United States. Congress. Senate. Committee on the.

The Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) online

. (page 30 of 140)
Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 30 of 140)
Font size
QR-code for this ebook

the Group, a need for further reliability and serviceability improvement.

Other concerns include the expected difficult GSA negotiation, the need for
accounts receivable recovery in the Federal arena, some new technology manage-
ment problems, the challenge to retain key personnel, and the maintenance of
a competitive position in the marketplace against possible new announcements by
other members of the industry.


Here we have clear signs of trouble which seem unavoidable to a great degree.
As 1968 performance trends upward, our comparison of 1969 over the preceding
year become worse. The potential year-to-year decline can grow from the
presently anticipated nominal amount to a number well in excess of $100 million
if we go toward the high end of the range of equipment sale in 1968. The afore-
mentioned pressure for higher development expenditures this year will be a
serious problem in terms of going rate in 1969. The backlog difficulties of 1967 and
1968 could be further accentuated by the lack of adequate demand in 1969 during
the late stages of System/360. At this reading, I would anticipate the need for a
price adjustment in the late third or early fourth quarter 1968.

1970 and Beyond

Except for the distinctly possible rental inventory upset, we have good prospects
for continued growth in the longer term. The year 1970, however, is a problem to
us in both financial and manufacturing capacity terms. The Strategic Plan will
tell us more about our ability to contain.

♦ «•»**•■

Equipment Sale — The most serious business problem we face is the sharp
acceleration of equipment sale. I believe this high rate of sale will continue into
and possibly through 1968, and the company should, in my opinion, take several
positive actions to correct this severe imbalance of purchase and lease. The
maintenance price increase seems to me to be absolutely essential and thoroughly
justified. I do not think it is wise that the company would take a nominal
increase as against a completely corrective increase as we have recommended.
However, if we are forced to the position of a nominal increase, we should imme-
diately set plans for two additional annual increases to correct fully our highly
vulnerable cost-to-price condition. Realistic multipliers on all new product
announcements are a must, and we possibly may require a price adjustment on
some of the high end products such as Model 65 and 75 CPU's, and their
associated memories. If we cannot have a purchase price increase, then I believe
a rental price reduction may be in order. This is being studied by my people
at this time. Further, I believe we need to modify sharply or perhaps eliminate
the purchase option. None of these actions is easy or without risk. The conse-
quences of a continuation of our present trends are more serious, in my opinion.

Backlog — This office is somewhat concerned over recent sales records in view
of the near t^rm under-scheduled positions which appear chronic, the high
cancellation rates, and our inability to achieve unit sales and new accounts in
adequate number. There is every indication that the Model 20 is suffering a
decline. I will predict that the Model 25 will bring the Model 20 to a stalled
position, and I believe that further investment in the Model 20 other than the
D program is highly questionable. The key problem here is our ability to create
enough demand across the line to sustain the resources we now have. This
condition must be monitored on a regular basis and marketing action programs
must be implemented to correct the weaknesses.

Information Systems — This is a case of darned if we do, and darned if we don't.
Highly advanced new systems may be the only real lever we have over control of
this complex data processing business, which has defied subdivision up to this


time. It may be that highly sophisticated and well developed information systems
are a good alternative to further divisionalization and splitting of our business.
Aside from that factor, systems such as the Manufacturing one and the AAS
may be crucial to future oi>erations — getting the job done. These systems are late
and the cost is high. They are hard to justify in dollars and cents, but our recent
experience in Field Administraiton points out the importance of advance thinking
and anticipation of operational problems. In my opinion we should continue at a
reasonably fast pace to develop the new systems and we should take all
deliberate steps to eliminate marginal systems projects, and certainly all
duplicative systems projects. Along this line, I am holding a review of all
planning systems to determine the extent of possible duplication and to recom-
mend a coordinated, modified attack which will give us good planning and
reasonable information systems economy.

Wage Cos<— Convergence of several major corrective programs within a short
space of time represents a cause of deep concern on our part that we are driving
our expenses and costs up too far, too fast and without the prospect of pricing
offsets. We have seen a rapid succession of what I deem to be relatively short
fused and uncoordinated responses including the UAW adjustment, the accelera-
tion adjustment, the Exempt Overtime recommendation, and the possible mil-
age allowance modification. Each one of these makes sense by itself. When
you put them all together, I believe it is too much, too fast. Some hard-headed
business positions should be taken to modify one or more of these programs and
bring the aggregate more in line with reality and preserve our cost and expense
competitiveness. I recognize the problem of keeping up with the economy and
offsetting inflation for our employees. We do not want to do this on such a scale
that that we will look back with regret over such rapid fire acceptance of every
proposition that the Personnel Department can dream up for the Group. Also on
the personnel scene, we recognize the diflicult situation we face with regard to
key losses. Broad programs of accelerated wages will not solve this problem,
in my opinion. Neither will IBM's current stock option program. The answers to
this problem are not simple, but I would be thinking along the lines of contracts
for key personnel and some kind of arrangements which will provide the pos-
sibility of grabbing the "brass ring" if product and market developments are

Product Strategy — The present trend toward customer and third party owner-
ship in the marketplace calls for astute modification of our strategy. Assuming
stability of architecuture, we probably need to consider such as the following :

1. A bigger jump in price performance in the next round.

2. Faster and bigger memories.

3. Great function.

Mea.sures of this kind are essential to keep the market moving and growing
despite a high level of purchase and cut-rate leasing of current 360 offerings.
They are also essential to the maintenance of price level in future products. Every
aspect of product introduction needs to be reconsidered, including timing. At the
present reading I am not sure we have enough leverage over development manage-
ment to make these things happen. Decentralized systems management has clear
disadvantages here, and while I would not abandon the concept, I think we must
be careful to give enough central direction to make sure that important factors
are not overlooked in the changing environment. We still face the traditional
CPU mentality throughout much of our development area. Is ACS right?? How
important is the advance of logic? We have seen recently that the advance of
memory technology may be much more important. Business management will
have to put out extra effort to keep up with rapidly advancing technology in
the immediate future.

Development Programs — In our opinion, the business cannot afford many
"TOAKE'' programs. Since the problem with that program is variable cost,
additional volume leaves us in the bad situation. Along these lines, we strongly
recommend that we stop the transaction recorder now. and we believe that it
should have been stopped a year ago. The hydraulic printer is a similar example.
Further, SBS will not be a successful program as presently constituted, and needs
to be sharply modified or terminated. Action to stop the weak programs is neces-
sary if we expect to devote our resources on any kind of an optimized basis
toward a maximized ultimate result. There are many good programs around in
which to invest our money, and these receive our full support. We would further
caution that big programs with .serious question marks such as the 27XT .should
be closely monitored and should not be committed on an absolute basis until the


problems are resolved. Finally, I offer again the position that we are spending
too much money on ACS in 1968 and 1969.

Systems Management — There is some turbulence oyer responsibility and profit
management in this area. We are working with the key players in Division and
Group Staff to get clarification of responsibilities and appropriate reporting. We
will be meeting with both of you soon on this.

/S/ Ted

T. C. Papes

plaintiff's exhibit 112

August 6, 1968

Office of the Controller IBM Confidential




System/360 Price Increases

Your memorandum of July 31, 1968

Mr. T. A. Spain

Tommy, in paragraph 3 of your letter you suggest that it is in IBM's best
business interest to increase the level of outright purchase of System/360's.
This point of view is in such fundamental disagreement with what I believe to be
in the Company's best interests that I suggest we seek to reconcile our different
points of view. I would appreciate hearing from you what rationale or support
you have for the position that higher levels of purchase are in our best interests.

It is clear to me in the context of the Strategic Plan of the Data Processing
Group and for that matter of the IBM Corporation that a reasonable balance of
lease and purchase is in our best business interest. The level of purchase we
have experienced for the last twelve months is in my opinion totally out of line
with this appropriate balance of lease and purchase and is in fact excessive
when viewed in context with the long-term growth objectives of the Corporation.
If you have any data or rationale that changes this fundamental picture we
should make it known to the Group and to the Corporation.

AJK:ls A. J. Krowe

cc: Mr. J. E. Cuth, Jr.
Mr. D. T. Kearns

plaintiff's exhibit 267.

IBM Registered Confidential

May 1, 1969

To: Mr. W. C. Hume

Mr. G. E. Jones

Mr. B. Marshall

Mr. M. B. Smith
Subject : DPG 1969-1975 Goals— Financial Critique

Corporate Finance does not agree with the profit margins in the DPG Goals,
in that margins are restored too slowly and profit turnaround from 1968 occurs
too late.


The large purchase bubble between mid-1967 and mid-1969 has advanced
significant revenues which would have been received during the Goals period
under a more traditional pattern of purchase activity. Although the compound
growth rate in revenue and profit between 1968 and 1975 or 1969 and 19(5
shows a relatively balanced growth pattern (Revenue growth of 12.7% profit
growth of 13.0% from 1969 for example), refiecting a return at the end of the
period to a more traditipnal financial performance level, the annual pattern of
revenue and profit shows the heavy impact of the purchase activity (see Exhibit
1). Relatively minor changes to normal expense growth patterns when combined
with significant shifts in revenue produce this irregular profit pattern.

The long-term profit degrading effect of purchase can be seen by comparing
the profit margin of the aggregate revenues and profits during the period with
similar figures for the preceding period :


1968-75 :

Revenue $39, 850

Profit $10, 680

Percent 26. 8

1960-67 :

Revenue $14, 154

Profit $4, 181

Percent 29. 5

In addition to the effect of differences in product mix, market and competi-
tive environment, etc., one of the major reasons for this decline in profit levels
can be attributed to the continuing service requirement of the purchase sinventory.

Another important facet of this increase in purchase, and another way to
view its long-term profit effect, is the distortion of conventional expense to
revenue measurements caused by the service requirements of the rela-
tively large purchased inventory, (see below on Expenses)

Finally, this large purchase inventory has the further effects of slowing ac-
ceptance of next generation systems and increasing the average yields during
the acceptance period. Through a combination of price discounting and special
service or programming offerings by third party organizations NS will be pre-
vented from or delayed in replacing a larger portion of the installed inventory.
When replacement of a purchased system does occur, especially in the early
years of the program, it will most probably replace an installed IBM rental
system of smaller point value, thereby producing higher yields (see section
below or business volumes) .

>l< >i> * * 9|c :(: 4c


Exhibit 2. — Report on Market Appraisal, Bendix Computer Division, Bendix


Appraisal of the Market for the Bendix G-20, G-20, and G-25 Computers

bendix computer division bendix corp., los angeles, calif.

(This report is confidential and intended solely for the information of the client

to whom it is addressed.)

January 1961

BOOZ -ALLEN & HAMILTON -w.o.,.. c■.cvcL.^Nu ..

A\ci/ia(/cinc nt (onsullanis iosai.o.i l



January 6, 1961

Mr. M. W. Horrell, General Manager
Bendix Computer Division
Bendix Corporation
5630 Arbor Vitae Street
Los Angeles, California

Dear Mr. Horrell:

We have completed the work as outlined in our proposal letter of
October 18, 1960. To the extent we found it possible to do so, we have
answered the five questions set forth therein, and these answers are
presented later in this summary report. However, it became apparent
early in the study work that the complexity and rapidly changing nature
of available computers and the market for them was such as to make it
impossible to answer with any real confidence the question of what pene-
tration Bendix can reasonably expect to obtain with the G-20.

Thus, in order that the findings and conclusions we do present may
be viewed in their proper perspective, we feel it necessary to precede
them with a general discussion of the events that have led to the present
situation. The combined effect of these various prior events presents


Bendix with a real need to make some major decisions as to the best
future course of action for the computer division. In order to assist in
this decision process, certain basic courses of action have been set down
and then examined as to their feasibility of accomplishment. These are
presented later within this report.



At the time the G-15 was introduced to the market, a significant
gap existed in the IBM product line between the IBM 604 punched card
calculator and the 650 stored program computer. Several companies,
including Bendix, Royal McBee, and Burroughs, recognized the oppor-
tunity presented by this gap and introduced small scale, stored program
computers. These products met a real need in the engineering-scientific
market and attained immediate sales success. An attempt by IBM to
counter this threat with the model 610 was not effective.

The problem facing IBM with respect to the introduction of a
machine fully competitive with the G-15 class and thus competitive with
their own 650 was discussed in our 1957-1958 report. The opportunity
presented by this situation was clearly apparent at the time and the
recommendations and forecasts presented therein were predicated
upon its exploitation. Sales of small scale E&S computers proceeded


generally in accord with the estimates and programs developed at that
time up to this past summer when IBM introduced the 1620. This new
equipment fills the gap in the IBM line and, in addition, offers many-
significant advantages over the G-15, placements of which dropped off
sharply. This move by IBM, combined with the recent introduction of
other small scale computers utilizing the latest solid state circuitry
and core memories, such as the Packard Bell 250, Control Data 160,
Royal Precision 4000 and the Autonetics Recomp II, will force the G-15
into the last part of its life cycle.

Bendix does not now have a fully competitive computer to offer in
this part of the market; the G-20 does not replace or supersede the G-15
with an improved model, but enters an entirely different segment of the
raarket .


The G-20, rather than being a small to medium computer that
many of your 300 satisfied G-15 users could advance to, is actually
in the area of sophisticated computers classed as "medium" in our
earlier study. The more elaborate G-20 installations will, in fact,
be well within the large scale portion of the market in terms of price
and performance.


Exhibit I, following this page, shows the relative position of the
Bendix G-15 and G-20 with respect to the computers currently in use or
available for procurement. The units are listed from top to bottom in
order of descending monthly rental for a "typical" system. It can be
seen that the G-15 user has the choice of a number of different com-
puters with superior capabilities to the G-15 before reaching the G-20
class system. The significant increase in competition facing Bendix is
also shown on this exhibit, in that the division's competitors at the time
of our earlier study are shown as squares [ * | , while virtually every
model shown is now in competition with either the G-15 or G-20.

In its class, the G-20 is a top quality computer offering consider-
able computational capability per dollar of rental.

The system is generally equivalent to competitive equipments
offered by IBM, Remington Rand, Minneapolis-Honeywell, and other
manufacturers. As pointed out above, with very few exceptions, it does
not provide the G-15 user, who has fully loaded his machine, with an
opportunity to trade up to a more capable computer in the Bendix line
without increasing his expenditure more than fivefold. The G-15's that
have or will soon become saturated will be replaced with t'he IBM 1620,
Recomp II, or similar equipment; the G-20 will have to be leased or
sold to new customers entering the market or to replace obsolete

40-927 O - pt. 7-20




' S 2 g o ■

s £ s 3 £ s

5 -J .?. S T ,■*, S ?, s s S S S i



medium scale installations such as the IBM 650 or Burroughs
Datatron 205, or large systems such as the IBM 704 or 709.



Since 1957, a whole new family of IBM computers has entered the
market place. In addition. Remington Rand, Minneapolis -Honeywell,
RCA, GE, Control Data Corporation, and others have introduced new
systems covering most fields of application.

(1) An Almost Continuous Range of Choice Now Exists;

Market Categorization by Large, Medium, and Small
Has Lost Much of Its Former Meaning

Along with the new equipment developments that have
taken place, such as the transition from tubes to transistors
and drums to cores, several other significant changes have
taken place in computer design concepts. Through the use of
accessory devices and a wide variety of peripheral equipment,
the range of applications of a given base unit may be extended
over a relatively broad portion of the market. This is true of
the G-20 computer, as well as its competitors. As a result of
these developments, the relatively clear lines of demarcation
that did exist between small, medium, and large computers


have pretty much disappeared. There is, of course, graduation
from the minim.um useful equipment obtainable up to the very
large and powerful systems, but no longer can any one base
unit be said to occupy a specific or somewhat limited portion
of the range.

A general idea of the confusion in the market place brought
about by these developments can be seen by superimposing the
system price range for the minimum useful installation up to the
super de luxe model with all of its associated "bells and whistles"
on the exhibit shown earlier. This is presented as Exhibit II,
following this page. This "modular" blending of equipment offers
the prospective user a wide selection among equipments and

(2) There Is No Longer a Major Distinction between
"Engineering -scientific" Equipment and Business
Data Processing Equipment

Of special significance to Bendix is the merging of
"engineering-scientific" and "data processing" equipments
into one as a result of advances in technology, computer design
practices, programming techniques, and the requirements of
the customer. Various systems have been slanted toward the
engineering or data processing markets through the selection







i 8 «. a s s s 5 I

IS- 00°-"

t 5 lis §1 H i iiii i ^1 i lis



of design alternatives during the development process or through
the orientation of the marketing effort. Both methods have been
used in an effort to "optimize" the Bendix G-20 for the engineer-
ing market. It is important to note, though, that all manufac-
turers are increasing the number and type of accessories avail-
able to their equipment in order to satisfy as large a segment of
the market as possible.

This change did not come about by pure happenstance; it
meets a real and growing demand in the market place. In tech-
nical research centers, for example, many of the calculations
now required to solve technical problems are of a data proc-
essing nature, requiring the handling of large quantities of
input and output with relatively simple computations. Thus,
the scientist needs certain data processing capabilities in his
computer. Similarly, as more information becomes available
to corporate management through the use of automatic data
processing equipment, growing uses are being made of various
operations research techniques drawn from the fields of science
and mathematics. The solution of these problems requires many
of the same computational capabilities previously considered
necessary only for scientific computers.


Another factor calling for equipment equally adept at
scientific or data processing work is an essentially economic
one. Generally speaking, the cost per unit solution declines
as the size and capacity of the newer computer systems in-
creases, making it more imperative than ever that a single
installation be used by the accounting and clerical areas of
the business and also by the engineers and scientists.

Finally, if two equipments are needed in companies with
really great capacity requirements or in those instances where
the scientific and clerical functions cannot share equipment for
some reason, it is advantageous to have their computers iden-
tical or at least fully compatible, so that, in case of breakdown,
vital work can be quickly transferred to the other machine.

Thus, the opportunity for uniqueness and specialization in
the "engineering-scientific" area, seemingly a major point in
Bendix policy, seems to no longer exist, at least so far as equip-
ment design is concerned. Of course, it can still exist in the
orientation of the sales and support effort, if such a course is




It is evident that the unique position in a discrete section of the
market held by Bendix only a few years ago has completely disappeared.
The conditions under which, in spite of its relatively small size and limited
sales effort, the Bendix Computer Division competed successfully in the
computer business no longer exist. This is partially because the giants

Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 30 of 140)