costs and fueling customers’ perceptions that K-C was more
focused on its own needs than on
its customers’ requirements.
K-C’s growing product portfolio
also created problems with order
fulfillment. To support its distribution, K-C relied on a “dynamic
sourcing” model that would
process an order, examine available
inventory throughout the company’s network, and then assign the
order to the plant or warehouse
that could ship directly to the customer at the lowest cost while still
meeting the customer’s delivery
requirements. If the first choice could not fill an order for a
particular product, K-C’s sourcing system would reassign
that product to whichever facility the software deemed the
next-best option.
However, executives found that dynamic sourcing created forecasting problems for the warehouses and sowed
uncertainty among carriers because they could not adequately plan for what would be shipped on any given day.
“One of our key metrics for success is better forecasting at
the DC level,” says Mark Jamison,
who was then vice president of
Kimberly-Clark’s North America
customer supply chain. “Yet every
time you reassign product sourcing,
you introduce variability, which
drives down forecast accuracy.”
ter predictability and cost savings.
Midway through the project, Jamison convened a group
of academics, consultants, retailers, and third-party logistics service providers to offer input and advice. The group
met three times over a nine-month period. Jamison’s project team reported to a board of K-C’s senior leaders, who
would be responsible for approving the blueprint.
The team’s guiding principle was simple but crucial: Cozy
up to your customers. The supply chain team quickly realized that K-C’s failure to locate facilities near its customers
was contributing to its distribution problems. To rectify
that, K-C decided to retain its own warehouses and DCs
connected to its factories, but would phase out the public
warehouses and replace them with nine mega-distribution
centers located near major customer touch points.
The group also decided to abandon the dynamic sourcing
approach in favor of a “fixed sourcing” strategy that aligned
each SKU with a specific factory and DC. For the first time,
K-C would know in advance the movement of every prod-uct, where it was coming from and going to, and how it
would be transported.
By introducing predictability where little previously
existed, “we could become more consistent in our relationships with our truckers, and we could give our production
people a more stable plan to work from,” Jamison says.
With greater certainty and more
accurate forecasting, Jamison’s team
reasoned, they could plan shipments
much earlier. This would allow K-C
to divert more of its traffic from
over-the-road truck to lower-cost
and less fuel-intensive intermodal
service.
Jamison’s team was given four
years to complete the project. It didn’t have unlimited resources, and it
would have to work around customers’ day-to-day needs. Crucially,
the team would have to persuade
higher-ups to effectively dismantle the only supply chain
most of them had ever known.
The project got the green light in 2004. Within 30
months, K-C had leased and occupied eight huge “mixing
centers” or mega-DCs, spanning nearly all of the United
States and Canada. (See map.) A ninth facility, located
Kimberly-Clark Corp.
at a glance
; Founded: 1862
; Headquarters: Dallas, Texas
; 2008 sales: $19.4 billion
; Global operations: 37 countries.
Products sold in 150 countries
; Lines: 11 personal and family care
consumer brands; health care, professional, and partnership products
Cozy up to the customer
Faced with increasingly complex
and costly distribution, uncertainty
about inventory and forecasts, and
customers’ concerns, K-C’s management recognized that it had to
revamp the company’s North American supply chain.
Jamison headed a multi-disciplinary team that was charged
with the task. The team’s objective was to build a network
that would place K-C’s personal and family care products
under the same roof, located as close as possible to its customers and supported by a sourcing model that would fos-
