Conventional wisdom says the wrong reason to migrate to the cloud is to save money, with cost reductions typically topping out at no more than 15%. Don’t tell that to Cintas Corp., though.

The $8.8 billion maker of uniforms, cleaning supplies, safety equipment and other business services recently migrated its suite of more than 200 SAP SE enterprise resource planning and related applications from on-premises infrastructure to the Google LLC cloud. In the process, it cut its hosting expenses by two-thirds.

The efficiencies surprised even Chief Information Officer Matt Hough, who arrived in 2013 with a charter to build a foundation of SAP knowledge and optimize the ERP software for maximum value.

Doubling down on SAP

Cintas had already gained significant efficiencies since first installing SAP in 2008. “It gave us automated connections between the supply chain and our operations as well as visibility into sales to ensure that we have all the inventory that we need to be optimized for distribution to our customers,” Hough said. With 540 locations across the U.S. and Canada and the world’s fifth-largest fleet of delivery trucks, the payoff from tighter logistics orchestration was significant.

But the company was paying dearly for those benefits in information technology costs. Running SAP’s Enterprise Central Component software with a local hosting provider required Cintas to overspend on mainframe hardware to accommodate future growth needs, but much of that excess capacity went unused for months or longer.

“When you buy a frame, you can’t just buy one or two CPUs,” Hough said. “You have to buy bundles of hundreds. It gets costly, and you’re not going to use much of that capacity for the lifecycle of consumption.” Cintas’ reliance on on-premises infrastructure also limited the company’s flexibility to connect external data sources to its SAP HANA-based analytics data lake.

Cost savings and capacity control

Google and Cintas already had a good relationship, so Cintas chose to stick with the hyperscaler to take advantage of its microservices, cloud-native automation and Cortex integration framework.

“We decided to move our whole SAP system to Google Cloud Platform to save money and also to get a lot more compute power at the switch of a button,” Hough said. “Their costs were the best of the hyperscalers we saw and their approach to problem-solving fits our culture well. They showed strong support across the stack from migration to getting more out of the data.”

The move involved migrating all of Cintas’ SAP workloads and more than 130 terabytes of uncompressed data running on Oracle Corp.’s database management system to the Google Cloud Platform while also implementing a RISE with SAP business transformation project in one of its business units. To facilitate the move, Google recommended Lemongrass Consulting, a service provider specializing in SAP migration and cloud management.

The plan initially called for a lift-and-shift migration, but Lemongrass recommended a more heterogeneous approach that would transition Cintas from an IBM Corp. Power Platform to Google Cloud and IBM’s Db2 relational database management system. The consultancy figured it could cut the size of the database by half while significantly reducing costs using its Lemongrass Cloud Platform, a governance, management and automation orchestration layer tuned to SAP operations.

Over nine months, Cintas and Lemongrass migrated 212 SAP applications on production, test and staging servers to Google Cloud over a series of weekends. It also shifted multiple databases from Oracle to DB2. Total downtime during the move: zero.

“We selected Lemongrass because they’re very innovative,” Hough said. “They’re the only ones we’ve seen that have been able to move a system this size.”

Over a weekend, Cintas migrated the database and compressed its size by 60%. “We were on Oracle on Friday and DB2 on Monday,” Hough said. “We’re seeing huge performance gains.”

Analytics on tap

With modernized operational systems, Cintas is looking to dig deeper into applying analytics to transactional data. HANA analytics has already been put to work to refine prospecting for the sales force. And Google’s BigQuery and Duet generative artificial intelligence have the potential to advance personalization and use mapping data to route trucks and supplies more efficiently.

“We can sell people in the north the jackets they’re going to need for the winter or get breathable, flexible clothing to the people down south,” Hough said. “We can start narrowing down and optimizing inventory to sell intentionally to locations and customers.”

That could make the business payoff of moving to the cloud far greater than the savings on IT costs. 

Photo: Cintas