The B group is the traditional retailing still strong in Italy, the A group are
the branded stores. We didnt particularly pursue online sales although
they are a growing C group. A and B groups each are roughly 50% of
the retail marketplace in Italy. Group A is growing of course and B is
shrinking
D
B
Hold less
Reduce Costs
Inventory
A
Manage well
C D
Protect Hold More
Throughput Inventory
D
B
Hold less
Reduce Costs
Inventory
A
Manage well
?
C D
Protect Hold More
Throughput Inventory
Pull Distribution
Pull replenish stores from a centralized Warehouse
OLT daily OLT Daily OLT Daily PUSH
Inventory
SLT = 1-4 wks SLT 1-4 Days SLT 1-2 Days
Distribution
Reorder daily, replenish as often as possible
Replenishment LT
Order Suppy
LT LT
100
Efficiency Gap
(surplus)
ROP
25
Buffer Management
DBM protects sales without unduly increasing inventory
the solution?
The preceding steps insure that
goods are delivered centrally where demand variability is smallest
and consumption highest
goods flow to stores pulled by demand in a balanced way
buffers adapt in size to protect sales from variations in demand
7 weeks
Near East 95%
Far East
Delivery two
5% 1 week Italy months before
In season delivery - start of season
two months before
the end of season
90%
10%
success, still
This little story helped them to decide: Two cars leave Rome for Milan, a
FIAT compact and a Ferrari. Which one gets to Milan first?
They both get to Milan at the same time because they stick to the
maximum speed rule
Meaning: no advantage from Pull Replenishment if the business policies
are not changed accordingly
5 stores
Spring-Summer 2015 Season
An internal single-piece Distribution Center created ad-hoc to feed the
stores twice a week
The single-piece DC being fed in real time by de-kitting enough
additional inventory
Test performed on 20% of SKUs that could sell additional inventory (up
to 300% of the overall budget)
Pull Replenishment for stores and warehouses managed by Elucidate
a web application by IDEA LLC, our partner
Concept ?
Two conditions
In order to actually improve the company must be ready to adopt the
subordination principle (see next chart)
The company must agree that perfect availability is a decisive
competitive edge for their retail business
important?
Subordination means that all the company parts work to attain more of
the goal - perfect availability in our case
In practice this means that all main functions, production/purchasing,
discounting, product packaging, visual merchandising, logistics,
warehousing, even IT, must be revised to comply with the subordination
requirement
One key result of subordination is, a bit surprisingly, more harmony in
the company. When instead each function is striving to attain
independent optimization, almost all steps in this direction are causing
undesirable effects elsewhere
In our POC there were many undesirable effects due to the resistance
to subordination. The POC results though were very comforting
obtained
In the five stores the selected POC SKUs sold 35% more pieces,
the baseline being the store budget
Some item-store
exhibit even an
explosive behavior!
infinite availability?
Yes and no. No, they already had suppliers not further away than two
trucking days which could replenish fast enough. This change would
not involve the Far East suppliers and would have been relatively easy
to test
Yes, in time, they had to change not simply Purchasing, but some
other functions that had better align themselves to the new
environment
And, Yes they had to really want to achieve exceptional results so in
the end it was the CEOs vision of the company future that was
important
During the discussions one issue stood out, namely the difficulty of
procuring the died fabrics required for the additional sales
Curiously in one next project we found a solution for the died fabric
problem wed like to share
problem?
in a collection
In a collection typically 33% of the SKUs sell all their budget, another
33% sell passably, and the last 33% do not sell much
We can reasonably assume that same died fabric items share the same
distribution
If this is true, why ship all
Ratio consumption vs budget, all sku-store
items in one single batch
when we know we wont
sell the whole batch?
Why not ship an initial
batch of 50% of forecast
without cutting the
remaining fabric?
Our fast supply
capability will allow us to
replenish to avoid stock-
outs
See the results next chart
www.tocpractice.com 32nd International Conference of the TOC Practitioners Alliance - TOCPA
The one single batch is the Your logo
norm in fashion
The seven models have many SKUs and all share the same fabric
Here is what happens if the supplier delivers the forecast in one single batch
Season Actual Total season
Unsold
forecast monthly sales sales
2 fast models
700 / 600 1000
2 avg. models
7 models 600 400
1000 x 150
4 months
season model 3 slow models
100 400 600
* The unused died fabric allows only 3500 4.400 3.500 300 items
more items. Actual sale rates set at 20% less items items
First batch sales amount to 3.200. The slow sellers leave 300 items unsold. Total sales
are 6.700 items, 96% of total production and unsold quantity is 4%.
Average inventory is less than a half the previous value, while sales
increase more than 50%. The results is a sharp increase in inventory turns
More turns
+279%
Excerpt from the The Choice, by Eliahu M. Goldratt, The North River Press, 2008