February
17, 2004 |
 |
 |
 |
 |
 |

|
 |

|
|
 |
We
received 112 responses to our survey questions last week regarding
RFID, with interesting results.
We
asked companies to identify their overall approach to RFID
adoption. Results were as follows:
23% |
Be
an industry leader/early adopter |
12% |
Leader/But
wait for others to be pioneers |
27% |
Go
forward prudently, but cautiously |
38% |
Wait
until technology/ benefits are more proven |
While
the large percentage that will approach the technology
cautiously
is not unexpected, the large percentage of those that intend
to be leaders is interesting.
We
also asked what would be the biggest driver of RFID adoption.
50% |
Customer
compliance |
42% |
Internal
distribution/ logistics improvement |
4% |
Closed
loop manufacturing systems |
4% |
Cargo
tracking/security |
Not
surprisingly, customer compliance led the results, followed
by the related area of internal logistics automation.
|
 |
|
|
 |
 |
   |
 |
 |
|
Dan Gilmore
Editor-in-Chief |
 |
 |
 |
Demand
Planning solutions have been around in one form or another
for about 20 years, and customer adoption really exploded
in the mid-1990s. I've worked with many companies that
have achieved huge benefits from implementing demand
planning solutions from leading players like Demantra,
i2, Logility, Manugistics and others. ERP vendors are
increasingly becoming players in this game, and newer
entrants like Steelwedge are approaching the demand
planning problem from unique angles.
Unfortunately
though, after an initial jump in forecast accuracy,
and substantially improved demand planning processes,
results often plateau, with companies still experiencing
relatively high levels of forecast inaccuracy. This
is especially true in the complex world of consumer
packaged goods, which is still characterized by heavy
use of trade and in-store promotions, "forward buying"
from retailers, and other factors that distort demand
and order patterns between CPG manufacturer and retailer,
while overall end user demand chugs along rather smoothly.
Of
course, Continuous Planning, Forecasting and Replenishment
(CPFR) was created to solve much of this problem, following
the modestly successful Efficient Consumer Response
(ECR) initiatives of the early-mid 1990s. While there
have been some isolated CPFR successes, it seems we
still have a long way to go before we see broad results
across the supply chain.
The
net of all of this is that average forecast accuracy
at the SKU/DC/week level for many manufacturers, can
consistently be off by 40-50%. The negative impact of
this forecast error is well known: customer service
problems, excess pipeline inventories, retail stock-outs,
and expedited transportation costs.
Proctor
& Gamble has announced plans to tackle this issue
in several ways, including an initiative that will eventually
enable it to schedule its plants much more flexibly
(see SCDigest
archives ). But is an easier fix simply being more
intelligent about actual demand and order inflow as
they are occurring, rather than basing forecasts primarily
off of historical patterns and expected lifts?
Talk
to demand planners in most consumer goods companies,
and they will describe a similar operating process,
in which they take the demand planning system forecast,
compare it to what else they know (actual orders), add
a little intuition, and then adjust the plan. But this
"second guessing" and re-adjustment of the forecast
is tough to do this well over hundreds or thousands
of SKUs.
Campbell
Soup has found a way to largely solve this problem by
combining traditional forecasting processes and technology
with a "bolt-on" solution that improves the forecasting
result by taking greater consideration of actual order
patterns during the forecast period. I first heard the
Campbell's story at the Manugistics user conference
last fall, and then gained some additional insight from
Rob Byrne of Terra
Technology, which provided the Campbell's
solution.
Click
here for a summary of the Campbell's solution
and results. Here's the bottom line: by adjusting short-term
demand forecasts through increased consideration
of
actual orders, and the inter-relationships between
demand in adjacent time periods, Campbell's was able
to reduce
percentage forecast error from the high 40's to the
mid-teens.
What
are the keys to improving forecast accuracy? Do we need,
as Campbell Soup has done, to consider actual order
patterns more clearly in understanding near term demand
patterns? Has your company solved the forecast challenge
through improved processes or technology?
Let
us know your thoughts. |
 |
|
|
|
 |
|
 |
The
Truth About Supplier Lead Times
Improving
Order Management for Complex Engineered Products
Summary
and comment below.
|
|
|
|
Supply
Chain Investment News
The
Supply Chain stocks in our index had another
mixed week, with no really big moves either
up or down. Aspentech and Agile
were the biggest losers, falling 4.29% and 5.23%, respectively,
while i2 reversed a recent slide with a gain
of 4.9%. |
 |
 |
|
Click
here to see performance over the past week, month,
quarter and year >> |
|
|
|
It was also a mixed week for our Logistics
software, 3PL and transportation stocks
with the most noticeable moves from FedEx
(up 4.20%), Yellow Roadway (up 5.34%), Symbol
(down 4.2%) and Vastera (down 9.4%).
|
 |
 |
|
Click
here to see performance over the past week, month,
quarter and year >> |
|
|
 |
|
|
The
Institute for Supply Management's well-known PMI
Index that tracks U.S. manufacturing health is
the weighted average of what five economic measures?
Answer below
|
|
|
 |
|
 |
Reader feedback from the
topics in these newsletters is what SupplyChainDigest
is all about. Give us your thoughts on this week's
logistics topics at feedback@scdigest.com.
|
 |
 |
 |
 |
You will be seeing an increased
level of formal and informal market research from
SCDigest.
Our
reader survey questions this week focus on demand
planning and forecasting. End users
only on the questions. Comments welcomed from
all (consultants, vendors, etc.).
|
|
|
|
|
 |
View
Full Article >>
Wal-Mart's
CIO Linda Dillman did a brief Q&A with Business
Week magazine on - of course - RFID.
No
new revelations, but some points worth highlighting:
 |
For
Wal-Mart, it's not so much about cost reduction
as it is increasing top line through fewer stock
outs at store shelf: "Costs savings isn't the primary
benefit for us of RFID, keeping goods in stock is.
We'll see better tracking and moving of inventory,
faster receiving and shipping, improved quality
inspection, fewer out-of-stock items resulting in
improved shopper satisfaction, greater predictability
in product demand, and better value for the shopper
as efficiencies occur," says Dillman. |
 |
Dillman
downplays the amount of investment that will be
required of both Wal-Mart and its suppliers for
hardware and software enhancements. "The amount
of money suppliers will need to spend on RFID technology
isn't as big as some people fear. In our stores,
we essentially put in a black box that sits at the
top of our legacy IT systems and sends us information
from the tags. For suppliers, the setup will be
similar. RFID also won't lead to a significant change
in the amount of data we'll have to deal with --
so it won't, in most cases, require extra spending
on basic software and hardware." |
 |
RFID
technology is evolving rapidly to deal with previously
existing challenges. "Only in the past 60 days,
researchers have figured out how to tag cases containing
liquids. We expect that, by 2006, most of these
issues will be past us, and that prices [for RFID]
will fall substantially - leading to an explosion
in the technology's adoption." |
The
Key Take-Away: Taking all this at its face, it says
Wal-Mart will go "slow but steady" in its RFID roll-out.
They "hope" to be live with the top 129 suppliers in
one market - Dallas - by January 2005. While suppliers
need to begin preparing, I think it's good news that
this will roll-out in a more measured way, recognizing
the need for continued technology advances, and that
you have to walk before you can run.
Do
you think Wal-Mart's RFID roll-out will go fast or slow?
Did anything in CIO Dillman's comments surprise you?
Let us know your thoughts.

|
  |
 |
 |
View
full article>>
Interesting
article in last month's APICS magazine. Based on years
of experience by the author, it suggests that far too
often, supplier lead times factored into complex MRP
or supply chain planning applications, come from the
supplier's sales rep, and may not be accurate.
As
author Randall Schaefer notes: " We had been working
toward Just-in-Time (JIT) manufacturing so our internal
lead times were under control. But a handful of purchased
components prevented us from meeting our lean targets.
We had reacted to these lead times by trying the usual
techniques: large safety stocks, vendor stocking programs,
consignment inventory, etc. All of them worked well
if judged on the criteria of not running out of parts.
But none of these techniques were acceptable if compliance
to JIT standards was the criteria."
Digging
into this issue, he found the lead time data in their
systems came from the parts buyer, who in turn got it
from the vendor's sales rep. So, he "asked the purchasing
manager to personally talk to a high-level operations
person, at each supplier, and explain our problem with
the lead times for these parts. I also asked the purchasing
manager to then provide the suppliers with our maximum
acceptable lead time, for a target, and to ask the suppliers
what they could do. I was surprised when the purchasing
manager succeeded in bringing all but four of the parts
within the maximum acceptable parameter."
Finally,
"Purchasing must make acceptance of the maximum allowable
lead time a condition for accepting quotes. Period.
Don't tolerate whining that this policy restricts purchasing's
authority or complaining that low cost suppliers are
eliminated. Purchasing's goal should not be to use the
low-cost supplier but to use the low-cost supplier that
can perform to quality, delivery, price, and lead time
standards. Old, favored suppliers whose quotes are no
longer accepted will quickly figure out how to synchronize
their delivery lead times with your allowable lead times."
Key
Takeaway: Don't assume your supplier lead times are
fixed based on what is "in the system." Having operations
work directly with its counterparts at the supplier
can often reveal insight or opportunities for both parties.
Do
companies too often just accept the sales reps' version
of lead times? Is this information often dated, or not
the true state of affairs? Do you have examples of stories
similar to the one above? Let us know your thoughts.

|
  |
 |
 |
View
full article>>
Ran
across an interesting piece from consultants PRTM on
the need for companies manufacturing complex-engineered
or configured-to-order products to dramatically improve
order management processes.
The
report cites one company that processes 1000 total orders
per year, but 200,000 changes to those orders, as during
the long lead times suppliers change parts, customers
make changes, the design is altered, etc.
These
changes are costly - first, the internal overhead to
manage the multitude of changes, and the drain on engineering
resources. While some think these "change orders" actually
increase profits, PRTM suggests they actually drain
profit margins. The key is improvement in the entire
order management process, which can significantly reduce
cycle times and the total cost of managing an order
through its lifecycle.
 |
Order
management process: Detailed look at current process
steps and opportunities to take out time and cost.
|
 |
Customer
order teams: Cross functional teams jointly manage
individual customer orders. |
 |
Order
management governance and measurement: Development
and detailed reporting on key order management process
metrics, with executive oversight. |
 |
Business
rules: More rigorous definition of policies regarding
pricing, options, configuration, material/component
suppliers, etc. to provide a decision framework
and internal discipline. |
 |
Improve
technology: But there are no magic bullets for this
problem from existing solutions. Key is avoiding
a morass of time and cost when selecting and implementing
order management systems in this environment. |
Key
Takeaway: While most complex, engineered products companies
would probably agree on the problem, this tends to fall
into the "important but not urgent" categories, and
never really gets addressed. But there are lots of dollars
on the table in terms of both cost reduction and increase
market share, for companies that substantially improve
order management processes.
Does
the order management process need to be fixed in most
complex, engineered-to-order manufacturers? What are
the key barriers to making these improvements? Is lack
of the right OMS systems for this market an issue? Let
us know your thoughts.

|
  |
 |
|
|
|
|
 |
Q. |
The
Institute for Supply Management's well-known PMI
Index that tracks U.S. manufacturing health is
the weighted average of what five economic measures?
|
The
PMI is a composite index based on the seasonally
adjusted composite of these five indicators: (New
Orders represents 30%, Production represents 25%,
Employment represents 20%, Supplier Deliveries
represents 15%, and Inventories represents 10%).
Now you know. |
|
  |
 |
|
|
|
© Copyright 2004, SupplyChainDigest. All rights reserved. |