Published by
Published on
August 6, 2026
How one of the world’s largest CPG companies is transforming network design from periodic modeling into a real-time, always-on capability — built on Optilogic
“What would change if your network could think as fast as your business moves?”
For most of network design’s history at Unilever, it was far from the reality, as models were built when needed, handed off, and rarely revisited until the next major business event.
The scale of Unilever’s U.S. network makes that untenable. With 7,000+ SKUs, 400+ customers, 1,200+ ship-to points, and 100+ factories and co-manufacturers feeding multiple distribution centers, too many variables move too fast for periodic modeling to keep pace. A sourcing decision made six months ago may no longer reflect current transportation rates.
From a journey started ten years ago, Elie Silow-Carroll, Associate Director of Supply Chain Strategy for Unilever North America, set a mission: transform network design into a tech-enabled collaboration engine, connecting granular supply chain data to real-time decisions across teams and partners. “When we set this goal 10 years ago, the tech was not there to support it.” All that has changed now.
Network 1.0 was regional DC placement by Excel and maps — one size fits all. Network 2.0 brought end-to-end SKU-level optimization from factory to customer, with supply chain software enabling periodic medium and long-term projects. Unilever North America built the company’s only dedicated regional network modeling team in the world, creating a sustained pipeline rather than one-off engagements.
Network 3.0 — where Unilever operates today, built with Optilogic — is the collaboration engine. Segmented networks allow Unilever to design distribution around each customer’s actual SKU profile and needs, meaning two customers in the same geography may be served from different DCs because the model shows that’s optimal at the product level and for the customer.
The bigger shift was collaboration — upstream and downstream. Upstream, Unilever now brings planners, manufacturers, and business groups into the analysis phase rather than presenting finished recommendations. “A lot of times the teams see an analysis and get nervous about change,” Elie noted. “If they’re part of the actual build with that upstream collaboration, we’ve seen that make a huge impact on how we’re able to execute.” Downstream, Unilever works directly with retail customers on network overlays — modeling both networks together to find savings neither party could see alone. In one case, this uncovered a direct-ship node in a customer’s network, bypassing Unilever’s DC entirely, with both sides sharing the cost benefit.
Apps and network overlays — developed in partnership with Optilogic — can unlock this further. Instead of running one-off analyses every time a manufacturer asks where to place a contract facility, an app would let them test any location and immediately see the downstream logistics cost impact. Instead of a finished network handed to a retail customer, a network overlay would let both sides visualize, share, and model together — turning network design from a handoff into a conversation.
Upstream collaboration has changed how recommendations get adopted — teams that help build the analysis don’t fight the result. Downstream collaboration has uncovered strategic opportunities that reduce miles and cost for both Unilever and its retail partners. The model means the North American team can serve business users continuously rather than through a queue of ad hoc requests.
These are the foundation Unilever is using to build toward Network 4.0 with Optilogic: real-time data feeding continuous optimization, AI recommendations for decisions that involve trade-offs, agentic actions for those that don’t, and automated scenario building that surfaces opportunities the team wouldn’t think to run. “We had the vision 10 years ago,” Elie said. “The technology is finally catching up.”
How one of the world’s largest CPG companies is transforming network design from periodic modeling into a real-time, always-on capability — built on Optilogic
“What would change if your network could think as fast as your business moves?”
For most of network design’s history at Unilever, it was far from the reality, as models were built when needed, handed off, and rarely revisited until the next major business event.
The scale of Unilever’s U.S. network makes that untenable. With 7,000+ SKUs, 400+ customers, 1,200+ ship-to points, and 100+ factories and co-manufacturers feeding multiple distribution centers, too many variables move too fast for periodic modeling to keep pace. A sourcing decision made six months ago may no longer reflect current transportation rates.
From a journey started ten years ago, Elie Silow-Carroll, Associate Director of Supply Chain Strategy for Unilever North America, set a mission: transform network design into a tech-enabled collaboration engine, connecting granular supply chain data to real-time decisions across teams and partners. “When we set this goal 10 years ago, the tech was not there to support it.” All that has changed now.
Network 1.0 was regional DC placement by Excel and maps — one size fits all. Network 2.0 brought end-to-end SKU-level optimization from factory to customer, with supply chain software enabling periodic medium and long-term projects. Unilever North America built the company’s only dedicated regional network modeling team in the world, creating a sustained pipeline rather than one-off engagements.
Network 3.0 — where Unilever operates today, built with Optilogic — is the collaboration engine. Segmented networks allow Unilever to design distribution around each customer’s actual SKU profile and needs, meaning two customers in the same geography may be served from different DCs because the model shows that’s optimal at the product level and for the customer.
The bigger shift was collaboration — upstream and downstream. Upstream, Unilever now brings planners, manufacturers, and business groups into the analysis phase rather than presenting finished recommendations. “A lot of times the teams see an analysis and get nervous about change,” Elie noted. “If they’re part of the actual build with that upstream collaboration, we’ve seen that make a huge impact on how we’re able to execute.” Downstream, Unilever works directly with retail customers on network overlays — modeling both networks together to find savings neither party could see alone. In one case, this uncovered a direct-ship node in a customer’s network, bypassing Unilever’s DC entirely, with both sides sharing the cost benefit.
Apps and network overlays — developed in partnership with Optilogic — can unlock this further. Instead of running one-off analyses every time a manufacturer asks where to place a contract facility, an app would let them test any location and immediately see the downstream logistics cost impact. Instead of a finished network handed to a retail customer, a network overlay would let both sides visualize, share, and model together — turning network design from a handoff into a conversation.
Upstream collaboration has changed how recommendations get adopted — teams that help build the analysis don’t fight the result. Downstream collaboration has uncovered strategic opportunities that reduce miles and cost for both Unilever and its retail partners. The model means the North American team can serve business users continuously rather than through a queue of ad hoc requests.
These are the foundation Unilever is using to build toward Network 4.0 with Optilogic: real-time data feeding continuous optimization, AI recommendations for decisions that involve trade-offs, agentic actions for those that don’t, and automated scenario building that surfaces opportunities the team wouldn’t think to run. “We had the vision 10 years ago,” Elie said. “The technology is finally catching up.”
Fill out the form to unlock the full content
.jpg)
