How PepsiCo Achieved Data Quality at Scale with Monte Carlo

Behind any company with $79 billion in annual net revenue is a data team managing the volume and quality of their data.

As the largest food and beverage company in North America—and second-largest in the world—PepsiCo products are enjoyed more than one billion times a day in more than 200 countries. And each of those billion brand interactions creates opportunities to ingest and leverage new data. 

How PepsiCo uses Monte Carlo

With a product portfolio that includes Lay’s, Doritos, Cheetos, Gatorade, Pepsi-Cola, Mountain Dew, Quaker, and SodaStream—each generating more than $1 billion in estimated annual retail sales—PepsiCo has some big data quality needs. 

So, when PepsiCo wanted to improve data incident detection and reporting across its myriad of data pipelines and table volumes, the team turned to Monte Carlo for help. 

Talking about how PepsiCo utilizes data observability, Data Platforms Leader Sagar Saraiya said, “My team uses Monte Carlo to provide data quality information on our data assets to our engineering teams as well as our data users.” 

By leveraging Monte Carlo’s automated quality monitors, Sagar’s team is able to monitor and share insights on data health for the full breadth PepsiCo’s Snowflake tables right out of the box. 

“We try to track incidents on Monte Carlo, as well as getting the Slack alerts. So, we use Monte Carlo heavily for those activities.” 

Improving data reliability with field-level lineage and dbt support

Every team has something they love about Monte Carlo. And Sagar’s team has a couple favorites of their own. 

“My favorite Monte Carlo feature is not just one, but two. I love data lineage within Monte Carlo, as well as the dbt integration that provides data cataloging and searchability within Monte Carlo.

Data lineage paired with Monte Carlo’s incident reporting allows PepsiCo’s data team to root cause and resolve issues in real time. At the same time, Monte Carlo’s dbt integration extends observability into the transform layer of PepsiCo’s data platform, supercharging the team’s visibility into—and control over—their production pipelines. 

What’s next for PepsiCo’s data team

“We have a lot of things prepared for our roadmap down the line,” Sagar said. “We are planning to better define SLIs and SLOs using Monte Carlo.”

Sagar went on to share that the team is aiming to develop a holistic view of the data issues impacting PepsiCo’s data assets, as well as how the team can most efficiently tackle those issues using Monte Carlo.

“Our main goal is to drive data servability with data observability and that’s what we’re planning to do with Monte Carlo.”

Check out Monte Carlo and data observability for yourself