1. BY ARC ADVISORY GROUP DECEMBER 2001
Achieving Profitable to Promise in
Distribution Centric Supply Chains
Profitable to Promise Is Possible.............................................................. 3
PTP Is as Simple as ABC......................................................................... 3
ABC Analytic Engines............................................................................. 5
Pricing Planned and Spot Buys ............................................................... 7
ABC Formulas Must Be Accurate........................................................... 10
PTP, Product Cost, and Dynamic Pricing................................................ 12
PTP and Customer Service.................................................................... 13
Enterprise & Automation Strategies for Industry Executives

2. ARC Strategies • December 2001
Profitable
2%
Customers
Profit
Margin
0%
-2%
Unprofitable
Customers
Customers
(Ranked Least to Most Profitable)
Taking Orders from Unprofitable Customers Is Bad Business
ABC Data Long Term
Sources
Contract
ERP ABC Negotiation
General Ledger
Analytic
Allocation Logic
Order Mgmt.
Engine
WMS Customer Profit Configurator
3PL Billing Rating Customer
Labor Mgmt. 1 A Specific
2 B
Spot Pricing
TMS 3 B
4 C
5 B
CSR Portal
Other 6 B
Problem
Resolution
The ABC Analytic Engine
2 • ARCweb.com • Copyright © ARC Advisory Group

3. ARC Strategies • December 2001
Profitable to Promise Is Possible
First came Available-To-Promise (ATP), which allows product to be prom-
ised if there are sufficient quantities in the warehouse. Then came Capable-
To-Promise (CTP), which allows the Make-To-Order (MTO) manufacturer to
check production capacity and see if the customer order date can be met. But
the holy grail of order promising is Profitable-To-Promise. The question
should not be “can we meet the customer’s requirements?” But rather,
“should we?” How profitable is this customer? How profitable will this or-
der be?
Because PTP involves a customer touch point surrounding the
The time cost of each process step
sales process, it is a Customer Relationship Management (CRM)
required to sell and deliver an item
application. Traditional CRM solutions cannot tell you how
must be accounted for.
profitable your customers are. Profitability should be the central
focus of managing customers, and PTP lies at the heart of ad-
vanced CRM solutions. Doing PTP will still require CTP or ATP. It is bad
business to take orders that cannot be fulfilled in a timely manner. Eventu-
ally we will see PTP and CTP technologies merge, but that is still a few years
out.
Order profitability has three main pillars: customer behaviors, product and
service costs, and the price charged. Activity-Based Cost (ABC) Analytics
Engines should be used to provide the core customer profitability data for
robust PTP within the Distribution-Centric Supply Chain sector. Distribu-
tion-Centric Supply Chains are supply chains in which the key constraints
fall within the sphere of fulfillment, from the finished goods Distribution
Center (DC) out to the customer. The Activity-Based Cost (ABC) methodol-
ogy is virtually a perfect fit for doing PTP in the Distribution-Centric Supply
Chain. While ABC will play a role in PTP in the Manufacturing and Sourcing
Supply Chains, its fit is not quite so seamless.
PTP Is as Simple as ABC
Customers do not behave the same way. Some place simple orders with long
lead times. Others place complex multi-product and multi-ship orders, call
back to change the quantities, call back again to change the delivery time,
and then all too often call to cancel the order. Furthermore, some customers
Copyright © ARC Advisory Group • ARCweb.com • 3

4. ARC Strategies • December 2001
require Value Added Services (VAS) such as special packaging, kitting, and
shipping. The only way to know whether a customer is profitable is to en-
gage in an Activity-Based Costing (ABC) analysis. Activity-Based Costing is
at the core of profitable order promising.
More Profitable Customers Less Profitable Customers
Order large quantities Order small quantities
Order high margin products Order low margin products
Order standard products Order special products
Pay standard prices Require discounts
Pay on time Pay late
Cancel orders rarely Cancel orders frequently
Use standard packaging, cases & pallets Require special fulfillment services
Use standard deliveries Require special deliveries
Rarely return goods Often return goods
Do not use charge backs Frequently use charge backs
Require low technical support Require high technical support
Profitable and Unprofitable Customers Can Be Distinguished by Their Behavior
McNeilus Steel, a Minnesota-based light manufacturer and distributor of car-
bon steel mill products, implemented a PTP application in just a few months
and received payback for it in less than a year. Their focus was generating
granular Customer Profit & Loss (P&L) statements that allowed their sales
force to understand which customers were profitable and more importantly
why. What were the customer behaviors that made an account a money-
losing proposition?
Once armed with that data they worked to change the customer’s behavior.
Changing behavior, rather than dumping customers, is particularly impor-
tant for a distributor because the more customers buy, the more volume
discounts a distributor can get from their suppliers. McNeilus’ advice for
others contemplating this kind of a project: have wide cross-functional par-
ticipation in the project. But even with wide participation, McNeilus found
an initial unwillingness among the sales staff to believe the Customer P&L
reports.
4 • ARCweb.com • Copyright © ARC Advisory Group

5. ARC Strategies • December 2001
Activity-Based Costing
The vast majority of manufacturers do a poor job of understanding customer
profitability because existing applications are not designed to capture activ-
ity-based customer order costs. Traditional accounting assumes that labor
and direct materials represent the vast majority of a company’s costs. Auto-
mation, efficient manufacturing, and the proliferation in value added
services combine to reduce the proportion of costs represented by direct ma-
terial and labor. The goal should be to gather customer-based ABCs in a way
that allows managers with profit and loss responsibility to make better deci-
sions. In particular, this data can be used to decide how to price products for
particular customers.
ABC projects can be approached in one of two ways. A predictive model can
be created to pull data from a variety of systems and predict the profitability
of certain customers based upon their predicted behaviors. The problem
with this is that customer behavior does not always reflect what the model
predicted. The alternative is to use an ABC Analytic Engine that reports the
historical ABC results that are based on actual activities.
ABC Analytic Engines
An ABC Analytic Engine captures the Activity-Based Costs and allocates
overhead by Customer, Product, and Territory. The system pulls data from
ERP systems (particularly the General Ledger and Order Management mod-
ules) and other systems, and uses algorithms to allocate costs
(both direct and “overhead”) by order line and customer. ABC Analytic Engines Provide PTP
Reports Useful for:
For example, one set of activities in order fulfillment is based • Changing customer behaviors
on order taking. The cost of taking different orders depends • Renegotiating customer contracts
on the order taking activities. Regardless of the order, it may • Deciding which customers to drop
take a Customer Service Representative (CSR) 30 seconds to
pull up the right computer screen. The number of line items
in the order are also a factor; it may take 30 seconds to handle each line item.
If the item is not in stock, the CSR must call the vendor; that takes 10 min-
utes. Rush orders may need approval from a warehouse manager; that
usually takes 15 minutes. A customer on a credit hold takes 10 minutes, on
average, to get approval from the finance department. In short, ABC alloca-
Copyright © ARC Advisory Group • ARCweb.com • 5

6. ARC Strategies • December 2001
tions depend on getting the right data from a system, an analysis of behavior,
and if/then formulas that do the math and apply costs to customer activities.
In generating the algorithm to allocate Activity-Based Costs, the analysis will
frequently need to go across departmental boundaries. For example, the cost
of canceling an order is largely external to the customer service department.
If an order line is cancelled, the costs of the cancellation vary depending on
whether the cancellation was done over the Internet (virtually no costs if
enough lead time is given), whether the problem is handled by a CSR (a
small ABC associated with that CSR’s time), and the lead time (if sufficient
lead time is given, no cost, if the order has dropped into the Warehouse
Management System (WMS) for fulfillment, the costs for undoing work can
be substantial).
While there are companies that provide models for predicting Activity-Based
Costs, Acorn Systems is one of the few that provide a true standalone ABC
Analytic Engine. Profit Analyzer from Acorn does not predict ABCs, it meas-
ures these costs and generates monthly reports that include customer,
product, and vendor profitability.
Order Entry is 10 times
company average. Delivery
expense is 15 times company
average. The drill down shows
many small orders delivered
separately.
Processing in the South facility
is highly inefficient. The drill
down shows a majority of the
orders placed by customer
were processed out of the
South facility.
Large ERP companies also claim that their solutions have strong ABC com-
ponents. While ARC does believe that leading ERP companies can provide
6 • ARCweb.com • Copyright © ARC Advisory Group

7. ARC Strategies • December 2001
strong product costing in discrete industries whose products tend to have
longer product lifecycles, ARC is not impressed with their capability to track
costs from the finished goods Distribution Center (DC) out to the customer.
Further, while ERP style ABC is used widely in Europe, particularly Ger-
many, it is used as a budgeting tool rather than a dynamic CRM decision
support and execution solution.
Acorn Solutions does offer a solution that
can be used in a more dynamic fashion in
distribution centric supply chains. Most of 20.0%
the companies that had implemented an
< 1 Year
Acorn System’s solution had the highest
1-2 Years
praise for the company. ARC interviewed
seven customers, including five with im- 80.0%
plementations of more than a year. Eighty
Payback Period for ABC Analytic Engines
percent of the users report a payback pe-
riod of less than a year; one reports a
payback period of between one and two years. While the most common and
largest source of payback came from better management of customer profit-
ability, users also reported payback from better management of products,
vendors, and operations.
Pricing Planned and Spot Buys
The goods that are purchased by large companies fall into two classes: spot
buys and planned buys. Spot buys are somewhat spontaneous ad hoc pur-
chases whose price is usually based on a price list. Planned buys are
negotiated rates for larger quantities of goods to be delivered over a longer
period of time and are often negotiated on an annual basis.
Pricing Planned Buys
The ABC Analytic Engine produces a monthly customer profit and loss re-
port. Customers can be profitable or not based upon the product mix they
purchase, delivery factors, order size, and the type of Value Added Services
required. The results from this report can be used to change customer behav-
ior or renegotiate contracts. When renegotiating contracts, either the price
for products can be raised or special charges can be accessed for Value
Added Services or other customer behaviors that drive higher costs.
Copyright © ARC Advisory Group • ARCweb.com • 7

8. ARC Strategies • December 2001
All of the companies ARC interviewed, except one, report using the results of
the customer P&L to change customer behaviors. The company that is the
exception will eventually confront their customers. But they want to put
their own house in order first, so that customers cannot argue that it is their
inefficient operations, rather than the customer’s behavior, that is the source
of the problem.
The other companies that ARC interviewed report that being armed with
hard data and examples of customer behavior that drove higher costs made
the process of changing behavior easier than they had expected. One man-
ager said, “he who has the hard data wins.” Another manager reported that
some of the activities that were costly for the company (handling odd sized
pallets, for example) were also costly to its customers, something the cus-
tomer would not have discovered if the issue had not been raised.
Only one executive mentioned using the data to drop customers, and only
two used the data to renegotiate contracts. Users of this system that have
renegotiated contracts report that while they approached the renegotiation
with a good deal of trepidation, the process was often easier than expected.
They discovered that they were more important to some customers than they
had realized.
This process of setting a price for customers, particularly large or important
customers, may need to involve a team that includes sales managers as well
as managers with profit and loss responsibility. Historical profitability in-
formation is often only the starting point for making customer pricing
decisions. At times, unprofitable customers are accepted in an initial en-
gagement as a way of getting a company’s foot in the door. Fur-
ABC Analytic Engines Provide PTP
Information That Can Be Used to: thermore, a customer may have some divisions that are
• Change customer behaviors; unprofitable and others that are very profitable. Doing business
with the unprofitable divisions may be the price of doing busi-
• Negotiate or renegotiate
customer contracts; ness with the profitable ones. Finally, if facilities are
• Drop customers. underutilized, accepting a large, slightly unprofitable customer
is still good business.
Another company with an interesting solution is ABC Technologies. ABC
has developed a predictive modeling tool. While Acorn’s solution is inter-
nally focused, ABC Technologies’ Oros Value Chain Analyzer helps
companies examine collaborative activities that can lower costs for both
partners. These lower costs can than be reflected in a lower negotiated price.
8 • ARCweb.com • Copyright © ARC Advisory Group

9. ARC Strategies • December 2001
Pricing Spot Buys
Acorn Systems also has developed a predictive modeling tool called “Profit
Optimizer For Orders.” ABC Technologies has a similar solution. While the
ABC Analytic Engine can tell Account Managers after the fact whether an
order has been profitable, tools are needed to help them during the negotia-
tion phase. The ABC Analytic Engine, in turn, can close the loop and make
sure that the predicted customer profitability was actually achieved.
The predictive model used in the initial contract negotiation can be used on-
line, in real-time or off-line. On-line usage requires greater implementation
complexity. The predictive model allows key profit driving variables such as
whether the goods require special packaging, the skid size, and the extra
costs for rush and special orders to be factored into the contract. The simula-
tion engine is asked to present a set number of options for improving order
profitability and given a set number of simulation runs to discover those op-
tions. This tool can also be used in reverse auction situations and allows a
company to know when they should drop out of the auction.
The ABC Analytic Engine can also be used to price spot buys in a different
way. Periodically, perhaps every month or two, a customer’s ordering be-
havior can be reviewed, and customers can be classed as “A” customers
(highly profitable), “B” customers (marginally profit-
able), and “C” customers (unprofitable). C customers can All the managers interviewed believe that
be charged a higher flat rate or a price with surcharges sales people should be compensated
for specific behaviors ($100 every time an order line is based on profit rather revenues. This will
be critical in changing behaviors. All but
cancelled within 72 hours of due date delivery). The cus-
one have moved slowly to implement this
tomer-specific pricing can reside on Internet product
policy; changing compensation policies is a
configurators and catalogs or internal systems only the tough transformation.
sales people see.
The Account Manager Portal
It is not possible to completely automate the process of pricing goods and
services for specific customers. For example, if a customer calls to complain
that they have been inappropriately charged for an expedited shipment, an
Account Manager will still need to make resolve the issue. Account Manag-
ers will need a portal that pulls relevant data from the ABC Analytic Engine
and other systems to handle such complaints.
Copyright © ARC Advisory Group • ARCweb.com • 9

10. ARC Strategies • December 2001
ABC Formulas Must Be Accurate
Certain customer costs are easy to calculate. Calculating shipping costs for
particular orders by customer is a straight feed from the pertinent field in a
database. In other cases, for other calculations, the necessary data does not
exist. One interviewed company had to begin capturing customer order data
they had not captured in the past. The general methodology for determining
these costs is to have a cross-functional team that includes operational, fi-
nance, and IT people work with implementation personnel from the software
supplier. This implementation team interviews personnel from across the
company to understand processes, activities, and the time associated with
those activities.
Some users wish they had more closely supervised the Acorn Systems im-
plementation team that wrote the cost allocation formulas. They say that
while the Acorn people were diligent, they were not knowledgeable about
their business. Therefore, at times, they misinterpreted what company per-
sonnel said during the interviews. It was suggested that either someone
knowledgeable about the company’s operations participate in all interviews,
or at the very least, closely monitor the interviewing process and the conclu-
sions that were being drawn.
Indirect costs, at first blush, seem very difficult to allo-
“The process of interviewing a wide range of cate. They can, however, be logically allocated. For
people (in order to write the formulas) example, accounting is considered a support depart-
generated interest in what we were doing
ment with indirect costs. But it deals with accounts
and accelerated acceptance of the system.”
receivable and accounts payable, and customer and
- Brad Hillery, IS Manager, Webb Chemical vendor behaviors do impact the time spent on these
activities. The Human Resources and IT departments
are true support departments, and yet there are logical ways to allocate their
costs. HR’s work is related to the number of people in that company. The
more people a department has, the higher the allocation they deserve. IT’s
work relates to the number of computers in the company. The more PCs a
department has, the higher that allocation should be.
While these indirect cost allocations are not perfect (there can be rational ar-
guments for different allocation formulas), it is better to be 90 percent right
and making informed decisions than to act blindly. It is important that 100
percent of the costs from the general ledger be allocated to customers, prod-
ucts, and vendors. Otherwise, a company is operating with impaired vision.
10 • ARCweb.com • Copyright © ARC Advisory Group

11. ARC Strategies • December 2001
Phases Description
1. Pilot Model Building Project team analyses operations. Algorithms are cre-
ated to model work processes. This process can take
up to 2 months and involve scores of interviews with
managers and workers from multiple departments.
2. Model Validation Model results are reviewed by key personnel from mul-
tiple functional areas for the purpose of validating and
refining algorithms. This process requires multiple it-
erations spanning several weeks.
3. Implementation The software is implemented and appropriate person-
nel are trained. This can be done in 2 weeks.
4. Value Capture Business processes are changed to capitalize on the
new system. Process changes will continue to occur for
as long as the system is utilized by power users.
5. Model & Process This is a living model - the output depends upon the
Maintenance model’s fidelity. Up to a week of model maintenance
per quarter is required.
The Implementation and Maintenance Process for ABC Analytic Engines
Supply Chain Execution Systems Have Key Data
While all interviewed users pulled almost all of the data from their ERP sys-
tems, some said the process of determining formulas used data from
execution systems to verify the ERP data. Execution system data is critical to
these formulas. A good WMS shows data on slot sizes, important in allocat-
ing inventory carrying costs. And the labor costs captured in a good WMS
for receiving, put-away, picking, and different Value Added Service activities
can be quite granular. Shipping & routing systems show the distance and
costs of serving different customers. Often a PTP project indicates places
where data in different systems, such as the ERP and execution systems, are
inconsistent.
SCE Companies Offer Complementary Solutions
Certain WMS companies provide software and services that are quite helpful
in PTP projects. Suppliers such as McHugh Systems International and EXE
Technologies offer both a Labor Management System (LMS) and a service for
determining how long different tasks in a DC should take. Other companies,
like Optum and MARC Systems, have a LMS solution, but do not offer the
engineered labor services.
One of the largest cost buckets in distribution is the cost of warehouse labor.
Labor Management Systems (LMS) determine the target time for each task
based on its work characteristics (distance, weight, equipment, etc.) and
Copyright © ARC Advisory Group • ARCweb.com • 11

12. ARC Strategies • December 2001
compare target times to actual completion time. These solutions, imple-
mented correctly, have a high ROI based on reduced labor costs. An LMS
solution is highly complementary to an ABC project. Both projects put a mi-
croscope on actual activities in an attempt to improve profitability.
This Is a Living Model, Maintain Its Fidelity!
It should be apparent that a substantial amount of effort is involved in gener-
ating the ABC allocation model. That work is wasted if the model is not kept
up to date. At larger companies, the work associated with keeping these
models up to date can run to one week per quarter. The ABC Analytic En-
gine generates a report once a quarter when the books close; therefore this
work needs to be done each quarter. The work consists of changes to the
formulas based on new customers, vendors, overhead allocations that arise
as the result of new expenditures, and changes to processes.
PTP, Product Cost, and Dynamic Pricing
In addition to customer behavior and fulfillment costs, order profitability
depends upon product costs and the price charged. Within the Distribution-
Centric Supply Chain, product cost is related to the cost paid for the product,
the terms and promotions associated with the purchase, and the cost of han-
dling and putting away product that comes into the Distribution Center.
Advanced WMS solutions can tell you which suppliers are more costly on
the inbound side, which can be related to whether they use Advanced Ship
Notices, how they palletize and shrink wrap the product, and how much
product damage is associated with the inbound shipments.
The price charged for product also obviously effects profitability. Some dis-
tributors distribute large numbers of products that are only slightly
profitable or even slightly unprofitable. These companies generate a large
percentage of their profits on a relatively few products, yet they cannot
eliminate most of the marginal products or risk losing customers. For these
companies, one key to profitability is based upon deriving the maximum
profit for products that suddenly are in short demand while minimizing
losses from slow moving products.
The masters of flexible pricing based upon variable demand and customer
segmentation have been the airlines, hotels and, more recently, rental car
12 • ARCweb.com • Copyright © ARC Advisory Group

13. ARC Strategies • December 2001
companies. Airlines, for example, adopt numerous pricing strategies to
maximize profitability. In the course of normal day-to-day operations, they
have to decide whether to quote a price low enough to guarantee early sales
or quote a higher price and risk that some units will remain unsold. In allo-
cating seating capacity, airline managers trade off between two types of
losses. Yield loss is selling at a low price and losing a better price later, while
Spoilage loss is waiting in vain to sell at a high price and losing the opportu-
nity of an earlier low price offer. So airlines typically sell some seats early at
low prices and guard against the risk of yield loss by blocking out some units
in the hope of selling them later at a high price. However, in high demand
travel periods, such as during holidays, airlines endeavor to garner the
maximum returns for a product that is in high demand.
Manugistics offers a dynamic pricing solution called PRO
(Pricing and Revenue Optimization) and is increasingly fo- Other corporate KPIs, in addition to
cused on helping manufacturers deploy these solutions. PRO profitability, need to be measured on
an ongoing basis. Too narrow a
forecasts the response of different customer segments to prices
focus on profitability can cause a
of products throughout the product lifecycle. PRO provides
company to under invest.
analysis on what a customer segment is willing to pay and
then generates specific pricing recommendations by period.
This solution depends upon strong customer data, and a large part of the im-
plementation can involve data cleansing. If the initial data is not strong, the
ongoing collection of customer responses to prices will allow the optimiza-
tion model to improve with time.
PTP and Customer Service
Among the most important issues in implementing PTP will be cultural is-
sues. Companies know they would not exist without customers, and it is
natural to want to keep those customers happy. It is also widely argued that
superior customer service leads to happy customers who will not defect to
competitors. Furthermore, it is also understood that it is far more costly to
gain new customers than to retain existing ones. It is natural for many peo-
ple in a company to resist the move to PTP because it just feels wrong.
These are valid feelings, and they need to be directly addressed before a
company can hope to succeed in a PTP project. There are some companies
that do compete on superior customer service, and upscale customers are
Copyright © ARC Advisory Group • ARCweb.com • 13

14. ARC Strategies • December 2001
willing to pay a premium for their services. But only a relatively small num-
ber of companies compete solely on the basis of superior service. The three
main ways companies compete are product (technically advanced or superior
product quality), price, and service. In many instances, some combination of
the three provides a niche in which a company can profitably compete.
Furthermore, PTP will enhance service for the majority of customers. One
measure of service is the Perfect Order Metric. A Perfect Order is an order
where undamaged products are delivered on time, in the proper quantities
with no unauthorized substitutions, with the proper Value Added Services
performed and the correct billing. Most companies perform miserably
around this ultimate service metric, with many companies having well under
60 percent of orders being perfect. Complex customer ordering behavior is a
major reason for poor performance on this Key Performance Indicator. The
most unprofitable customers will often be the customers that are adding the
most complexity to a company’s processes. Serving these customers’ needs
harms the service level for core customers.
A company’s assets, processes, and people skills determine what types of
fulfillment complexity a company can profitably handle. If companies de-
cide to add assets and improve processes to handle greater fulfillment
complexity, than that is a strategic decision. The payoff to such a decision
will depend upon the volume of the resulting business, the actual costs of
doing the more complex fulfillment, and the margin generated. Very often,
however, the most complex service requirements are best handled by Third
Party Logistics providers, and they will charge a set fee for the product but
variable fees based on the particular services.
Many people in a company have a vested interest not to move to PTP. If
salespeople are compensated based on sales rather than profitability, they
will resist any system that shows that they may be hurting corporate profit-
ability rather than enhancing it. Moving forward will definitely require
support from top management. Good luck!
14 • ARCweb.com • Copyright © ARC Advisory Group

15. ARC Strategies • December 2001
Analyst: Steve Banker
Editor: Ed Bassett
Distribution: All EAS Clients
Acronym Reference: For a complete list of industry acronyms, refer to our web
page at www.arcweb.com/arcweb/Community/terms/indterms.htm
3PL Third Party Logistics ERP Enterprise Resource Planning
ABC Activity-Based Costs IT Information Technology
API Application Program Interface KPI Key Performance Indicator
ATP Available-to-Promise LMS Labor Management System
B2B Business-to-Business MIS Management Information System
B2C Business-to-Consumer MRP Materials Resource Planning
BPR Business Process Reengineering OMS Order Management System
CAGR Compound Annual Growth Rate P&L Profit & Loss
CMM Collaborative Manufacturing PAS Process Automation System
Management PTP Profitable-to-Promise
CRM Customer Relationship Management ROI Return on Investment
CSR Customer Service Representative SCE Supply Chain Execution
CTP Capable-to-Promise SKU Stock Keeping Unit
DC Distribution Center TMS Transportation Management System
EAI Enterprise Application Integration VAS Value Added Service
EAM Enterprise Asset Management WAH Web Application Hosting
EC Electronic Commerce WMS Warehouse Management System
Founded in 1986, ARC Advisory Group is the leader in providing strategic plan-
ning and technology assessment services to leading manufacturing companies,
utilities, and global logistics providers, as well as to software and solution suppli-
ers worldwide. From Global 1000 companies to small start-up firms, ARC
provides the strategic knowledge needed to succeed in today’s technology driven
economy.
ARC Strategies is published monthly by ARC. All information in this report is pro-
prietary to and copyrighted by ARC. No part of it may be reproduced without
prior permission from ARC.
You can take advantage of ARC's extensive ongoing research plus experience of
our staff members through our Advisory Services. ARC’s Advisory Services are
specifically designed for executives responsible for developing strategies and
directions for their organizations. For subscription information, please call, fax, or
write to:
ARC Advisory Group, Three Allied Drive, Dedham, MA 02026 USA
Tel: 781-471-1000, Fax: 781-471-1100, Email: info@ARCweb.com
Visit our web page at ARCweb.com
Copyright © ARC Advisory Group • ARCweb.com • 15

16. Cambridge, U.K.
Düsseldorf, Germany
Munich, Germany
Hamburg, Germany
Tokyo, Japan
Bangalore, India
Boston, MA
Pittsburgh, PA
San Francisco, CA
Visit ARCweb.com for
complete contact information
Three Allied Drive • Dedham, MA 02026 USA • 781-471-1000 • Fax 781-471-1100