw that people where more likely to place an order when the app started faster. Now, the team has improved the app load time and increased revenue.

You can do a lot with data, provided it’s easy to access and manipulate. All the data in the world doesn’t do a business much good if it sits in a data warehouse and is only available to data analysts and people who can author SQL queries.

That’s why data success stories don’t start with the quantity of information. The heart of these stories involves tools that allow teams to trust and interact with data, giving it the gravity to transform your business.

Data success stories don’t start with the quantity of information. The true transformation takes place after adopting tools that allow teams to interact with data, giving it the gravity to transform your business.

I spent a decade on infrastructure, working on Google Cloud Platform on a variety of products including Google BigQuery. I was one of the people who worked on the tools data analysts use every day. It was challenging and rewarding work, but I wanted to try something different, so I joined RBI in 2021. You may not have heard of our company, but you know our brands. Restaurant Brands International is the parent company of Burger King, Popeyes, Tim Hortons, and Firehouse Subs. It is one of the world’s largest quick service restaurant companies with more than $35 billion in annual system-wide sales and over 29,000 restaurants in more than 100 countries.

My job as Head of Data Analytics is to make our data useful for our internal teams and franchisees. I lead a group of engineers, data analysts, and contractors. We manage and deploy analytics tools for the company and our brands. We also build dashboards for our internal use and our brands. We tailor the latter to the restaurant business while making them general enough that any of our companies can adapt them to their needs.

The best part of my work at RBI is that our product isn’t technology. Each of our brands has an internal data analytics team, but my team builds the data platform for the entire company. Every system we deploy and every dashboard we write has an impact, directly affecting the bottom line for our franchisees.

Snowflake and Amplitude: a dynamic duo

My first big project at RBI was integrating Snowflake into our existing tech stack, which included Amplitude Analytics.

Our analysts and product teams leveraged Amplitude to gather behavioral intelligence. Usage ran the spectrum from people who wanted to monitor the performance of a new digital offer to a product manager on the growth team who wondered whether a new app feature led to more conversions.

In one instance, the team was planning changes to improve the app load time. We wanted to determine whether there was a connection between startup speed and customer spend. An Analytics graph confirmed our hypothesis—we saw that, yes, when the app started faster, people spent more money with us. By decreasing app loading speed by 43% on Android devices and 16% on iOS, we saw conversions increase by 4%. It confirmed that investing in this initiative was a move in the right direction.

As helpful as Amplitude was, we were limited to the data that it was ingesting from front-end applications, like our mobile apps . There was no way to correlate this digital behavioral data with business data from many other sources. For our backends, we used AWS DynamoDB but couldn’t query our database efficiently to get the right answers. This setup prevented us from asking some of the most fundamental questions for our business. For example, our analysts couldn’t measure the impact of store schedules on sales because our locations’ hours weren’t stored in Amplitude.

We needed a way to route data to Amplitude, which meant rethinking where and how we stored our data. In October of 2021, we started building a Snowflake data warehouse. We chose Snowflake because it was fast, cloud-based, and could integrate seamlessly with our other tools, including AWS DynamoDB and Amplitude. We routed some of our key data sources such as order records from our backends and store metadata to Snowflake immediately.

Snowflake stood out for its ability to integrate and analyze massive data sets. It allows us to clean and store our data in a format we can use in all our analytics and BI platforms. For example, we route order data from AWS DynamoDB to Snowflake and make it available in Amplitude. In the past, Amplitude would have had to convert this data into a format it could use for behavioral analytics, creating a different data set from the original that might lead to discrepancies. With Snowflake, every application receives the same data, so we can be confident that what we see in Amplitude is consistent and reliable.

Insights for everyone and SQL for those who need it

Snowflake is a powerhouse, but we didn’t need many high-touch, in-person engagements during deployment. Instead, we created documentation and made it available to our people, who are savvy enough to read it and figure things out independently. Before we knew it, they started finding their own use cases for Snowflake and Amplitude.

By bringing together Amplitude and Snowflake, we leveraged the best of both platforms. The combination of Snowflake’s clean, consolidated data and the dashboards of Amplitude Analytics empower our teams to ask meaningful questions and extract valuable information from our datasets. Both platforms are easy to use, too. Whether or not they’re a data scientist, anyone in the company can stitch together pieces of information from different analytics tools and systems, reconciling various data sources to craft clear and comprehensive stories about our business.

Layering data tools enables anyone in the company to stitch together pieces of information from various data sources and craft clear and comprehensive stories about the business.

One of the most impactful projects enabled by Snowflake and Analytics is our Suggestive Sell Engine (SSE). SSE is a machine learning (ML) application that uses historical purchase data from our mobile app residing in Analytics to build models that suggest additional purchases. For example, if you order a Whopper on the Burger King app, it’ll ask if you also want fries or whatever side people typically order with a Whopper.

It works on our mobile apps and websites and powers the interactive outdoor digital menu boards at our drive-throughs to offer suggestions when customers place their orders. Our SSE has generated increased revenue for different brands. In some markets, we also have self-serve kiosks at our restaurants that also make suggestions using this technology.

RBI also leveraged Amplitude and Snowflake to build rInsights. This franchisee-facing tool uses Amplitude data sets from our loyalty backend servers to help franchise owners better understand the behavior of returning customers. Franchisees can make better operational decisions by tracking what causes people to return to their stores and what keeps them away. If we wanted to develop this platform without Amplitude and Snowflake, we would’ve had to build infrastructure to connect our servers to the app. Instead, we sent all the data to Amplitude and then created a dashboard on top of that. This simplified approach has enabled our franchisees to know more about their customers and run their stores accordingly.

Linking the two platforms is now easier than ever. Amplitude’s new datashare integration to Snowflake allows us to use Amplitude data without leaving Snowflake, further simplifying our workflows and helping our engineers move faster in an environment where they are most productive.

Data offers a competitive edge in a crowded space

Incorporating our data sources into Snowflake is an ongoing process, but the benefits have made a big difference in our organization. From data analysts to product managers throughout our brands, people see the possibilities data offers and are asking more questions than before. The insights they’re gleaning by leveraging Snowflake data and Analytics dashboards help us develop new tools like our ML project and franchisee-facing tools that drive revenue at RBI. The more visibility they have into our data, the deeper they want to dig and the more they want to do.

The most valuable analytics tools offer the ability to harness data, turn it into something that’s easily understood, and drive better business decisions.

Data is not a magic bullet. Too much data can bog you down, especially if you don’t know how to harness and transform it into useful information. Amplitude and Snowflake have given RBI the ability to harness our data, turn it into something that’s easily understood, and drive better business decisions. It enables us to stay competitive in the fast-paced food services industry and support our franchisees as we adapt to changing tastes and find new ways to serve our customers within our app, online, and at our retail locations.