Copyright
Thomas A Barocci.

The information systems decision online

. (page 1 of 3)
Online LibraryThomas A BarocciThe information systems decision → online text (page 1 of 3)
Font size
QR-code for this ebook


)O0



1 3(sD I'S^



WORKING PAPER
ALFRED P. SLOAN SCHOOL OF MANAGEMENT



THE INFORMATION SYSTEMS DECISION

By

Thomas A. Baroccl*, Mary Thron
Rolf Gaertner and Klrsten R. Wever



WP #1361-82



JANUARY 1983



MASSACHUSETTS

INSTITUTE OF TECHNOLOGY

50 MEMORIAL DRIVE

CAMBRIDGE, MASSACHUSETTS 02139



DRAFT - COMMENTS WELCOME



THE INFORMATION SYSTEMS DECISION

By

Thomas A. Barocci*, Mary Thron

Rolf Gaertner and Kirsten R. Wever

WP #1361-82 JANUARY 1983



This report is one of a series under the aegis of the Productivity/
Quality project within MIT's Sloan School of Management. The Project Director
is Thomas A. Barocci, Associate Professor of Management. For further
information contact Ms. Carolanne Foilb, Project Administrator, Sloan School
of Management, MIT, 50 Memorial Drive, E52-454 , Cambridge, MA 02139.



THE INFORMATION SYSTEMS DECISION

By

Thomas A. Baroccl*

Rolf Gaertner

Mary Thron

Klrsten R. Wever



Investment decisions about information systems and computing hardware
are usually made on the basis of faith. Most often, the competitive necessity
to produce efficiently (whether perceived or real) forces the decision.
Managers generally address the problem by simply investing and hoping. But by
projecting and carefully considering the expected effects of the decision they
could sharply reduce the faith component of this investment.

Almost all firms fighting to stay competitive in today's markets have
to make information systems (l/S) investment decisions. At the same time,
they must pay more attention to productivity than they did a decade ago. So
when managers make these decisions, they seek to maximize expected pay-offs,
including productivity improvements that result from l/S installations. But
in order to compare l/S costs with expected benefits, managers will have to
use benchmark productivity measurements by which they can later guage the
value of an l/S investment. The point of l/S is not to confuse the



*Thomas A. Barocci is Associate Professor of Management, MJT's Sloan School of
Management. Rolf Gaertner and Mary Thron are former graduate students and
Klrsten R, Wever is currently a Ph.D. candidate and research assistant on the
Productivity/Quality Project. This report is in part based on Rolf Gaertner
and Mary Thron's unpublished Master thesis, entitled "Productivity of
Information Systems In Manufacturing and Engineering." The authors gratefully
acknowledge the time and financial support of the corporate study
participants. Without their time, advice and financial support this research
would not have been possible.



productivity Issue, but to raise productivity. The most fundamental challenge
facing management Is thus the necessity to view l/S and productivity in
conjuntion with one another .

The productivity gap between the U.S. and its major competitors is
widening rapidly. "Computerization," the primary concern of the seventies,
has resulted in growing l/S costs, low software productivity, and the problems
of coordinating complex computer tasks within and outside the firm. So the
main problem we consider here is how l/S affect end user productivity. Our
five cases are in the high-technology electronics and banking Industries. All
of the firms are large multinationals. We studied them by interviewing about
ten people in each company and collecting around 100 questionnaires.

In the electronics industry, we focused on manufacturing and
engineering — both areas in which recent investments in l/S are common. In
banking, computerization is noticeably on the rise. Manufacturing l/S include
material requirements planning (MRP), Inventory control, and shop floor
scheduling control. In engineering we examined Computer Aided Design (CAD)
systems, used to design printed circuit boards. Integrated circuits (IC, LSIC)
fnd mechanical parts. In banking we looked at how recent computerization
(analagous to plant automation, rather than manufacturing control) affect
process automation to increase productivity. The questions we addressed in
all three areas were how useful these systems were in producing information on
productivity, and how that Information was used.

The body of our analysis is structured as follows. After presenting
our company overviews, we first consider problems of measurement and the
effects of these on productivity. We then look at Intangibles that must be
addressed with qualitative judgements. Finally, we discuss the ways in which
Investment philosophies — justifications for the use of l/S — can affect the
usefulness of the systems and their impact on the firm and its productivity.

-2-



One major theme of our analysis Is that a really insightful
productivity analysis must consider outcomes rather than simply outputs.
Outputs are units of products weighted by their respective prices. Output
attributes and process constraints affect the outcome of the productivity
project. Product attributes that typically affect a productivity outcome
include quality, timeliness, adherence to specifications, variety of options,
user satisfaction, and the effects of engineering changes and updates on
customers, service, spare parts, training and documentation. Process
constraints are secondary effects encountered during production, which are
imposed by the interaction of the firm with its environment. Examples include
environmental regulations, antitrust laws, miscellaneous legal requirements,
financial obligations, and foreign exchange rates. The importance of
considering outcomes rather than outputs lies in the simple fact that the
productivity v;hole (outcome) is greater than the sum of its parts (outputs).
Tangible changes in, say, the capital/labor mix, are paralleled by intangible
changes in outcome; the latter are much harder to measure, partly because
their effects are usually felt most strongly in the long-run.*

Productivity is measured in ratios; the formula is generally output
divided by input in physical units, weighted in market dollar values. But
these ratios are often inadequate. And there is no specific, company-wide
nethod of setting up a meaningful system to improve them. Different
industries have different kinds of productivity objectives. They depend, for
example, on whether the



*Michael Packer, "Output-Mapping: A Productivity Analysis Technique
for R&D/Organlzations," Laboratory for Manufacturing and Productivity, M.I.T.,
19G1, conference paper.



-3-



iector Is capital- or laber-lntenslve. Productivity ratios can be measured oo
the basis of partial factor productivities, using only one Input (e.g.,
labor). Or ratios can ust aggregated Inputs (In dollar values); these are
called total, or total factor, productivity measures.

But even the firm with a clear understanding of these differences will
encounter difficulties in at least three areas. First, some kinds of expenses
(e.g., R&D. advertising or maintenance costs) may lower productivity figures
80 as to discourage management to incur these costs. Second, the gap between
necessary and obtainable data la usually large. And finally, exogenous
factors can result In uncontrollable or non-continuous changes In measured
productivity — changes which do not reflect the firm's actual productivity.
Wildly fluctuating Interest rates, for example, have no impact on the actual
productivity of a bank. To address these and other productivity problems
effectively, productivity programs must be related to the firm and function
under consideration. But just as important Is the need for managers to
question purely quantifiable productivity measurements and to take into
account the qualitative aspects of labor and capital investment productivity.

In designing quantitative and qualitative productivity measures, it
should be kept In mind that I/S productivity is a management issue. It is up
to management to design and evaluate efficiency and user satisfaction,
grouping all measurements Into qualitative and quantitative categories. To
reflect the importance of management involvement in the productivity issue, we
have based our study on the flow-chart pictured in Figure I. The chart
Illustrates the primary stages a productivity program goes through from its
inception to various aspects of the outcome.

The complexity of the productivity process and of the measurement not
only of outputs, but of ultimate outcomes, pose many problems that enforce the
haphazard nature of productivity-related l/S investment decisions. But any

-4-



FIGURE



1: The Productivity Process at the Company Level



CORPOPATE STRATEGY



SBU GOALS



PRODUCTIVITY PROGRAM
(CRITICAL SUCCESS FACTORS)



KEY VARIABLES
(INDICATORS)



PRODUCTIVITY OBJECTIVES
(OUTCOME/KEY-VARIABLE)



OUANTITATIVE QUALITATIVE



L



I

IMPROVEMENT PROGRAM
(BUDGETING)



r



CAPITAL
PROJECTS



MONITORING

L_



HUMAN RESOURCE
MANAGEMENT

I



DATA-ORIENTED
(CONTROLLING)



DECISION-ORIENTED
(JUDGEMENTAL)



J



IMPLEMENTATION



STOCKHOLDERS



MARKET FEEDBACK

(NEEDS, WANTS,

SATISFACTION)



INDUSTRY FEEDBACK

(PRICES, SEGMENTATION,

SUPPLY, REACTIONS)



INTER - COMPANY

COMPARISON

K



INTRA - COMPANY
COMPARISON



LEARNING &

EXPERIENCE

I



MEASUREMENTS
(QUANTITATIVE,
QUALITATIVE)
i



INPUTS, OUTPUTS



-5-



viable firm must be efficient, any efficient firm must face the productivity
quandary despite its complexities, and any efficient high-tech engineering or
manufacturing enterprise must link its I/S with larger productivity goals.
This last requirement is worth emphasizing, particularly since the explicit
relationship between l/S and productivity is not of major concern to the firms
covered in this study. The specific ways in which l/S and productivity are
interconnected are illustrated in Figure 2.

We conducted our field research with four large high-technology firms
and one major bank, on the basis of personal interviews and questionnaires
including a company survey, an interview guide and a check-off questionnaire
for each area of concentration. The company survey covered general
information, the firm's culture, productivity programs and measurements, and
reward systems. This information provided us with a feeling for each
company's broad approach to the productivity issue.

The interview guide was designed to give qualitative as well as
quantitative background information about each area we examined. It covered
the description of the systems in use, the relationship between each
functional unit and the l/S department, computer investment history and
philosophy, each unit's definition and measurement of productivity, and how
productivity information was used. We compiled a version of the guide
specific to each area of concentration: manufacturing, engineering, and
banking.

The check-off questionnaire was answered by checking boxes, rather than
writing short answers. It was composed for statistical analysis, using a
seven-point scale to measure perceptions of the factors affecting
productivity, the ways in which information systems affect these factors, the
effectiveness of productivity programs, and the relationship between these and



-6-



FIGURE 2



The Role of Information Systems



OUTCOME



Output
Attributes



Constraints



c

o

•H
-P

m

N

■H

e

■H
40

D.
O



O

CD
r)

-a

CD
CD



ORGANIZATION



I HumaPi




Inertia
Focus



Resistance
to Change



Resource
Manage-
ment




INFORMATION

Investments

Projects

User
Interface



CAM

i >

Manufac-
turing
Control
Systems


PROCESS
Production


Engineering




Test

QAS,R



REST OF
ORGANIZATION

Finance



Sales



Purchasing





Development



Communication



1/S
Development



-7-



other areas (e.g., the connection between an l/S group and manufacturing or
engineering). Again, we used three versions of the questionnaire, one for
each area of concentration.

We began by administering the company survey to a corporate manager
with productivity responsibilities. Then we interviewed an average of 10
people at various levels of plant management, l/S, engineering, corporate
manufacturing services, CAD groups, research and development (R&D), and
corporate engineering support. At the bank, we spoke with corporate staff
support, office managers and systems managers. The interviewees were for the
most part very open, particularly in the high technology firms. The
questionnaire response rate was 80%; the sample from each company is varied.

The firms we studied are referred to as: Atlasbank, a leading
international bank; Comptech, a computer systems designer; Gruber, a producer
of electronic instruments; Ratronics, a defense electronics firm; and Semcon,
a manufacturer of semiconductors. Company overviews appear in Figure 3.

Because the banking and high technology industries are becoming more
and more competitive, their focus on productivity has had to intensify.
Because of the low direct labor content of high technology products, and the
high percentage of knowledge workers involved, productivity programs in high
technology enterprises emphasize overhead costs. In banking, the focus Is on
automating manual processing at the lowest levels. Most firms use the
classical measure of productivity: output per employee hour. This measure is
simple, controllable, and actionable at low levels. But the four high
technology firms are also striving for quality-related productivity goals.
The general productivity programs and their situational settings are laid out
Jn Figure A,

Figure 4 illuminates several interesting trends. It is apparent, for
example, that Ratronics' hands are more or less tied at every level by the
narrowly defined standards and requirements of its government contracts. This
firm is not very eager to experiment with productivity. Its connection with



O M


33






w


cn 13


in


K Z


M






§


M »


M


ss a


cn






w o


tsi


3 tn


O






B


< o

M G


W


M H


»






K


n n




n »


K






PI


n 1^




W K








M
in


\




1 1
1 1


1 1
1 1

O S (11




a


1 1 II
1 1 II

Ml 'J pi rt 3'


1
1


1
1

cr ni


ft n> s


com




M


M ra o c p- <


Si


p- in


fD ft n


3 n ft




P


(B 3 3 M iQ O


^


p* in


m HI",






<


,a ft u1 3 3' M


p* (B


Qt n o




O


ri 1


p-


P- rt


k: H- (u


3 S H-




in


(B rf t: O P- i/v


3


O in


Ul






3 (B 'O < O


■a


3 „


kO 3 H-


fu pi in






ft 3 13 IB S -




o


H 3


ft ^l 3-






C (B M a rt (u
K P- w


W IB 1 P-


p-


P* P-


rr (B rt C o


p* in


Ifl O P- 3


o


Pi


(B IB 3 M


H-. pi


ft H-


t! C 3 vQ


3


cr 3


G. M &> -H 3


(B 3


p- p-


S H- 3


(B p) (t


U)





P- I - \ 1


3 P-


O (B


?§^


M 3 O (B
3 0, TD O 3




M


3 in (y)


rt n


3 in






iQ <


in




1)


(B i« IB 3 vQ
Pi ^ H ft P-






1 T) (B






M






M M






H-


ft r;. M M 3

(B r. O f!>

3 •- g






-




/


O






1




ro












1


?.




M


II II
II II

iQ 3 Ml IB M Ln


1
1


1
1

fl


H




P-


o in £ O


Ml


O 3


3-


H-




pj


a C in (B O


(B


O C


I -


ft




ft


3 P- 13 ft O


3


PI


•<


B>




P-


pi P- O rt C


in


3 M




5







pj O 3 M


(B


P-


O




3


cr 3 Pi M 3




P* in








in


P' \ :


-■B


fj p,


•a


8






M in in
<


p*

a


P- P'

(B'


(D


'O






-0 (B


o


3 in


ft


tr






W M


rt




p-


01






o »


2




ft


m










H"-


H-








3







VI








P-
O
in




1 1

Hi O P-


1 1
1 1

Ml (1) o in Ml


l-<


1 1
1 1

^O 3


1 1 1
1 1 1

O IB Ml fB P" lO


1
1

in


1

1

-en P


(11 3


P- •« (-• M ft O


(B


O


1-1 3 (B X O


n


P- 3


W ."3 O


M 3" p- (B p- e

3 p- ft W M 3


H'





P- »Q in O C ~


g


3


rt -0 n


(U


a c


(B P- in (B o


p-


cr r:


ro (B


IB p- n p- Di


ft


3


3 3 P- '0 ft O


i


p- pi


>Q ft Cj


l/l c 1 p-


P-


!-■ P-


M ro rr C o


o


!•-' M


K H- Ul


10 O P- 3





p


ij (B 3 M


o


pi


ft !'•


^3 C 3 lO


3


cr r;


0. ►< Bi M 3


3


p- in


« H- 3


(B 0) ft


U)


o


P- P- ■-^v 1


D.


a


d la


>3 3 O (B




i-(


3 in in


C


u p"


b- 3


g a -0 3






■n <


O


IB


•a


(B „, O 3 lO
pi ^ M rt P-
ft p. W M 3
(B L, O H






1 m (B


ft


in


H






M t^


O





h-






o.»


'1




n






1


in




(D













G



O

c

>








to




in




n




o






1-3




c




G






W




t-i




in






c




►9




^3








o




G




O








g




S




9i








S








w




— computer aut
mation of fc
manual jobs


1
1

o Ml j: ro


1
ifl in tu




1 1 1 1 1
1 1 1 1 1

o g y 10 M, n

P P' O


o


1 1
1 1

O Ki




r: c p- pi


t1 P- C




!•'


(B




M y rl- o


B< rt




3 3 o a y


P


h( tS




rr o y y


13 y




3 p n £ 3 in


1-1


'0 in




c rt


y 0) y




p- tQ pi pro




o




h( p. 0, Oi


P- o




rt (B (T> n P" rl


g


p. 3




(D P- P-


n cr 3




rt 3 B' <


fu


P P




y ti, <


P '< o




(B (B t( ft)


y


rt P'




Ml P-


P- c




(By o ri


p


ro




P> ID in


iq in




rt y P-


I.U


P




y h< p-


Pi (B




e P' <


ro


th O






Oi (t o


HOD.




tr M y ro


3


P- O




y y


(B 1 P-




'< ft iQ


ro


3




rt


P) <




e <




P-




t^


rt


y 3






1




ro




1 It

in




— division I/S groups
report to division
heads and corpor-
ate I/S

— VERY LOCAL respon-
sibilities for I/S
development and
maintenance


rt

(B


1 1 1
1 1 1

< 0. n




II 1 1 1
II 1 1 1

'OinyiBytiD.innS'a
!nroogoroop-opP




1
1

P- o
n p-




t<


H y 3

rt pi 'o




Ohsp-ia innoyyn
Dpj -•


P (U


^-•• c' U' n)






P c


t: o(B tr OP ixiyro




rt ro


o w


U c W p- pj M p>






(n O


t\ DIB ■< P-M kCiroa




C fJ


R>


rr P> 3 o y






p- 1


ro 3 1 in rt fi M




M rt


(n


(D 11) ?!• -0 (X
M 1






in


1 p- p (B ro tr
O 1 in p




ro M
t1





1/1








3 1 rt
1




in y
1



10



FIGURE 4 — The Information System - Productivity Connection



ATLAS BANK



COMPTECH



GRUBER



RATRONICS



SEMCON



Productivity



-incrementally
increase out-
put/per
employee hour



-incrementally
increase out-
put/per
employee hour
-quality-rela-
ted productiv-
ity goals
-shipping on
time within
quality stan-
dards, at
mimimum cost



-incrementally
increase out-
put/per
employee hour
-quality-rela-
ted productiv-
ity goals
-shipping on
time within
quality stan-
dards, at
minimum cost



-incrementally
increase out-
put/per
employee hour
-quality-rela-
ted productiv-
ity goals
-shipping on
time within
quality stan-
dards, at
minimum cost



-incrementally
increase out-
put/per
employee hour
-quality-rela-
ted productiv-
ity goals
-shipping on
time within
quality stan-
dards, at
minimum cost



Program



-low-level
process auto-
mation systems
with corporate
support groups
helping to im-
prove produc-
tivity

-recent suc-
cess with QCs,
to improve
cost & quality



-first five
year plan now
in effect, de-
centralized
with each
group develop-


1 3

Online LibraryThomas A BarocciThe information systems decision → online text (page 1 of 3)