Effective Benchmarking

Published byInstitute of Management Accountants10 Paragon DriveMontvale, NJ 07645-1760www.imanet.org

Copyright © 1995Institute of Management Accountants

All rights reserved

Statements on Management Accounting

B U S I N E S S P E R F O R M A N C E M A N A G E M E N T

T I T L E

Effective Benchmarking

C R E D I T S

This statement was approved for issuance as aStatement on Management Accounting by theManagement Accounting Committee. The Institute ofManagement Accountants extends appreciation to TheSociety of Management Accountants of Canada (SMAC)for its collaboration in creating SMA 4V and to MohanKharbanda, Vice President, Service and MarketDevelopment, home and building controls, Honeywell,Inc., who drafted the manuscript.

Special thanks go to Randoif Hoist, SMAC Manager,Management Accounting Guidelines, for his continuingoversight during the development of the Statement, andto the members of the focus group that provided adviceand counsel, including Management AccountingCommittee member Dennis Daly and the staff directorof IMA's Continuous Improvement Center, WilliamMcMahon. Dennis Daly also suggested modifications tothe final draft that significantly improved Statement 4V.The IMA is grateful to members of the 1994-95Management Accounting Practices Committee and its Subcommittee on SMA Promulgation for their earlier contributions.

Statements on Management Accounting

T A B L E O F C O N T E N T S

I. Rationale . . . . . . . . . . . . . . . . . . . . . . . 1

II. Scope . . . . . . . . . . . . . . . . . . . . . . . . . 1

III. Defining Benchmarking . . . . . . . . . . . . . .2

IV. The Role of the Management Accountant . .3

V. The Benchmarking Process . . . . . . . . . . .3

VI. Benchmarking Tools and Techniques . . . .5

Selecting and Prioritizing Benchmarking Projects . . . . . . . . . . . . . .5

Organizing Benchmarking Teams . . . . . .11

Documenting Own Work Processes . . . . .13

Researching and Identifying “Best-in-Class” Performance . . . . . . . . .17

Analyzing Benchmarking Data andIdentifying Enablers . . . . . . . . . . . . . . . .26

Implementing Benchmarking StudyRecommendations . . . . . . . . . . . . . . . .32

VII. Management Accounting Challenges . . .33

VIII. Conclusion . . . . . . . . . . . . . . . . . . . . . .34

Appendix The Benchmarking Code of Conduct

Glossary

Bibliography

ExhibitsExhibit 1: Benchmarking Phases and Activities 4

Exhibit 2: Criteria Testing . . . . . . . . . . . . . . .6

Exhibit 3: Hierarchy of Defined Characteristics . . . . . . . . . . . . . . .8

Exhibit 4: Comparing Criteria . . . . . . . . . . . .9

Exhibit 5: Performance Maturity Analysis . . .10

Exhibit 6: Key Business Process Assessment . . . . . . . . . . . . . . . .10

Exhibit 7: Materials Acquisition Functional Map . . . . . . . . . . . . . .14

Exhibit 8: Process Classification Scheme:Overview . . . . . . . . . . . . . . . . . . .16

Exhibit 9: Business System Decomposition .17

Exhibit 10: Database . . . . . . . . . . . . . . . . . .19

Exhibit 11: Sources of Benchmark Performance Information . . . . . . .20

Exhibit 12: Pros and Cons of InformationGathering Methods . . . . . . . . . . .21

Exhibit 13: Potential Partners Profile . . . . . . .24

Exhibit 14: Best Practices Matrix . . . . . . . . .25

Exhibit 15: Root Cause Analysis . . . . . . . . . .27

Exhibit 16: Radar Chart . . . . . . . . . . . . . . . .29

Exhibit 17: Business Performance Comparision . . . . . . . . . . . . . . . .30

Exhibit 18: Z Chart . . . . . . . . . . . . . . . . . . . .30

Exhibit 19: Force Field Diagram . . . . . . . . . . .31

Exhibit 20: Force Field Analysis . . . . . . . . . . .32

Effective Benchmarking


I . RAT IONALEThe long-term viability of any organization—whether in the competitive or not-for-profit sectors—depends largely upon how well it understands andhow effectively it meets its customers’ require-ments. Most organizations still have a long wayto go, but some have begun embracing TotalQuality Management (TQM) to meet customerrequirements. Yet TQM is only the first step inthe restructuring of institutions; productivitymanagement and business process reengineeringare also essential. Benchmarking is a tool tohelp companies with these two steps.

Benchmarking involves evaluating the other company’s business processes and adoptingthem to incorporate best practices to improveperformance, search for innovative ideas, andgain a competitive advantage. The corporateworld recognized the benchmarking process inthe 1970s. At first, it was used primarily by manufacturing companies to improve their products. Benchmarking techniques are nowcommonly used by service companies and appliedto customer service and staff departments.

Starting in the 1980s, service companies soughtdata about the internal workings of companies intheir own industry. Now companies are lookingbeyond their own industry to benchmark againstthe best in any given function as long as the lessons can be applied to their circumstances.They may also benchmark against their own customers and vendors.

Benchmarking provides a rational method for setting performance goals and gaining marketleadership and a broader, more accurate organiza-tional management perspective. Since it is basedon what the best are doing, it takes the emotionout of arguments about the need to change.

I I . SCOPEThis Statement of Management Accountingdescribes tools and techniques that improve theeffectiveness of benchmarking studies, with special emphasis on their practical applications.

It is addressed to management accountants andothers who are already familiar with the basicbenchmarking process. This tools and tech-niques statement will provide guidance to thosewho want to effectively conduct a benchmarkingstudy and accelerate the introduction of bench-marking into their organizations.

If you are not already familiar with the basicbenchmarking process you need to becomefamiliar with the types of benchmarking (internal,competitive, functional, and generic) as well asthe differences between a strategic and an oper-ational focus. The best sources of informationon benchmarking types and focus can be foundin Robert Camp’s Benchmarking: The Search forIndustry Best Practices that Lead to SuperiorPerformance and Business Process Benchmarking:Finding and Implementing Best Practices, andGregory Watson’s Strategic Benchmarking whichdemonstrates how to conduct each type ofbenchmarking project.

The recommendations above will furnish thereader with the essential background necessaryto effectively utilize the tools and techniques presented in this statement.

The tools, techniques, and case study includedin the guideline are structured to apply to:

l all levels of an enterprise;l all functions in an enterprise;l enterprises in all business sectors;l the public and private sectors; andl small and large organizations.

1


This SMA will be useful to benchmarking teamleaders, facilitators, and individuals who act asliaisons between benchmarking teams andupper management. It will help these bench-markers understand how to:

l identify benchmarking projects that are consis-tent with senior management’s strategic plans;

l analyze a process flowing across functionalareas;

l structure an approach to effectively gather“best-in-class” information;

l develop a framework for identifying, organizing,and analyzing process performance; and

l analyze benchmarking data effectively to compare performance measures and underlyingcauses and enablers.

I I I .DEF IN ING BENCHMARKINGBenchmarking involves continuously evaluatingthe practices of best-in-class organizations andadapting company processes to incorporate thebest of these practices. Webster’s (unabridged)Dictionary defines benchmarking as: “A standardor point of reference in measuring or judging...”According to Robert Camp: “Benchmarking is thesearch for industry best practices that lead tosuperior performance.” D.T. Kearns, former CEOof Xerox, states, “Benchmarking is the continuoussystematic process of measuring products,services, and practices of companies that arerecognized as industry leaders for the purpose ofachieving superior performance.” Gregory J.Balm of IBM has a similar definition:“(Benchmarking is) the ongoing activity of com-paring one’s own process, product, or serviceagainst the best known similar activity, so thatchallenging but attainable goals can be set anda realistic course of action implemented to efficiently become and remain best of the best in a reasonable time.”

Common to all of these definitions is a processdesigned to allow both an internal and externalassessment to develop and implement a plan forleadership in the marketplace. The key itemsstressed in these definitions are:

l Products, services, and practices: Benchmarkingtoday is much broader than the competitiveanalysis that traditionally focused on productfeatures or price comparisons. Benchmarkinganalyzes and measures the key outputs of abusiness process or function against the bestand also identifies the underlying key actionsand root causes that contribute to the perfor-mance difference.

l Ongoing process: Since the external environ-ment is continually changing, benchmarkinghas to be a continuous process as well. It cannot be this year’s management slogan or a fad. Competitive market forces tend to driveperformance to ever higher levels.

l Measuring: Measuring involves both quantitativeand qualitative analysis. It includes measuringthe difference in business processes with thebenchmark company. The quantitative compo-nent could define the gap between “as is” andthe “desired state,” such as a nine-month-longer product development cycle. The qualitativecomponent can include specific practices,such as early vendor involvement in the product development cycle.

l Industry leaders: Achieving parity against thebest in the industry may not always guaranteesuccess. The goal of benchmarking should beto compare one’s performance with the best inthe world. The best may be a non-competitor inanother industry.

2


Benchmarking should not be confused with competitive analysis. Benchmarking focuses on process and “best-in-class” comparisons.Competitive analysis is an investigation of thecompetitor without the help of the competitor.With benchmarking, the “best-in-class” openlyshare information with those conducting thestudy. The goal of benchmarking is to learn fromothers, adapt, implement, and improve.

IV. THE ROLE OF THE MANAGEMENT ACCOUNTANTBenchmarking was initially the domain of the marketing, product design, and businessstrategy functions. These functions are generallystaffed and operated by non-accountants. Today, benchmarking is seen as a company-widefunction that includes management accountants.Because management accountants are trainedto gather, analyze, measure, and report informa-tion, they can easily adopt the tools and techniques described in this guideline. The man-agement accountant can serve as an advocate,promoting the value of benchmarking and activelyhelping people to implement it. This will help makemanagement accountants become valuable mem-bers of the cross-functional teams to implementspecific benchmarking projects and systems.

The management accountant should integratebenchmark-derived standards, analysis of com-petitive gaps, and recommendations for improve-ments and tracking into existing managementreporting systems. These efforts will enhancethe role and responsibility of managementaccountants in the enterprise. They can highlightthe fact that a performance gap can be thesource of a major competitive disadvantage andkeep management focused on closing the gap.

The management accountant should be wellversed in benchmarking tools and techniques in order to:

l make the benchmarking process more effective;l ensure that the benchmarking projects are

prioritized, and based upon strategic criteria,to support company objectives;

l ensure that benchmarking team members andbenchmarking partners are selected to identifythe greatest areas of opportunity to close thecompetitive gap;

l ensure that internal and external performancedata are normalized and analyzed using thecorrect data analysis tools to identify gaps andtheir root causes;

l ensure that benchmark performance is projectedinto the future and periodically recallibrated;

l create a system of financial and performancemeasures to monitor ongoing progress againstinternal and external benchmark-based standards; and

l incorporate benchmarking as a key ingredientin the strategic planning process and theenterprise’s TQM efforts.

V. THE BENCHMARKING PROCESSThe benchmarking process has been formalizedby the leading practitioners into several phases.All of the practitioners use an integrated, sys-tematic, and measured approach to benchmark-ing reflected in the following five general phases:planning; data gathering; analysis and integra-tion; implementation/execution; and recalibra-tion. In each of these phases there are specificactivities to be completed. Exhibit 1 portrays theessential phases and activities in this structuredprocess. While organizations can modify them tomeet a particular situation, these activities arerecommended as a guide for implementingbenchmarking.

3


4


Acti

vit

ies

Benchm

ark

ing

Phases

Iden

tify

and

Prio

ritize

area

s to

be

benc

hmar

ked

• Se

lect

ing

and

prio

ritizi

ng

be

nchm

arki

ng p

roje

cts

• Or

gani

zing

benc

h-

mar

king

team

s

• Do

cum

entin

g ow

n

work

pro

cess

es

• Re

sear

chin

g an

d

iden

tifyi

ng “b

est-i

n-

clas

s” p

erfo

rman

ce

• An

alyz

ing

benc

hmar

king

da

ta a

nd id

entif

ying

en

able

rs

• Im

plem

entin

g

benc

hmar

king

stu

dy

reco

mm

enda

tions

• Re

calib

ratin

g

benc

hmar

ks

Esta

blis

h be

nchm

arki

ng

team

s

Dete

rmin

e da

taga

ther

ing

met

hod

Iden

tify

orga

niza

tions

to b

e be

nchm

arke

dGa

ther

ben

chm

ark

perfo

rman

cean

d en

able

r dat

a

Dete

rmin

e cu

rrent

perfo

rman

ce g

ap

Gain

man

agem

ent

acce

ptan

ce o

fco

nclu

sion

s an

d ac

tions

to c

lose

gap

Esta

blis

hpe

rform

ance

go

als

and

obje

ctive

s

Impl

emen

tac

tions

Asse

ss p

rogr

ess

towa

rd g

oals

Revi

se G

oals

,if

nece

ssar

yVa

lidat

e/up

date

benc

hmar

ks

Iden

tify

caus

es fo

r cur

rent

and

futu

re g

aps

Estim

ate

futu

reat

tain

able

per

form

ance

Dete

rmin

e re

leva

ntbe

nchm

arki

ng

mea

sure

men

ts

Stud

y pr

opos

edin

tern

al b

ench

mar

kpr

oces

s(es

)

Iden

tify

cust

omer

s of

be

nchm

arki

ng a

nd th

eir

requ

irem

ents

EX

HIB

IT 1

: B

EN

CH

MA

RK

ING

PH

AS

ES

AN

D A

CT

IVIT

IES

EXHIBIT 1. BENCHMARKING PHASES AND ACTIVITIES

This structured benchmarking process is thesole focus of Robert Camp’s Benchmarking: The Search for Industry Best Practices that Leadto Superior Performance. Camp also includes acomplete set of management presentation chartsin the Appendix for a clear and comprehensiveunderstanding of the benchmarking process.

Benchmarking is deliberate, time consuming,and at times, difficult. It requires organizationaldiscipline to be sustained in the face of day-to-day pressures. Benchmarking has significantresource requirements that need to be focusedduring implementation. These are often drawnfrom other areas in the organization and theseareas may suffer. There are two critical elementsthat need to be present to ensure a successfulbenchmarking activity: 1) direct involvement ofthe CEO, senior managers, and process ownersand 2) a willingness to adapt and learn from others. Bogan and English describe these twocritical elements including an in-depth casestudy in Chapter 4 of Benchmarking for BestPractices: Winning Through Innovative Adaptation.

VI .BENCHMARKING TOOLS AND TECHNIQUESThis statement describes several key tools andtechniques, covering all aspects of benchmark-ing. These are grouped into seven phases, asshown in Exhibit 1:

l selecting and prioritizing benchmarking projects;l organizing benchmarking teams;l documenting own work processes;l researching and identifying “best-in-class”

performance;l analyzing benchmarking data and identifying

enablers;l implementing benchmarking study recommen-

dations; andl recalibrating benchmarks.

Selecting and Prioritizing Benchmarking ProjectsBenchmarking will be doomed to frustration andfailure if the subject is not chosen carefully. Inorder to appropriately select and prioritize bench-marking projects, organizations need first tounderstand their critical success factors andbusiness environment. This helps them identifytheir key business processes and drivers. It alsohelps them develop the parameters that definewhat processes to benchmark. The criteria forselecting what to benchmark are related to the reasons for the processes’ existence andtheir overall importance to the organization’sbusiness mission, values, and strategy.

The reasons for the existence of a businessprocess can be judged by the practices associatedwith customer satisfaction. How important is it inthe internal supplier-customer chain or in satisfy-ing end users’ or customers’ needs? Ultimately,a process exists for this purpose only andwhether it should be benchmarked dependsheavily on answering this question.

Further criteria would determine how importantthe decision for improving a practice is and howapplicable benchmarking is to the decision. The importance of plans and strategy also is a relevant criterion.

Once the decision criteria are selected, organiza-tions can use one or more of the following toolsto prioritize benchmarking projects:

l criteria testing;l the analytical hierarchial process;l maturity analysis; andl key business process assessment.

Since many of the benchmarking tasks may notbe feasible due to high costs and the detailed

5


analysis required, smaller organizations willmost likely purchase benchmarking informationrather than develop in-house programs.

Criteria TestingSince resources are usually limited for extensivebenchmarking studies, it is important to haveagreement on which business processes shouldreceive this degree of analytical attention. Whenseveral benchmarking opportunities are avail-able for consideration, a decision matrix may beused to weigh improvement opportunitiesagainst various criteria for success. The criteriaused in the example in Exhibit 2 are: marketshare, gross margin, customer satisfaction, newbusiness opportunities, and defensive strategies.

Criteria testing includes:

l criteria and their relative weights with respectto each other; and

l key processes and their relative ability to influence the selected criteria.

Some subjectivity may be required in the prioritiza-tion process. It is, therefore, necessary to havethe subject matter experts involved in the develop-ment of the relative weights and priorities.

Criteria testing begins by assigning relativeweights to the selected criteria. In the examplein Exhibit 2, these weights are:

Criteria WeightsCustomer satisfaction = 3 (High)Defensive strategy = 3 (High)Market share = 2 (Medium)New business opportunities = 2 (Medium)Gross margin = 1 (Low)

Similarly, the key business processes to be prior-itized for benchmarking in this example, are:

l new product/market development;l customer acquisition;l supply chain management; andl manufacturing.

6


Criteria

Weight

Weights

1 = low2 = medium3 = high

Marketshare

Grossmargin

Customersatisfaction

Defensivestrategy

TotalKey business processes 2 1 3 2 3

9 2 6 9 6 32

8 2 3 5 6 24

2 2 6 2 2 14

3 5 3 2 3 16

New product/marketdevelopment

Customer acquisition

Supply chain management

Manufacturing

Newbusiness

opportunities

EXHIBIT 2: CRITERIA TESTING

EXHIBIT 2. CRITERIA TESTING

Using available facts and judgments, a relativepriority is assigned to each key business processon a scale of 1 to 5 (with increments of 0.5)relating to its ability to influence the criteria. Forexample, in Exhibit 2, the new product/marketdevelopment process is rated 4.5 in its ability todeliver market share.

The resulting product (that is weight x relative priority) is entered for each process. For newproduct/market development the entry undermarket share is 9 (i.e., 4.5 x 2). Similarly, thenew product/market development process israted 2 in its ability to affect gross margin(weight of 1) and therefore its entry is 2.

In this example then, the process priorities forbenchmarking, based upon the scores in thetotal column, would be as follows:

i) new product/market development;ii) customer acquisition;iii) manufacturing; andiv) supply chain management.

The Analytical Hierarchical ProcessThe Analytical Hierarchical Process (AHP) is anexcellent tool for establishing benchmarking priorities based on agreed-upon decision criteria.It is a structured way to use a consensus of subject matter experts to build a prioritized listof characteristics of the process being analyzed.It can be used at any level of detail, from prioritizing process goals to defining importantcharacteristics of a process. It has a wide varietyof uses in business analysis where a number ofcomplex alternatives need to be evaluated at thecrucial subprocess level.

It begins by the subject matter experts building ahierarchy of defined characteristics of the business process. The group starts with creatingconsensus on an overall (level 0) goal for theprocess and then dividing the goal into a few(level 1) subgoals in support of the overall objectives. These subgoals are then further sub-divided at (level 2). The hierarchy can be quitedetailed, though most applications need no morethan three levels, as shown in Exhibit 3.

Once the levels and elements are determined,the subject matter experts assign relative weightsto each defined characteristic within each hierarchy level using a consensus method, asshown in Exhibit 4. For example, gross margin is considered nine times more important thanmarket share. This places a value of 9 in column1, row 2. This naturally puts a value of 1/9 in row1, column 2. Similarly, customer satisfaction isdeemed by the team to be seven times moreimportant than market share, placing a seven inrow 3, column 1 and 1/7 in row 1, column 3.Each objective is considered simply to be asimportant as itself, thereby placing 1 in the diagonal column.

The paired comparisons are determined usingthe values scale shown below:

1 = equal importance to both elements in thematrix

3 = moderate importance of one element compared with another

5 = strong importance of one element comparedwith another

7 = very strong importance of one element compared with another

9 = extreme importance of one element compared with another

7


8


Goa

lLe

vel 0

Leve

l 1

Leve

l 2

Goa

l

Obje

ctiv

es

Optio

ns

Mar

ket s

hare

Units

sold

New

cust

omer

s

Prem

ium

pric

eco

mm

ande

dCo

stre

duct

ion

Prod

uct

mix

Old

cust

omer

sre

tain

ed

Delig

hted

cust

omer

sRe

duct

ion

in c

ompl

aint

sPr

oduc

tpe

rform

ance

New

cust

omer

sad

ded

New

busi

ness

reve

nue

cont

inue

d

Mar

ket s

hare

cust

omer

bas

em

aint

aine

d

Gros

s m

argi

nCu

stom

er s

atis

fact

ion

New

busi

ness

Defe

nsive

stra

tegy

EX

HIB

IT 3

: H

IER

AR

CH

Y O

F D

EFI

NE

D C

HA

RA

CTE

RIS

TIC

S

EXHIBIT 3. HIERARCHY OF DEFINED CHARACTERISTICS

The resulting numbers are added horizontally inthe total column. Exhibit 4 shows that gross mar-gin is likely to yield 41% of the goal as comparedto 17% for the other two nearest (level 1) goals,i.e., customer satisfaction and new business.These weights can be used for pair comparisonsto arrive at the normalized averages (that is, thesum of all relative weightings equals 1.0) asshown on the last column.

Recently, several types of PC-based AHP soft-ware tools have come onto the market that con-vert such a pairwise comparisons matrix into nor-malized averages or relative weightings.

Performance Maturity AnalysisA simpler benchmarking prioritization tool isPerformance Maturity Analysis. A maturity matrixprovides an overview and a look at the strengthsof work processes relative to the organization’sobjectives.

9


EXHIBIT 4: COMPARING CRITERIA1

Marketshare

Marketshare

1 1/9 1/7 2 5 8.25 0.15

9 1 5 3 4 22.00 0.41

7 1/5 1 1/3 1/4 8.78 0.17

1/2 1/3 3 1 4 8.80 0.17

1/5 1/4 4 1/4 1 5.70 0.10

53.3 1.00

Grossmargin


Newbusinessenabled

Defensivestrategy

Aggregates

Grossmargin


Newbusinessenabled

Defensivestrategy Total

Normalizedaverage2

Total Weighting Scale Interpertation

0-10: average importance of criteria11-20: important21-30: very important

EXHIBIT 4. COMPARING CRITERIA1

1 Normalized averages are calculated by dividing the row totalsby the aggregate. For example, the market share criteria total of8.25 53.53 = a normalized average of 0.15 and the gross margin criteria of 22.0 53.53 = a normalized average of 0.41.

–..

–..

A uniform set of characteristics for each workprocess is created. As shown in Exhibit 5, thesecharacteristics are listed down the left side todefine the rows in the maturity matrix. Across thetop (column headers), a few maturity (or achieve-ment) levels are defined. There are five levels,starting with absent and proceeding through low,medium, high, and mature. Values are assignedbased upon available facts and judgments. Acritical business process with low maturity sig-nals an immediate area in which to initiatebenchmarking. Conversely, a lower performance

may mean lack of strategic interest due tochanges in focus.

Key Business Process AssessmentAnother way to assess key processes is to analyze their influence across major businessfunctions, such as production, product design,and marketing. Senior management may need tofocus a benchmarking study on the processesthat have the most influence on the company. A key business process assessment is shown inExhibit 6.

10


EXHIBIT 5: PERFORMANCE MATURITY ANALYSIS

Documented

Processcharacteristic

Absent1

Low2

Medium3

Maturity levels

High4

Mature5

Scale

Able to deliver

Predictable

Repeatable

Efficient

EXHIBIT 5. PERFORMANCE MATURITY ANALYSIS

EXHIBIT 6. KEY BUSINESS PROCESS ASSESSMENT

Productdevelopment

Keyfunctions

Keybusinessprocesses

Assess marketrequirements M H H

Developcompetitive M H Mresponse

Developbusiness Hstrategy

Improve productquality, cost H L L H Mdelivery

Marketing

Relationship: L = Low M = Medium H = High

Sales Manufacturing Distribution

EXHIBIT 6. KEY BUSINESS PROCESS ASSESSMENT

Organizing Benchmarking TeamsThe planning, organization, and execution of awell-designed benchmarking study involves aconsiderable amount of time and energy.Engaging individuals to share the workload anddeveloping an equitable division of labor are notonly practical considerations, in most cases theyare required.

However, there are other reasons why bench-marking is an appropriate team activity. Forexample, those employees responsible for implementing changes based on the results ofthe benchmarking study are usually motivated to participate in the study and to make it a thorough and meaningful exercise. Another goodreason for the use of teams is the level of func-tional expertise and work experience a team represents. Different perspectives, special skills,variety of business connections, and physicallocation are some of the dimensions that individual team members bring to the bench-marking process.

There are two basic types of benchmarkingteams, defined by their structure and reportingrelationships. They are:

l Intact work groups: They are usually situated ina single location with all members of the groupreporting to a common manager. Typically, allmembers of the work group participate as abenchmarking team member. Intact workgroups are often the customers for their ownbenchmarking studies.

l Cross-functional, interdepartmental, and inter-organizational teams: These are often struc-tured as task teams or task forces with specif-ic charters and definite sets of customers. Theindividuals selected for these teams are chosen for their specific knowledge or skill

levels, but they also act as representatives of theirrespective departments, locations, or divisions. Inmany cases, these types of teams are broughttogether to work on one issue or problem. Oncetheir benchmarking study is complete, the teamdisbands. These teams often produce recom-mendations or reports and present those find-ings to their sponsors or to senior management.

It’s often convenient to put together a bench-marking team based on nominations or easyavailability of potential participants. However,organizations need to select their team carefullyto organize a team of motivated and skilled individuals who can effect change.

The team selection criteria should include suchappropriate elements as:

l Knowledge of the function: Effective bench-markers should represent the best and thebrightest in an organization in terms of theprocess being benchmarked. Not only shouldthey know the right questions, but they shouldbe able to interpret the responses from bench-mark partners.

l Credibility/respect: Since the recommendationsresulting from the benchmarking study couldsignificantly change the status quo, the personssuggesting change should be well respectedwithin the organization.

l Communication skills: Effective speaking skillsensure that questions are presented clearly andsuccinctly to benchmarking partners. Effectivebenchmarkers must also be good listeners,because the typical benchmarking interview isabout 25% speaking and 75% listening.

l Team player: Benchmarkers need to be able tointeract as team members for many months,often in locations removed from their normalwork environment.

11


l Interest/motivation: Benchmarkers must wantto expand their learning, be innovative, andsupport cross-functional problem solving.

l Project management skills: The team leaderand the team facilitator must have good projectmanagement skills, because the benchmark-ing effort may cut across many functionalboundaries and span several months.

These team selection criteria are shown in thefollowing table. According to this example,Peter, Kathy, and Nancy would be considered aspotential participants in the study.

In putting together a benchmarking team, it isalso important to consider team dynamics. Forexample, the following roles may be allocated tothe various team members:

l Team leader: to provide overall project direction and liaison with the management;

l Facilitator: for such functions as project coordi-nation, scheduling, and meeting facilitation;

l Data collector/analyst: to identify data needs,normalize the data collected, and analyze itusing the techniques described later; and

l Interviewer: to conduct on-site interviews,telephone, or mail surveys.

The size of benchmarking teams varies.However, for many organizations, the averagesize seems to be about six persons. Whenteams get too large, the level of qualitativereview and analysis by individual team memberscan decrease proportionately. (For more information on managing teams, see SMA 5C,“Managing Cross-Functional Teams.”)

12


BENCHMARKING TEAM SELECTION TEMPLATE

John 3 2 4 3 4 3 19

Kathy 5 3 2 4 4 4 22

Paul 3 3 2 4 3 2 17

Peter 5 4 4 5 3 2 23

Nancy 3 4 5 3 3 2 20

Potentialmember

Functionalknowledge

Credibility/respect

Communicationskills

Team-ing

skills

Projectmanagement

skillsInterest/

motivationTotalscore

Scale: 1 = Lowest, 5 = Highest

Benchmarking Team Selection Template

Documenting Own Work ProcessesDocumenting own business processes involvesself-study of the proposed benchmark processand examining those factors that influence business process performance.

Without a thorough investigation of its internalprocesses, an organization may not realize theextent of its improvement opportunities. Forexample, without an accurate understanding ofitself, an organization cannot calculate thepotential gap that exists between its outcomesor activities and those of the best-practicesorganization it has investigated. Furthermore,benchmarking is always an exchange of informa-tion. Organizations cannot exchange informationthey do not have.

In addition, organizations may be bypassing someimportant internal benchmarking opportunitieswithout a thorough internal analysis. They maynever discover the possible sources of informa-tion and assistance available to them internally.

Seeing an organization as a series of businessprocesses rather than as a fixed structure is particularly useful, because, to improve perfor-mance, senior management must optimize theorganization as a whole. As used in this guide-line, a business process is a network of relatedand interdependent activities linked by the outputs they exchange. Product creation, forexample, cuts across research, engineering,manufacturing, and marketing, while supplychain management cuts across a variety oforganizations: the firm's suppliers, their suppliers,the company itself, and its customers.

Some of the tools that benchmarkers use to document business processes are:

l cross-functional maps;l family of measures; andl process taxonomy.

Cross-Functional MapsIn organizations with complex activity chains, itcan be quite difficult to identify the primary char-acteristics of a process. One way to solve thisproblem is by tracing the path a product or servicerequest takes through the organization. The pathis a flow of work that leads to some final outputor product that is valued by the customer; it is avalue chain that knits the organization together.

Cross-functional maps help benchmarking teamsto identify the business process output and thecustomers for that output, to examine individualactivities, and to collect and analyze process performance data.

In detailing the flow, a series of questions needto be asked in order to help unravel the complexsequence of events that results in the delivery ofa product or service to an internal or an externalcustomer. Some of the preliminary questionsthat need to be addressed are the following:

l Who is involved in delivering the product or service?

l Why are they involved?l Why are they doing it?l Is what they are doing adding value in the

customers eyes?

Cross-functional maps help to focus the busi-ness processes down to the few essential inter-actions, as shown in Exhibit 7 for a typical mate-rials acquisition function for vendor deliveries.This helps explain:

13


l process flow (for example, Procurement orderentry as it progresses from Procurement toVendor, Receiving, Quality Inspection, Inventory,and the processing of invoices by AccountsPayable);

l hand-offs and organizational seams as thework elements flow from one internal function-al organization to another (for example, theincoming materials require hand-offs betweenReceiving, Quality Inspection, Inventory, andAccounts Payable);

l cycle times or completion time associated witheach of the above work elements;

l baseline transaction volumes for the businessprocess (for example, the expected, maximum,and minimum number of incoming materialsdaily, expressed both in different part numbersas well as the absolute volume for all materials);

l defect rates for each key element of theprocess (for example, frequency of miscodedinvoices);

l cost for each key element of the process; andl other relevant characteristics for each step of

the business process flow.

14


EXHIBIT 7: MATERIALS ACQUISITION FUNCTIONAL MAP

Copy to vendor (Alsoto: receiving, quality

control, and accounts payable)

Order generated

Vendor deliveryreceived

Partsinspected

QualityOK?

Yes

Procurement

Vendor

Qualityinspection

Inventory

Accountspayable

Yes

NoNo

Matched partnumber, units,

price withpurchase order

?

Vendor contacted for appropriate action

Partsinventoried

Checkissued

Authorizepayments

The invoice + goodsreceived notice sentto accounts payable

Function Activities

EXHIBIT 7. MATERIALS ACQUISITION FUNCTIONAL MAP

The mapping process usually begins as a verygeneral flow diagram, showing the main businessprocesses. The general, high-level maps can thenbe broken down into more detailed ones, until avery fine level of detail is obtained for all tasks inthe process. Generally, reengineering businessprocesses requires less detailed informationthan continuous improvement efforts, whichrequire a much finer detail. At detailed levels, morenumerical data are included. Adding levels of detailis not difficult, but it does require discipline andthe involvement of the staff who actually performthe work.

The mapping process itself may show improvementopportunities that can be acted on immediately.These may be critical to gaining the acceptance ofall involved in making the other fundamentalchanges.

Family of MeasuresGenerally, the benchmarking team would want todevelop a balanced set of measures or family ofmeasures that look at the process performancefrom the perspectives of the customers, employees,and shareholders, keeping in mind the need forcontinuous improvement. A family of measuresprovides a conceptual framework for identifyingprocess measures.

The units of measure, expressed as ratios or percentages, are the key gauges that the bench-marking study will want to uncover for any newbest practice. Knowing these in advance allowspreparation of a questionnaire to incorporatethem and ensures their documentation. They willbe used later on to determine the size of thebenchmark gap.

The measures selected should be true indicatorsof the process performance. For example, mostTQM programs consider the basic categories of

process performance measurement to be quality,cost, and cycle time. Using a generally acceptedset of measures is much better than using company-specific measures that may not matchanother company's measurement system.

Examples for process performance measures ofquality, cost, and cycle time are:

l First-pass yield (quality): the measure of effec-tiveness for a process in performing its transaction for the first time without defects;

l Value-to-cost ratio (cost): a measure of processeconomy in terms of its ability to produce highervalue of output for lower levels of processcost; and

l Cycle time: a measure of process efficiency interms of the amount of time consumed foreach transaction.

Other process performance indicators providedetail that supports the measurement of thesegoals. For instance, work-in-process inventory,process changeover time, and process down-timeall influence the cycle-time efficiency of the process.

Effective process performance measures shouldfit the following criteria:

l are driven by the external and internal customer;l support organizational goals and critical success

factors;l provide a good basis for comparison with

external and internal benchmarks; andl are easy to collect, reliable, and achievable.

These may be the same measures used by theorganization today or redefined measures thatcomprehend the true performance differences.The actual process of benchmarking may changesome of these measures if a significant re-engineering of the business process is desired.

15


While most business activities are measurable,it is sometimes not possible to develop anappropriate measure. Examples of this are organizational philosophy and strategic planningprocesses. In these cases, the team can use acase study approach to identify important lessons for the organization. (For more informa-tion on performance measures, see SMA 4U,“Developing Comprehensive PerformanceIndicators.”)

Process TaxonomyBecause the internal business processes,accounting practices, and measurement sys-tems are likely to vary from one organization tothe next, a process taxonomy, accompanying thefamily of measures, is desirable. A process taxonomy is a set of work process elements,measures, and phrases, with their definitions,that describes the process to be benchmarked.It ensures a common language with the bench-mark partners so that like elements can begrouped, measured, and compared.

16


EXHIBIT 8: PROCESS CLASSIFICATION SCHEME: OVERVIEW

Understandmarkets &customers

Developvision &strategy

Design products

& services

Market&

sell

Produce&

deliver

Invoice &service

customers

Opera

ting p

rocesses

Managem

ent

& s

upport

pro

cesses

Develop and manage human resources

Manage information

Manage financial and physical resources

Execute environmental management

Manage external relationships

Manage improvement and change

EXHIBIT 8. PROCESS CLASSIFICATION SCHEME: OVERVIEW

Source: American Productivity & Quality Centres

The act of creating a process taxonomy is some-times referred to as business system decomposi-tion. This approach looks at organizationsaccording to their various levels of performance:work and support processes, activities, andsteps. For example, an organization begins bydefining its work and supporting processes, asshown in Exhibit 8. It then decomposes each ofthese work and supporting processes to thelevel of activities and steps, as illustrated inExhibit 9.

Researching and Identifying “Best-in-Class” PerformanceThe researching and identifying “best-in-class”phase of the benchmarking study as describedin “Implementing Benchmarking” is often themost difficult for organizations. During thisphase, the benchmarking team needs to use theknowledge gained in its internal work processstudy to determine which organizations are mostappropriate for external process comparisons.

17


EXHIBIT 9: BUSINESS SYSTEM DECOMPOSITION

Develop new serviceconcepts and plans

Design and evaluateprototype services

Refine existingservices

Develop service enhancements

Eliminate qualityproblems

Eliminate outdated services

Test effectiveness of new or revised

services

Prepare for Production

Manage the servicedevelopment process

Understand marketsand customers

Develop vision and strategy

Design productsand services

Managebusiness

System Processes Activities Steps

Market and sell

Produce and deliver

Invoice and service customers

EXHIBIT 9. BUSINESS SYSTEM DECOMPOSITION

Data and information cost money to obtain andanalyze. Care should be taken to access thebest benchmarking method to derive therequired data at a reasonable cost. In manyinstances, it may be necessary to confirm thedata from several independent sources toensure valid results. These steps take time.

The following critical techniques support research-ing and identifying “best-in-class” performance:

l setting up databases;l information gathering methods;l formatting questionnaires; andl selecting benchmarking partners.

Setting Up DatabasesA database is a useful tool for organizing infor-mation for analysis. Databases are particularlyuseful for coordinating multipartner benchmark-ing studies. They can be used during the actualprocess of collecting data. There are two typesof databases:

l numeric (measures); andl descriptive (practices and enablers).

Numeric databases can be developed with anystandard spreadsheet or database software.Computerized databases allow for easy manipulation of numbers, which comes in handyduring data normalization and gap analysis.Descriptive databases are less rigid. These usually are broken down into summaries of best practices and their supporting enablers.

Creating a database involves compiling the fragments of information found in a variety ofsources, so that a coherent picture of the targetbenchmark companies can be formulated. Oneway to organize this information is to create achart of relevant performance measurements

and process steps for each of the companiesunder consideration.

Exhibit 10 is an example of a typical database.Part 1 records information collected from fourorganizations on the subject of factors affectingdeveloping in-house training material. Thecolumns of the matrices list the four organiza-tions that participated in the benchmarkingstudy, beginning with information from CompanyA conducting the study. The remainder of thecompanies are recorded as letters B through D.This is an example of what is called a blindstudy, in which the organizations participating inthe investigation are not identified.

Also, as the members of the benchmarking teamrecord their responses on these types of matrices,the information can easily be summarized andtabulated in the form of tables.

The five rows of the table record information aboutthe topics measured. A simple review of the infor-mation in the columns allows the benchmarker toidentify trends or numbers that indicate potentialfor further investigation. In addition to the trends,Part 2 also includes practices and enablers(shown for Company B) that contribute to thatorganization’s performance.

Information Gathering MethodsNot all benchmarking requires site visit interviewswith benchmarking partners. The benchmarkingteam will need to decide on the best method forinformation gathering based upon:

l amount and accuracy of information required;l cost of obtaining the desired information;l how the information will be used;l time available to perform the benchmarking

study; andl sources of data and ease of their access.

18


Based upon this, the benchmarking team maydecide on one or more of the following approaches:

l internal information sources;l external information sources; andl original benchmarking research.

It is good practice to explore the internal andexternal public domain sources before conduct-ing an original study. For example, several peopleinside the organization may have access to thedesired benchmark company performance infor-mation. Some general sources of informationare shown in Exhibit 11.

Information in some of these sources, such as Dunand Bradstreet, Value Line, ABI Inform, and Standardand Poor’s, can be accessed easily with the help ofpersonal computers through keyword searches.

In recent years, national and international bench-marking clearinghouses have been set up to sharegeneric benchmarks among a variety of users. TheSociety of Management Accountants of Canada’sbooklet entitled “Benchmarking Information ReferralService,” which lists over 40 clearinghouses world-wide, can help organizations to quickly and efficient-ly track down clearinghouses for benchmarking.

19


EXHIBIT 10: DATABASE

Cycle time 2 days per 1 day per 1 hour 1 hour of material of material

Quality 0.03 errors 0.001 per material errors

per material

# of trainers 12 6

Knowledge medium highlevelrequired

Degree of moderate moderatetechnical content

• Modularized • Database material

• Process of update

• Change control

• Shared data • Trained network

• Input personnel

• Restrict access

• Contract • Personnel people as experts • Good location

• Modern system

Organization A B C D

Practice Enablers

Part 1Subject: Develop In-House Training Material

Part 2Organization B

EXHIBIT 10. DATABASE

If these sources do not meet the organization’srequirements, an original study may be required.This will generally require more time, money, andspecial skills but may be justified if more complexinformation or a higher level of accuracy is desired.

In some cases a combination of two or more infor-mation gathering methods may be necessary(such as a telephone follow-up after a site visit, toclarify key points).

The relative pros and cons of some of the morepopular methods of information gathering are sum-marized by Robert Camp of Xerox in Exhibit 12.

For original benchmarking research, the followingtechniques may be useful:

l Survey: If a mail or telephone survey is the preferred method of benchmarking, the bench-marking team will need to pay particular attentionto:

l selecting the appropriate target population;l obtaining mailing/telephone lists;l validating the quality of the survey list; andl providing an incentive for the target audience to

respond.

l Consulting firms: If a consulting firm needs to behired to assist in information gathering andanalysis, the team can look at several sources,such as: professional associations, consultingassociations, and directories of consultants. Forthe selected consultant, the team will need todefine the:

20


EXHIBIT 11: SOURCES OF BENCHMARK PERFORMANCE INFORMATION

Internal External Original Research

• Sales representatives• Surveys• Tech representatives• Procurement office• Customer visits• Library databases• Recent hires• Competitive analysis

• Mailings• Phone interviews• Consultants• Benchmarking site visit

• Industry publications• Professional associations• Seminars• Industry experts• University sources• Newsletters• Consultants• Financial analysts• National/international clearinghouses• User groups • Security analysts reports• Annual reports• Patent records• Newspapers• Buyers guides• Government documents• On-line services or exchanges

EXHIBIT 11. SOURCES OF BENCHMARK PERFORMANCE INFORMATION

l benchmarking project objectives;l key performance measures and their associat-

ed enablers;l desired industry or "best-in-class" benchmark-

ing partners; andl administrative requirements, such as overall

project schedule, checkpoint reviews, paymentterms, and the respective roles of the consult-ing firm members and the benchmarking team.

l Site visit and interview: This is the most crediblebenchmarking method and will require workbefore, during, and after the benchmarking visit:

l Before the visit, the benchmarking teamshould:l review all relevant data about the organiza-

tion;l prepare the visit objectives;l prepare a list of interview questions;l ensure that the internal process is well

documented and understood;

l identify the appropriate contacts and requesta visit; and

l select the team members who will conductthe site visit (the recommended site visitteam size is two to four).

l During the interview, the site visit team mem-bers should:

l represent themselves honestly and clearlystate the objectives of their visit;

l be prepared to answer why a particularquestion is being asked and also be will-ing to share their own organization's per-formance if asked;

l follow up responses for clarification, ifnecessary, and ensure that the enablersor best practices are understood;

l take notes; andl thank the benchmarking partner and offer

a reciprocal visit.

21

B U S I N E S S P E R F O R M A N C E M A N A G E M E N TEXHIBIT 12: PROS AND CONS OF INFORMATION GATHERING METHODS

Excitement Effort

Sitevisits

High

Low

High

Low

Databasesearches for best

practices

Secondary research/public domaininformation

International information/reports

Scanning and monitoring throughjournals/magazines/newspapers/conferences

EXHIBIT 12. PROS AND CONS OF INFORMATION GATHERING METHODS

l After the interview, the team should:l debrief the site visit team members; andl document the site visit findings.

The benchmarking team should be aware of thelegal and ethical considerations in data gathering.For example, covert photography is illegal. Mostcompanies have written policies regarding suchinformation sharing. In general, the followingpractices should be observed:

l Do not misrepresent yourself, your company,or the purpose of your research.

l Do not entice others to divulge informationthrough illegal means.

l Do not ask for or obtain data on proprietaryproducts or processes.

A good guideline is “not to do to your benchmarkingpartner what you wouldn’t want them to do to you.”

An excellent guide on accepted moral standardis the Professional Code of Conduct jointlyapproved by The Strategic Planning Institute’sCouncil on Benchmarking and The AmericanProductivity and Quality Center’s InternationalBenchmarking Clearing House; it is included inAppendix A. The code summarizes the protocolof benchmarking. There are similar guidelinesdeveloped by Xerox, AT&T and other companies

In addition to these guidelines, there are legalconsiderations that may govern the relationshipbetween benchmarking partners such asantitrust laws, industrial espionage, and restric-tive clauses related to intellectual property.When in doubt, the benchmarking team shouldconsult their company’s legal department.

Formatting QuestionnairesRegardless of the data gathering methodsemployed, they all require the benchmarkingteam to prepare a list of questions in advance toensure a productive outcome.

A questionnaire serves several important purposes. It ensures that all questions of interestare documented. It permits more extensive datagathering, which may not be readily available during a site visit. It can also be used to ensureanonymity. Where confidentiality of the organiza-tion’s sources of information is necessary,questionnaires may be the only method ofobtaining cooperative data gathering.

The questions will typically be about the following:

l problem definition;l process measurement;l process problems;l process improvements; andl process enablers.

For each of these areas, the team may selectone or more of the following types of questions:

l open-ended;l multiple choice;l forced choice; andl scaled.

For the selected type of questions, the phrasingcan also dictate the responses. For example, foran open-ended questionnaire; balanced phrasingmay elicit different responses than extremephrasing:

l How satisfied are you with the response timefrom your service vendor? (balanced)

l What is the best response time you can expectfrom your service vendor? (extreme)

22


Each has its merits and should be reviewed forapplicability to the data and information beinggathered. The company’s market researchdepartment can be an additional resource to theteam in structuring and conducting the question-naire, survey, and interview.

A mailed survey should be easy to fill out andshould communicate to the recipient exactly thetype of information sought. Fowler identifies fourpractical standards that all questions shouldmeet:

l Can this question be asked exactly the way itis written?

l Will this question mean the same thing toeveryone?

l Can people answer this question?l Will people be willing to answer this question,

given the data collection procedures?

One way to ensure this is to pretest the questionnaire by administering it to the internalprocess being benchmarked. This will helpsequence and polish the questions. Thesequence and appearance of the questions in asurvey can affect the responses given. Manydesign guides suggest starting with the easyquestions first to ease the respondents into thestudy. A second approach is to put the least sensitive questions first, gaining commitmentfrom the respondent before asking for the hard facts.

In deciding upon the sequence of questions, oneshould sort them into logical groupings (basedupon querying the same subject), and thenarrange them so that early questions lead naturally to the mindset needed to understandand respond appropriately to later ones. Simplequestions are those that are immediately under-stood and have no trace of ambiguity around

them. An example of a simple question is “Whatis the number of employees in your accountspayable function?” A not so simple question maybe, “What is the relationship between theaccounts payable and the procurement organiza-tions?” The ideal is to make every question“simple.” Failing that, it is important to ensurethat the question, even though complex, can beeasily understood and answered.

In addition to the questionnaire, providing askeletal process flowchart can help give therecipient a graphic reference.

Selecting Benchmarking PartnersOnce it has been determined if the output is abenchmark against internal, competitive, orworld-class organizations, the organizations toinclude in the benchmarking study need to beidentified. For example, if the output is world-class benchmarking in the area of consistentand timely service, the benchmarking team mayconsider companies in the express deliveryindustry.

In order to identify potential benchmarking part-ners, it will be necessary to define:

l “the best” for the performance to be bench-marked;

l trends in the key performance measuresselected; and

l trade-offs between key measures, if applicable.

Based upon this, the team should develop a profile for each of the potential benchmarkingpartners, as shown in Exhibit 13.

A potential partners profile enables an assess-ment of potential partners across the opera-tional and business variables that affect theimplementation of best practices. While the

23


ideal is a partner similar in size and structure,the team can also use the profile to normalizedata from organizations that aren’t exact matches.

For the prioritized benchmarking partners, theteam can then develop a best-practices matrix,as shown in Exhibit 14. The best-practices matrixis a second-stage screening tool for narrowing

down the list of potential partners. It shares theteam judgment on the selected practices foreach partner. In Exhibit 14, Company D is likelyto be a better benchmarking partner. Partnerscan be rated on any criteria that the team considers to be best practice.

24


EXHIBIT 13: POTENTIAL PARTNERS PROFILE

Business Measure Partner A Partner B Partner C

Type of business manufacturing holding company manufacturing

Business size– revenue $210 M $1,800 M $900 M_ # of employees 2,000 18,000 8,900

Ownership ofbusiness public public family

Organizational single site, multinational, multi site,structure single business multi site single business

Geographic Europe, Northrepresentation Canada America, Asia North America

low and mid rangeProduct array high-end products, complete product products, low costand complexity high functionality array and complexity

Product technology state of the art mixed mixed

Company culture open, TQM based centrally controlled TQM based

Key performanceindicators– R&D/revenue 10% 6% 8%– ROS 11% 5% 6%– Market share 21% 40% 20%– Debt/equity 40% 33% 30%

Distribution channels mostly dealers company and stores dealers and OEMs

EXHIBIT 13. POTENTIAL PARTNERS PROFILE

When all sources of external information havebeen exhausted, the benchmarking team is readyto propose the companies for the site visits. Todetermine which companies should be visited,the benchmarking team should compare theinformation it has gathered to decide which companies exceed its own organization’s performance in the chosen process. The follow-ing questions are usually addressed during this stage:

l Is there sufficient information to indicate thata performance difference exists?

l How recent and reliable is the information?l Are the companies sufficiently similar to

warrant a comparison?l Is the proposed company likely to be willing

to share the details of its process?

To help prepare a list of companies that are con-sidered to be industry leaders or “best-in-class,”the team should consult:

l customers;l members of professional or trade associations;l securities analysts;l business directories; andl people within the organization.

It should be recognized that no company, howeversuccessful, is best at everything it does. Even acompany with market share or quality advantageis unlikely to be the best at everything. The teamshould, therefore, resist benchmarking companiesjust because some other organization found that company to be the benchmark for that functional area.

If unable to find the best performer, the bench-marking team should consider the help of a business consultant who understands the enterprise's business. This may especially be necessary if the company being sought is outside the industry.

25

B U S I N E S S P E R F O R M A N C E M A N A G E M E N TEXHIBIT 14: BEST PRACTICES MATRIX

Benchmarking partner A B C D E

Process results 3 4 2 5 1

Process innovation 4 4 2 4 2

Customer satisfaction 3 4 2 4 2

Market share 2 3 1 3 2

Process maturity 3 2 3 2 3

Quality 2 4 2 4 2

Continuous productivity improvement efforts 2 3 1 4 1

Totals 19 24 13 26 13

(SCALE: 1 = Low, 5 = High)

EXHIBIT 14. BEST PRACTICES MATRIX

Analyzing Benchmarking Data and Identifying EnablersAfter the data have been collected and summa-rized, teams are ready for the next step in thebenchmarking study data analysis. The underly-ing objectives in data analysis are to identify performance gaps, understand the reasons forthe performance differences, and prioritize thesekey causals and enablers for implementation.

Truly sophisticated benchmarking studies takeperformance analysis beyond the realm of theaverage benchmarking team and into the worldof statistics and business analysis, using suchtools as:

l statistical analysis;l data stratification techniques;l data normalization techniques;l radar charts;l force field analysis;l performance gap analysis tools; andl identifying process enablers.

One or more of these tools may be needed whenthe study involves:

l a large number of variables;l testing of assumptions; orl presenting objective, quantitative results.

Statistical AnalysisGraphical or numerical methods can assist the benchmarking team to analyze a process orpopulation of events. Often these tools use asampling of the population to make a judgmentabout the entire population.

Listed below are the six most commonly usedstatistical tools for analytical problem solving:

l Check sheet: A check sheet is used to recorddata. It is useful in the data collection stage ofbenchmarking studies. The best check sheetsare simple to use, make use of the organiza-tion’s operational definitions (describes whatsomething is and how it is measured), arid display the data in a format that can revealunderlying patterns.

l Histogram: A graphical representation of dataas a frequency distribution, this tool is valuablein evaluating both attribute (pass/fail) and variable (measurement) data. Histograms offera quick look at the data at a point in time; theydo not display variance or trends over time. A histogram displays how the cumulative data look today. It is useful in understandingthe relative frequencies (percentages) or frequency (numbers) of the data and howthose data are distributed.

l Pareto diagram: This is a type of histogramthat helps to identify and prioritize problemareas. Pareto analysis consists of five steps:l identifying the occurrence, measurement, or

non-conformity for analysis;l determining the frequencies of the data;l calculating the frequency percentages and

listing the frequencies in descending order;l determining the scale for the Pareto

diagram; andl plotting the results on a histogram.

Pareto charts are useful throughout a bench-marking study: early on to identify which prob-lem should be studied, and later to narrowdown which causes of the problem to addressfirst. Since they draw everyone’s attention tothe “vital few” important factors where thepayback is likely to be greatest, Pareto chartscan be used to build consensus in a group.

26


l Control chart: This provides a graphic depictionof the quantified characteristics of a process,process element, or work activity. Controlcharts display the plotted values of theprocess and indicate if the process isapproaching an established limit. For bench-marking studies, control charts can determineif a process is trending and determine if aprocess is in control.

l Scatter diagram: A scatter diagram organizesdata using two variables: an independent variable and a dependent variable. These dataare then recorded on a simple graph with x and y coordinates showing the relationshipbetween variables.

l Linear graph: This is a series of numberedlines and data with a one-to-one correspon-dence to the columns of a related matrix ofnumbers arranged in rows and columns.

Data Stratification TechniquesData stratification is the process of classifyingdata into two or more subgroups based upon categories or characteristics. This is a powerfuland frequently used tool in benchmarking studies. When the data are stratified accordingto the variables that are thought to cause variation, the causes of variation can be detectedmore easily by the benchmarking team.

One technique that is easy, as well as helpful for stratifying data, is a root cause analysis, orfishbone, as it is sometimes called. An applica-tion is illustrated in Exhibit 15. The effect is usually indicated at the head of the diagram andthe potential causes are listed on the fishboneson the left. This method can help the benchmark-ing team build an intuitive understanding of themany interrelational possibilities and it also provides strong visual clues to interrelationshippossibilities.

27


EXHIBIT 15: ROOT CAUSE ANALYSIS

Market understanding

Pricing

Relativ

e pr

oduc

t qua

lity

and

func

tiona

lity

Adve

rtisin

g bu

dget

Number of sales reps.

Marketshare(effect)

EXHIBIT 15. ROOT CAUSE ANALYSIS

If the root cause analysis shows a large numberof variables for which the precise impact on theeffect is desired, the team should follow up this analysis with more sophisticated analysisusing design of experiments, analysis of variance, regression analysis, or some other formof multivariate analysis.

Normalizing Data for ComparisonOperations must be comparable in scope toobtain valid benchmark data. There is a tenden-cy to accept data and information that arebelieved to be comparable, especially from exter-nal visits. This could lead to the acceptance ofan observed performance as a correct statementof the benchmark. This measure may be closelytied to the internal processes but may be inap-propriate for benclimarking comparison withother groups who deploy different internalprocesses. This is shown below.

Company X Company YDirect labor cost $200 m $75 mOverhead rate 15% 30%

In this example, Company X appears to be twiceas efficient as Company Y based upon the laboroverhead rates. However, if Company Y hasinvested in automation to reduce labor content,then the overhead rate may be the direct resultof depreciation of capital investments that isallocated to a smaller direct labor cost pool. Theoverall manufacturing cost for Company Y maybe indeed lower.

If performance measures are not comparable,benchmarking teams should normalize the per-formance data so that they can make accurateassessments. Normalization factors are oftenused to normalize performance data on a ratiobasis to avoid “apples to oranges” comparisons.The common normalization factors are:

l size (for example, revenue/employee,sales/sq.ft., maintenance calls/technician,sales calls closed/rep);

l age (for example, plant age, generation of tech-nology); and

l working environment (for example, number ofshifts, make vs buy, amount of competition,general economy, access to raw material,degree of vertical integration, regulatory differ-ences, tax treatments, exchange rates).

In addition, normalization factors such asaccounting methods and calculation methodologywill also need to be considered.

Radar ChartsRadar charts can help assess an organization’srelative competitive position across the key perfor-mance measures. It is an excellent comparisonand communication tool.

Typically, it consists of various spokes within acircle. Each spoke represents a key measure.Assessment criteria are represented by circulargrades that gradually improve as the radius getssmaller, the centre signifying “best-in-class” per-formance. The relative value of a competitor’sstrength is measured on each spoke, as shownin Exhibit 16. It shows the survey results of acompany's service performance, on a scale of 1to 5 for the five measures listed. For example,Organization A scores 1 (excellent) on responsetime compared to Organization B, which scores 3(average).

28


Identify Performance GapsThe basic tool for analyzing performance andprocess gaps is a matrix chart with a list of per-formance measures for each company examinedas well as the internal performance. Using thisformat, many performance and process gaps areimmediately evident, as illustrated in Exhibit 17for a manufacturing process.

In evaluating the performance gaps, the bench-marking team should consider all enablers or

factors that influence the process, such as material movement, data, and resource flows.Determining the current level of performance isless important than understanding the trend inthat performance. Instead of aiming at today’starget, teams should project the benchmark intothe future to understand what level of performancewill be required and what enablers may helpthem attain that level.

29


EXHIBIT 16: RADAR CHART

Product reliability

(Scale: 1 = excellent, 5 = needs improvement)

Partssupport

Ability to repairfirst time

OrganizationA profile

1

2

3

4

5

Service technicianprofessionalism

OrganizationB profile

Responsetime

EXHIBIT 16. RADAR CHART

Benchmarking teams can use a Z chart to displayhistorical trend information and to project the historical information estimates of future perfor-mance levels required to achieve competitive parity.

Exhibit 18 shows an example of a gap analysis.As can be seen from the exhibit, the Z chart ismade up of three parts:

l the historical productivity trend for the perfor-mance measure being benchmarked;

l the benchmark performance gap, which isshown as a step function (these may requirestrategic actions to close the gap); and

l the projected future productivity to either main-tain or retain superior performance.

30


EXHIBIT 17: BUSINESS PERFORMANCE COMPARISON

Benchmarkingorganization

Organization

Parts per million defect rate 750 500 400 1,250 500

Overhead rates 22.5% 30% 15% 17% 17%

Labor rate $20.00 $19.50 $17.50 $15.00 $17.50

Output/hour(units/hour) 500 500 300 400 400

Cost of production peremployee $100K $90K $120K $110K $110K

Absentee rate(days/year) 3 7 4 3 4

Performancemeasure

PartnerA

PartnerB

PartnerC

Medianpartners

EXHIBIT 17. BUSINESS PERFORMANCE COMPARISON

EXHIBIT 18: Z CHART

1200

1000

800

600

400

200

1 2

Year

Defective parts permillion

Historical trend

Future productivity needed

Benchmark gap(step change needed)

Currentbenchmark

3 4

EXHIBIT 18. Z CHART

Identifying Process EnablersHaving analyzed process performance measuresin the benchmarking gap analysis, the team mustalso identify the key process enablers. Key processenablers are those activities that facilitate thekey behavioral or process changes; those activi-ties in the root cause analysis identified as thestimulants of the performance change. This taskof identifying enablers is actually a simple rank-ing of the enablers, assigning priority based oninformation received during the site visits.

Enablers are helpful in the implementation ofbenchmarking practices, but they should not beconfused with the practices themselves. Thebenchmark practices are specific new methodsor practices that require a change to meet a stated goal. Enablers are a broad set of activi-ties that enhance implementability. In sequenceof priority they can be shown below:

.

Force Field AnalysisThis approach identifies those forces that bothhelp and hinder an organization in closing thegap between where it currently is and where itwants to be. This method enables the organiza-tion of perceptions together with impacts so thatthe benchmarking team can uncover importantrelationships between drivers and barriers.

The steps to carrying out a force-field analysisare as follows. A line is drawn down the centre ofa flip chart page. This represents the present situation. At the right edge of the sheet, a secondvertical line parallel to the first is drawn. Thisrepresents the situation as it should be—thedesired state. After using one or more of thetools for generating and collecting information,the benchmarking team lists the helping forcesto the left of the centre line and the hinderingforces to the right. These forces are often shownas arrows, with the helping forces pushingtoward the “should be” state and the hinderingforces pushing away from it, as shown in Exhibit 19. It is often helpful to assess the relative strengths of both helping and hinderingforces. Some teams use a scale (for example,5 = very strong, 4 = strong, 3 = medium, 2 = low,1 = weak) to evaluate the relative impact of theforces. For graphic representation, proportionatelysized arrows show relative strengths.

31


HIERARCHY

Goal

Benchmark practice

Enablers

Reduced clericaldocumentation workload

Use of bar coding for automatic data capture

Training in useof scanners

EXHIBIT 19: FORCE FIELD DIAGRAM

Helping forces Hindering forces

As is Should be

EXHIBIT 19. FORCE FIELD DIAGRAM

Exhibit 20 shows an assessment of a product lineversus its competitors. The analysis graphicallydetails the helping and hindering forces that willaffect the implementation of “best practices.” Theorganization can capitalize on its strengths inproduct features, quality, reliability, marketing,distribution channels, and after-sales support. Itneeds to overcome its disadvantages in price andproduct costs compared to its competitors.

Once the force field analysis is complete, theteam can use this information to generate poten-tial solutions. For example:

l how to increase the number or strengths of thehelping forces; and

l how to decrease the number or strengths ofthe hindering forces.

Implementing Benchmarking Study RecommendationsThis final phase of the benchmarking process iswhere leadership abilities are most crucial. Theteam must be able to justify its recommendedapproach for improvement to senior manage-

ment. As well, process improvement teams mustbe able to manage change implementation andtrack their progress.

Obtaining Management Buy-inAs critical as the preceding benchmarking phasesare, so too is the need to effectively structurethe project findings and recommendations toensure management sponsorship. Generally, themagnitude of the changes will determine theextent of education of and communication to allthose directly involved. Common to all projectcommunications, however, are the following elements:

l the current internal performance and situation;l the benchmarking project objectives;l the team profile;l the benchmarking partners and the rationale

for their selection;l basic findings and their causal factors;l the benchmark gaps between the current and

desired state;l interim and long-term goals leading to the

desired state and their implications;

32

B U S I N E S S P E R F O R M A N C E M A N A G E M E N TEXHIBIT 20: FORCE FIELD ANALYSIS

(+) helping forces

Product features

(-) hindering forces

Uncompetitive price

Product costProduct quality,reliability

Product marketingcampaign

Number of outletsselling products

After-sales support

EXHIBIT 20. FORCE FIELD ANALYSIS

l comparison of costs associated with the present state and the desired state (savings),including the cost of change;

l action plans, investments, schedules, andresponsibilities; and

l feedback and monitoring systems.

Organizing Process Improvement Teams(PITs)The benchmarking team recommendations mayrequire a different team structure for implemen-tation. In these cases, the benchmarking team,working with senior management, should:

l identify criteria for selecting team membersconsistent with the recommendations. Thiscould include the recommended team membersand their roles in the implementation phase;

l define the team mission, objectives, and anyconstraints, such as time and costs; and

l hold a kick-off meeting with the processimprovement team to share benchmarkingfindings and formulate the basis for recom-mendations.

The process owner of the PIT will need then todevelop a structured approach to implementa-tion with applicable schedules, key tasks,resources, and investments.

Implementing ImprovementsThere are basically three alternative approachesto improving high-priority processes: incrementalimprovement, redesign, and reengineering. Thespecific approach will depend upon the cost/benefit justification.

l Incremental improvement consists of small,often sequential improvements in current workprocesses, often in response to problemsidentified by self-empowered work teams.

l Redesign of the process is a higher level ofimprovement in which management examinesthe process as a whole to determine if it canbe done better. Redesigning must be done bymid-level management who collectively ownthe entire process.

l Reengineering is a still-higher level of improve-ment in which organizations need to recon-ceive the entire way of doing business. This isparticularly useful when stakeholders areextremely dissatisfied, when there are con-flicts one cannot resolve within the current wayof running the business, or when the processsimply doesn’t exist.

Tracking and Reporting ProgressAfter the implementation phase begins, it isimportant that the benchmarking team, or theprocess owner, monitor progress against themileposts established in the action plans andtake corrective actions if necessary. Managementshould be kept apprised of the status to ensurecontinual involvement and commitment.

V I I . MANAGEMENT ACCOUNTING CHALLENGESManagement accountants have a significant rolein ensuring that “best practices” are institution-alized throughout the organization. In this role,the management accountant will find challengesin several areas:

l Strategic planning: Since the evolving role ofmanagement accountants takes them into therealm of strategic planning, they need to fullyunderstand the processes being bench-marked, and not just from a financial perspec-tive. They will need to be fully conversant withthe company's strategic intent, core competen-cies, process capability, product and R&D port-folio, key environmental and industry trends,

33


customer expectations, and the long-rangevision of the organization.

l Benchmarking training: The managementaccountant will need to participate in and be achampion for training the various levels of theorganization in benchmarking concepts, tools,and techniques to create a common knowl-edge base. Training should be from the topdown: managers who participate in staff training reinforce the importance the organiza-tion attaches to benchmarking. The trainingprogram should also identify contacts for further information and training and, if applicable, explain the company-wide bench-marking database.

l Business practices: Different companies, anddifferent functions within the same companies,are likely to use different accounting practices.If not understood, the benchmark results couldbe very misleading. The management account-ant will need to know how these figures werederived and normalize them for comparison. Inaddition to financial practices, the managementaccountant will need to normalize for businesspractices, such as number of shifts, regulatoryclimate, exchange rates, make vs. buys, compe-tition, and degree of verticalization.

l Performance reporting: As the benchmarkingprocess begins to be institutionalized, themanagement accountant will need to use thetools and techniques described in the guide-line to modify the existing costing and report-ing systems to create a system of benchmarkgap analysis and to evaluate exactly what isrequired by each organization to meet thebenchmark performance. As well, the manage-ment accountant will need to integrate thebenchmark information with other competitiveinformation to provide a complete picture ofoptions, and to establish reasonable goalsand implementation strategies to achievethese goals.

l Measurement systems: Crucial to the success ofbenchmarking is the ability of managementaccountants to develop a better understandingof their internal and external customer needsand expectations, and to develop measuresthat truly reflect these expectations. Frequently,companies find that traditional measures arenot only inadequate, but misleading and mustbe overhauled and discarded. They should ensurethat the organization performance measuresare based upon the viewpoints of customers,employees, and shareholders, and upon theneed for continuous improvement.

l Performance gap analysis: Managementaccountants will need to facilitate the perfor-mance gap analysis activity to determine whatthe organization lacks to move from one pointto the other, or to leapfrog the competition tobecome the new industry leader. To close theperformance gap, they should ensure that theorganization's goals, mission, and objectivesare tied to the benchmarking process.

l Other skills: To meet the above challenges, man-agement accountants will also need to enhancetheir proficiency in several areas, such as projectmanagement, problem solving skills, meetingmanagement, communication, TQM principlesand methods, and interviewing skills.

V I I I . CONCLUSIONBenchmarking, if applied properly, is a powerfultool with which to keep an organization competi-tive. The tools and techniques described in thisSMA enable an organization to effectively bench-mark and transform themselves to meet thechallenges of today's competitive environment.

The organizations that understand, embrace,and implement benchmarking and the relatedtools and techniques will be able to celebratethe turn of the millennium.

34


APPENDIX

THE BENCHMARKINGCODE OF CONDUCTTo contribute to efficient, effective and ethicalbenchmarking, individuals agree for themselvesand their organization to abide by the following prin-ciples for benchmarking with other organizations.

Principle of Legality. Avoid discussions oractions that might lead to or imply an interest inrestraint of trade: market or customer allocationsschemes, price fixing, dealing arrangements, bidrigging, bribery, or misappropriation. Do not discuss costs with competitors if costs are anelement of pricing.

Principle of Exchange. Be willing to provide thesame level of information that you request in anybenchmarking exchange.

Principle of Confidentiality. Treat benchmark-ing interchange as something confidential to theindividuals and organizations involved.Information obtained must not be communicatedoutside the partnering organizations withoutprior consent of participating benchmarking part-ners. An organization’s participation in a studyshould not be communicated externally withouttheir permission.

Principle of Use. Use information obtainedthrough benchmarking partnering only for thepurpose of improvement of operations within thepartnering companies themselves. External useor communication of a benchmarking partner’sname with the data or observed practicesrequires permission of that partner. Do not, as aconsultant or client, extend one company’sbenchmarking study findings to another withoutthe first company’s permission.

Principle of First Party Contact. Initiate contacts, wherever possible, through a bench-marking contact designated by the partner company.Obtain mutual agreement with the contact onany hand-off of communication or responsibilityto other parties.

Principle of Third Party Contact. Obtain anindividual’s permission before providing theirname in response to a contact request.

Principle of Preparation. Demonstrate commitment to the efficiency and effectivenessof the benchmarking process with adequatepreparation at each process step, particularly at initial partnering contact.

ET IQUETTE AND ETHICSIn actions between benchmarking partners,the emphasis is on openness and trust. The following guidelines apply to both partners in abenchmarking encounter:

l In benchmarking with competitors, establishspecific ground rules up front, e.g., “We don'twant to talk about those things that will giveeither of us a competitive advantage, rather,we want to see where we both can mutuallyimprove or gain benefit.”

l Do not ask competitors for sensitive data orcause the benchmarking partner to feel thatsensitive data must be provided to keep theprocess going.

l Use an ethical third party to assemble andblind competitive data, with inputs from legalcounsel, for direct competitor comparisons.

l Consult with legal counsel if any information-gathering procedure is in doubt, e.g., beforecontacting a direct competitor.

l Any information obtained from a benchmarkingpartner should be treated as internal, privi-leged information.

35


l Do not:l Disparage a competitor's business or opera-

tions to a third party.l Attempt to limit competition or gain busi-

ness through the benchmarking relationship.l Misrepresent oneself as working for another

employer.

BENCHMARKING EXCHANGE PROTOCOLAs the benchmarking process proceeds to theexchange of information, benchmarkers areexpected to:

l Know and abide by The Benchmarking Code of Conduct.

l Have basic knowledge of benchmarking andfollow a benchmarking process.

l Have determined what to benchmark, identifiedkey performance variables, recognized superiorperforming companies, and completed a rigorousself-assessment.

l Have developed a questionnaire and interviewguide, and will share these in advance ifrequested.

l Have the authority to share information.l Work through a specified host and mutually

agree on scheduling and meeting arrangements.

l Follow these guidelines in face-to-face site visits:l Provide a meeting agenda in advance.l Be professional, honest, courteous and

prompt.l Introduce all attendees and explain why they

are present.l Adhere to the agenda: maintain focus on

benchmarking issues.l Use language that is universal, not one's

own jargon.l Do not share proprietary information without

prior approval from the proper authority, ofboth parties.

l Share information about your process, ifasked, and consider sharing study results.

l Offer to set up a reciprocal visit.l Conclude meetings and visits on schedule.l Thank the benchmarking partner for the time

and for the sharing.

36


This code of conduct was co-authored by The AmericanProductivity and Quality Centers International BenchmarkingClearinghouse and The Strategic Planning Institute's Councilon Benchmarking, 3/1/92.

GLOSSARYBaselining – This is a fully systematic approach to

mapping the current business process. Itbegins by gathering information about the cur-rent business operations and creating amodel of the business processes being studied. The mapping process describes thework process flow, cycle times or completiontime associated with each work element,baseline transaction volume, and cost foreach key element of the work process.

Competitive benchmarks – These are the bestproducts, services, and work processes of anorganization’s strongest direct competitors inthe industry. Competitive Benchmarking isuseful in positioning an organization’s prod-ucts and services relative to the marketplace.

Critical success factors – These are the criticalvariables that have a direct influence onachieving the organization’s objectives. Theyare, therefore, critical to the success of theentire organization.

Hand-offs – These are the passing of a workprocess across the organizational boundariesas the work elements flow from one internalfunctional organization to the other (e.g., theincoming parts may require hand-offsbetween receiving, quality, storage, andaccounts payable). The product developmentprocess requires hand-offs between market-ing, design, and manufacturing.

Internal benchmarks – These are the best of anorganization’s own similar processes, prod-ucts, or services. This is perhaps the easiestform of benchmarking, since the potentialbenchmarking partners can be easily identi-fied and are usually willing to share the infor-mation.

Organizational seams – Similar to the hand-offsconcept, the organizational seams are thewhite spaces between the vertical functionalorganizations requiring hand-offs.

World-class benchmarks – These are most difficulttypes of benchmarks because they compareperformance to the world-class organizationsthat may be in a different industry but performa similar process. However, these can providebreakthrough opportunities.

37


BIBL IOGRAPHYAmerican Productivity & Quality Centre. 1994. The

Best of Benchmarking: A Best Practices Guide.Houston, TX: American Productivity & QualityCentre.

Balm, Gerald J. 1992. Benchmarking: APractitioner’s Guide for Becoming and StayingBest of the Best. Schaumburg, IL: QPMAPress.

Bogan, Christopher, and Michael English. 1994.Benchmarking for Best Practices: WinningThrough Innovative Adaptation. New York, NY:McGraw Hill.

Camp, Robert C. 1989. Benchmarking: The Searchfor Industry Best Practices That Lead toSuperior Performance. Milwaukee, WI: ASQCQuality Press.

Camp, Robert C. 1995 Business ProcessBenchmarking: Finding and Implementing BestPractices. Milwaukee, WI: ASQC Quality Press.

Deming, Edwards W. 1982. Quality, Productivityand Competitive Position. Cambridge MA:Centre for Advanced Engineering Study,Massachusetts Institute of Technology.

Eyrich, HG. 1991. “Benchmarking to become bestof breed.” Manufacturing Systems. April.

Fowler, Floyd J., Jr. 1984. Survey ResearchMethods. Beverly Hills, CA: Sage Publications.p. 99.

Fuld, Leanard M. 1989. Monitoring theCompetition: Find Out What Is Really Going OnOver There. New York, NY: John Wiley & Sons.

Hammer, Michael. 1990. “Re-engineering work:Don’t automate, obliterate." Harvard BusinessReview. July/August.

Harrington, James H. 1990. Benchmarking. LosGatos, CA: Ernst & Young.

Harrington, James H. 1991. Business ProcessImprovement. Milwaukee, WI: AmericanSociety for Quality Control.

Imai, Masaaki. 1986. Kaizen: The Key to JapaneseCompetitive Strength. New York, NY: RandomHouse.

International Benchmarking Clearing House.1992. Planning, Organizing, And ManagingBenchmarking Activities: User’s Guide.Houston, TX: American Productivity & QualityCentre.

Porter, Michael. 1985. Competitive Advantage:Creating and Sustaining SuperiorPerformance. New York, NY: Free Press.

Saaty, Thomas L. 1988. “Decision making forleaders: The analytical hierarchial process fordecisions in a complex world”.

Society of Management Accountants of Canada,The. 1991. Implementing Benchmarking.Hamilton, ON: The Society of ManagementAccountants of Canada.

Society of Management Accountants of Canada,The. 1993. Managing Quality Improvements.Hamilton, ON: The Society of ManagementAccountants of Canada.

Spendolini, Michael J. 1992. The BenchmarkingBook. New York: AMACOM/ The AmericanManagement Association.

Watson, Greg. 1992. The BenchmarkingWorkbook: Adapting Best Practices forPerformance Improvement. Cambridge, MA:Productivity Press.

Watson, Gregory H. 1994. StrategicBenchmarking. New York, NY: John Wiley andSons, Inc.

Zivan, Sy. 1992. “Are you ready for benchmarking?”Distribution. March.

38


Documents

Effective Benchmarking