Pitney Bowes Simplifies Data Sharing with Snowflake and AWS

Executive Summary

Pitney Bowes, a provider of commerce enablement technologies, was using an on-premises data warehousing solution to store and manage data. The company’s goal of expanding its data analytics abilities was hampered by the limitations of its current data warehouse solution. Attempts to work around these limitations led to continued downtime and unallocated IT costs. Pitney Bowes needed a solution that could scale with its needs and support widespread data sharing—while improving management of data and related IT costs. After comparing enterprise data warehouse solutions, Pitney Bowes deployed a Snowflake solution built on Amazon Web Services (AWS). The company has since reduced costs, increased collaboration, and accelerated data analysis.

Struggling with the Limitations of Legacy IT

Pitney Bowes hardware, software, and services have helped pioneer the flow of commerce and e-commerce for companies of all sizes. Employees would independently access data if they wanted to conduct their own analysis, leading to many siloed data versions. As the scale of data grew, along with the need for more complex data queries, the on-premises solution couldn’t keep up. To facilitate data sharing, the company migrated to an open-source cloud service and gave everyone access to the database. This solved the bottlenecks caused by resource constraints, but it also resulted in out-of-control resource usage and unallocated storage costs.

Unleashing Data and Demolishing Silos

Pitney Bowes compared a handful of data warehouse solutions: SQL Server, MySQL running on Amazon Relational Database Service (Amazon RDS), Amazon Redshift, and Snowflake, to see which met its requirements for cost effectiveness, flexibility, and ease of use. Snowflake came out on top. Working with Snowflake, Pitney Bowes built a solution on top of its Amazon Simple Storage Service (Amazon S3) data lake and connected its E-Commerce and SQL Server data marts on the back-end. With Snowflake, Pitney Bowes employees would now have access to the same data and have the freedom to share it with colleagues and stakeholders across the company. This new level of capabilities has led to greater efficiencies, better IT cost management, and more informed business planning.

Improving Business through Borderless Data Sharing

Prior to Snowflake, Pitney Bowes employees worked mostly in silos. The inability to share data led to duplicate work streams, a lack of collaboration, and drawn-out development cycles. With Snowflake in place, employees have a centralized dashboard to securely access and share data stored in the company’s Amazon S3 data lake. Having a greater level of data visibility also helps the company operate and develop ecommerce solutions at a faster pace. “With the ability to see and share data, development cycles have accelerated and time to market is much faster,” said Vishal Shah, Solutions Integration & Deployment Architect at Pitney Bowes.

“With Snowflake, more than ever before, we are collaborating with engineering and marketing to understand and share data.”

- Vishal Shah, Solutions Integration & Deployment Architect, Pitney Bowes

Metering Costs Based on Consumed Storage and Compute

Running Snowflake on AWS provides Pitney Bowes with a scalable, pay-as-you-go pricing model based on the volume of data and amount of compute time used. And with the ability to conduct multiple workloads on one copy of data, the company can delete numerous SQL Server replicas. Perhaps best of all, IT can now allocate costs to the respective business units because workloads all target the same data. “With Snowflake, we have the flexibility of a cloud infrastructure and the ability to distribute costs, lessening the impact on our baseline operating costs,” said Shah.

Extending Business Insights by Integrating with More Services

Pitney Bowes uses other AWS Cloud Services, such as Amazon SageMaker, AWS IoT, and AWS Elasticsearch Service and existing third-party services to create better business insights. SQL Server, Salesforce, and other back-office workloads running across Pitney Bowes’ network send batch data into an Amazon S3 data lake. From there, Pitney Bowes uses the machine learning capabilities of Amazon SageMaker to generate predictive analytics about factors such as the volume of returned Pitney Bowes equipment or a mail parcel’s delivery date. Meanwhile, the company uses AWS IoT to stream real-time data into the data lake, and AWS Elastic Search for a real-time view of it. “We can see what’s currently happening with the device, whether it’s up and running or if there’s an issue. Then we can perform processes on top of the data to identify a solution,” said Shah.

By running Snowflake on AWS, Pitney Bowes has simplified data sharing, controlled costs, and generated company-wide insights that are the basis for better business planning.

Pitney Bowes

About Pitney Bowes

Pitney Bowes power billions of transactions around the world through solutions, analytics, and APIs for ecommerce fulfillment, shipping and returns, presorting, and financing.

About Snowflake

Thousands of customers worldwide mobilize their data in ways previously unimaginable with Snowflake's cloud data platform—a solution for data warehousing, data lakes, data engineering, data science, data application development, and data exchange. Snowflake provides near-unlimited scale, concurrency, and performance, and delivers a single experience spanning multiple clouds and geographies.

Published March 2021