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 32 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 32 of 140)
Font size
QR-code for this ebook


the newer equipment will be to subject the G-15 to increasing
pressure from technical and economic obsolescence as time
goes on.

It is likely that the Bendix share of the unit market will
fall from its past 35% share to somewhere in the vicinity of
10%, for the reasons stated above. This, then, offers the
division a potential of around 120 units. Some of these would
be to new users who want a computer installed quickly, others
to prospects who wish to benefit from Bendix's extensive pro-
gram library and to current G-15 users who wish to add more
capacity that is compatible with their existing equipment.

While these 1, 200 units are entering the market to fulfill
new requirements, approximately 800 additional units will be
procured to replace existing installations. Since the G-15 will



5160

be under pressure from obsolescence itself, we do not believe
that it will participate in the replacement market - except as
a candidate for replacement.

In the opinion of your sales people, approximately 50%
or some 160 of the currently installed G-15's will approach
usage saturation in the next 12 to 18 months. In this event,
it is probable that a large portion of these units will be turned
back and more powerful equipment procured. As these G-15's
are returned, they will become available for re -lease to new
customers, or will be disposed of in some other way such as
outright sale at an attractive price.

Since at least a portion of the 120 unit new sales that we
see can be met with returned units, the number to be manufac-
tured is less than 120, probably between 50 and 75. This is
far below the 350 or so that our earlier study indicated could
be installed in 1961 through 1965. It is also far below the 500
small G-15 type computers called for in the Bendix plan dis-
cussed above in regard to the G-20. The reduction from our
original 350 estimate is in part due to a reduction in the number
of small computers that we believe the market will absorb, some
2, 400 to 2, 800 rather than around 3, 300 as seen in late 1957.



5161

The primary cause of the reduction in Bendix's probable share
of the market from around 350 to approximately 75, however,
is due to the introduction of considerably more aggressive com-
petition and better equipment to this market.

As in the case of the G-20, the quantitative data used above
may be off by a wide margin. Any probable variation, however,
can hardly change the main conclusion. The G-15 is going to
decline in number of net placements at a rapid rate. There is
no chance whatsoever of its producing the volume needed for
profitable operation in this business.

The G-20, while medium in size as computers go, is
nevertheless a fairly expensive piece of equipment. The G-15,
as we have shown, is definitely on the down side of its life cycle.
The requirements for profitable volume from the engineering-
scientific market simply cannot be met with these two equipments.



(3) The G-25 Would Provide the Best Market Opportunity for
Bendix at This Time, but It Is Unlikely To Completely
Offset the Volume Deficiencies of the G-15 and G-20



There is a definite market for a computer of smaller size
than the G-20 and more up-to-date than the G-15. This is being
met by IBM with the 1620 and 1410, RCA with the 301, by the



5162



NCR 315, Packard-Bell 250. Royal Precision 4000 and 9000,
North American Autonetics, Recomp II and III, Control Data 160,
and others.

Bendix does not now have a competitive computer to offer
to this market. We have been told by division management that
a system in this category could be announced in January of 1961
and initial deliveries made in early 1962. Thus, this product
must also be examined in order to determine its contribution
to the Bendix program.

The G-25, as now planned, would fall on or very near the
hypothetical price line between small and medium class computer
systems. Since we have no historical data with which to work in
this price class, it appears best to appraise the market for the
G-25 in terms of the small and medium markets. If we assume
that such a computer can compete for the top half of the "small"
equipment market and the bottom half of the medium category,
the potential would be approximately 825 new units and 550
replacements over the next five years.

If Bendix were able to capture 25% of the new equipment
market, replace 200 of the installed G-15's, and obtain 25% of
the remaining 350 unit replacement market, the G-25 potential



5163

would be around 500 units. We would consider this to be the
upper limit, based upon a threefold increase in sales and sales
support effort which, as in the case of the G-20, would raise
the marketing costs to a point where the program might not
continue to be profitable.

It is more probable that Bendix's share of this market
will fall considerably short of this 500 -unit upper limit. We
say this for four principal reasons:



The amount and caliber of competition has
increased more than fourfold since Bendix
achieved a 35% share of the small computer
market.

Bendix is almost a year behind competition in
the introduction and marketing of a computer
in this class.

The marketing force required to achieve this
volume will take a considerable period of time
to develop; in the meantime, sales will be lost
to competition.

During the period that the necessary marketing
force is being developed, the existing small
sales force will be further split, reducing its
over-all effectiveness in the sale of G-25's
and G-20's.



The net conclusion here is that Bendix's market opportunities
within the engineering and scientific field are limited by two key factors.



5164



The amount and caliber of competition for each segment of the
market is such as to make it extremely unlikely that Bendix can obtain
the volume of business necessary for profitable operation.

A marketing force of sufficient size and quality to obtain a profit-
able share of these markets has not been developed in advance of the
current need, and it is unlikely, if not impossible, that one can be
formed in time to seriously affect the outcome. Furthermore, even
if it could be, the cost of doing so would reduce the already slim
chances for a profit in this market.

It will be of interest to you to know that we hold the same opinion
with respect to other companies in the field; we do not believe that any
company can be successful limiting its attention to engineering and
scientific applications.

Thus, we can conclude that Bendix should expand its operations
into serving the data processing field, assume the role of a maverick
in the computer business, change the nature of the division, or withdraw
from the field.

The next point examines the- chances of success for Bendix under
an expanded operation.



5165



EXPANSION BY BENDIX INTO SERVING THE COMPUTER MARKET
IN ALL FACETS IS NOT WARRANTED BY THE PROBABILITY OF
SUCCESS



We consider it a fair amd reasonable presumption that, in addition
to IBM, there are other companies who will survive in the computer
business and end up with a profitable operation, at least when viewed
on a current basis.

The determining factors as to how many and which these companies
might be are:

Size of the total market and how much of it IBM and its
nearest competitor are likely to get.

The minimum size that any one company must achieve
to permit a profitable operation.

Which companies are able and willing to put forth and
finance the effort required at the necessary rate.

An examination of these factors from the Bendix point of view can
yield some appreciation of the magnitude of effort that would be needed
and the sizable risk involved.



(1) The Total Market for Computers of All Types Will Run
around $4 Billion over the Next Five Years, with IBM
and One Other Company Obtaining 85% of It



On a basis of selling price, rather than rental income,
this figure of approximately $4 billion, covering all computers



40-927 O - p'



5166

regardless of size or use, appears reasonable as a starting
point from which to examine the opportunity the market holds
for Bendix.

If we then allocate 85% to IBM and one other company,
this leaves $600 million, or $120 million per year, for all
others in the field. For comparative purposes, this is the
rough equivalent in dollars of about 120 to 160 G-20's per
year. Actually, it will be made up of all types of computers,
with around 70% or 350 in total being roughly comparable to
the G-20 in size. This $120 million in annual sales will have
to meet the needs of all but-the two largest companies in the
field. Even so, this looks like a substantial market and it is,
but as previously mentioned, high sales volume seems to be
a prerequisite to success in this business.

(2) A $50 Million to $75 Million Annual Sales Volume Appears
To Be Required for Continued Success

While the minimum size for profitable operation is con-
jectural, five key points stand out which substantiate the con-
tention that large size is a necessity.



5167



1. Research and Development Expenses Are High

It is our understanding, for example, that the
Bendix development cost for the G-20 will be in the
vicinity of $2. 3 million. This is only for a single
product to compete in this equipment generation.
Past history indicates that we may expect the intro-
duction of a new generation of computers every four
or five years, at least for as long as the market re-
mains highly competitive. This cycle may lengthen
as the art matures, and the degree of difference
exemplified by the move from tubes to transistors
may never occur again; nevertheless, there will be
change and this change will require heavy R&D ex-
penditures to remain competitive. We estimate that
it will require from $3 million to $5 million per year
simply to maintain an adequate, though not necessarily
full, line of equipment up to date. For comparison,
IBM reportedly spent close to $75 million for computer
R&D last year, while Minneapolis -Honeywell, a relative
newcomer, spent an estimated $7 million.

2. Marketing in the Computer Field Is a Very
Expensive Undertaking

This is especially true in the data processing
application area, for usually a complex system is in-
volved and a great deal of direct selling effort and
systems or application engineering work has to be
done to convert a prospect into a customer. This
sales effort must be directed to a number of "pros-
pects" for each customer that actually signs a lease
or purchase agreement. (This ratio has been con-
servatively estimated at 15 to 1, with even the most
efficient sales organization.) A relatively complex
systems analysis and detailed proposal must be de-
veloped for each of these prospects; all to obtain one
"sale. "

Furthermore, if it were possible to identify the
one best prospect, it is still a long-range necessity to
submit a full proposal to the majority of the other poten-
tial customers if the manufacturer wishes to continue to
be considered for possible future procurement. This
total effort requires a large, aggressive, well -qualified,
and well -trained direct sales organization.



5168



3. The Direct Sales Force Must Be Backed by a Large
and Highly Capable Applications Engineering Force

This part of the organization is charged with the re-
sponsibility for developing basic programs, proposals,
and system "software" packages for potential and actual
customers alike. Since the smaller company must exert
a total marketing effort as large and effective as all but
its largest competitor, yet write these costs off against
a relatively low volume of business, marketing costs will
be higher percentagewise for the second, third, fourth, . . .
ranking companies. The theory that the smaller company
can obtain more sales per dollar of marketing effort by
careful prospect selection is fine - except that RCA, GE,
Honeywell, and others are employing it also. Thus, in
any given situation, Bendix, or any other competitor for
that matter, must expect to face extremely strong com-
petition, the pace being set by IBM or its nearest com-
petitor.

4. Systems Service Costs Are High

Prompt service is a must in the computer business.
Bendix has established an excellent reputation in this
respect, but the cost has been extremely high. Here
again, size is a requirement for economy. Enough
equipment must be out on lease to make economical use
of a minimum size service organization. This is not
too difficult in the case of medium and large systems,
since a service man usually stays with the equipment.
It is a problem with small computers, for geographical
concentration is required.

5. Machine Population Affects the User's Choice of
Equipment

Size also plays a significant part in the customer's
ultimate equipment selection. Few computer users are
willing to run the risks associated with having the only
computer of its kind in an area. Equipment failure can
be a serious problem to the user and his only insurance
is the immediate availability of another machine that can
handle his work load. RCA, Honeywell, and GE have be-
gun the creation of service center networks in an effoii;
to offset IBM's advantage in respect to machine population.
This is also a very costly undertaking.



5169

These points are presented in a semiquantitative manner
because of the lack of detailed cost information; however, they
do indicate the nature of the problem and provide some measure
of the size required to compete effectively in this business.

It seems reasonable to expect that a sales volume of $50
million to $75 million would be needed to provide the necessary
sales and service organization, research and development, and
missionary work in this field.

If this is the case, the $120 million annual sales volume re-
maining for other than the first two in the field provides room for
only one or two additional companies.



(3) More Strong and Determined Companies Are After a Place
in the Computer Business than There Is Room for



The reasoning of the last two points indicates that in ajddi-
tion to IBM there is room for only two or three other companies
on a profitable basis.

Currently, the competitors for these two or three positions



Sperry Rand - Univac Division' . Royal Precision Corporation

Radio Corporation of America*' . North American Aviation, Inc. - Autonetics Division

General Electric Co. "^ . Control Data Corporation '^

National Cash Register Co.' . Digital Equipment Corporation

Minneapolis-Honeywell Regulator Co. ^ . El-tronics Inc.

Burroughs Corp. "' . Monroe Calculating Machine Company

Philco Corp. -^ . Packard-Bell Electronics Corp.

Bendix Corp. . Thompson Ramo Wooldridge Inc.



5170



and possibly others, almost all of which are knowledgeable,
well-financed firms with an avowed purpose of making a per-
manent place for themselves in the computer business. In our
opinion, all but three or four will fail to do so, but the economic
pressures and profitless sales volume that will result from their
attempts will be very costly to all concerned.

In two to five years, after the present competitive situation
has shaken down, at least two and maybe more, in addition to
IBM, will find a position in the field. The chances are that they
will come from the top of the above list.

It is by no means impossible for Bendix to join the race
and be one of the winners, but we do think it improbable and
feel even more strongly that the chance of success is not worth
the risk involved.

If it is agreed that the first two possible alternatives,
concentration on the engineering and scientific market and
expansion into the total market in all its facets, are not
feasible, the third "marketing-oriented" approach should be
exannined.



5171



9. IT DOES NOT APPEAR FEASIBLE FOR A MAJOR CORPORATION
TO PARTICIPATE IN THE COMPUTER BUSINESS ON A LIMITED
SCALE

It is well known that there are many instances in industry where a
small company may compete effectively and on a profitable basis with the
giants that dominate the market. An examination of the general character-
istics of such a maverick and specifically the relationships between these
characteristics and the computer business is pertinent to Bendix's problem.



(1) It Appears Unlikely That a Small Manufacturer Can Hold
a Lasting Position in This Industry



Ordinarily, the manufacturer who assumes the role of
maverick in an industry achieves success through one or more
of the following routes:



Lower total costs

Higher quality or technologically superior products

Geographical, industrial, or product specialization



These market opportunities and the likelihood of their
occurrence in the computer business are considered in the follow-
ing points.



1. It Is Improbable That the Small Supplier Can
Achieve Lower Total Costs

The high cost of the sales, support, and service
effort required in this business was discussed in detail
above. Essentially, it costs the computer manufacturer
the same to develop a computer program such as the



5172



Bendix SPACE, PAR, or SNAP systems for one unit or
one hundred. The unit cost is obviously lower with higher
volume and is generally reflected in the price to the cus-
tomer. The same is essentially true with respect to
product development and sales and service costs.

Thus, the one prime price reduction opportunity
available to the maverick computer manufacturer is in
production costs. In this industry, the cost of goods is
around 25% of the selling price; therefore, a 20% reduc-
tion in manufacturing costs would represent a 5% reduc-
tion in the price to be used at best. This is not of enough
significance to give the small supplier a lasting competi-
tive advantage, since the dominant suppliers can easily
meet it through volume economies in all other areas.

2. It Is Unlikely That the Small Supplier Can Maintain
a Significant Degree of Technological Product
Superiority

The maverick supplier is usually successful in
those industries that have largely achieved technological
stability and where the pace of product evolution has
slowed considerably. The maverick then exploits the
scientific principles and techniques that have been devel-
oped in the universities and large corporate laboratories,
in an effort to offer a better product without the expense
of extensive R&D facilities or a large technical staff.
Examples of this approach may be found in the fractional
horsepower motor, steel fabrication, paper products,
and automotive replacement parts businesses, as well
as in many others.

In the computer industry, the state of the art is
far from maturity - the rapid pace in technology forces
the manufacturer - large or small - to keep abreast,
and if possible, to stay ahead of developments in order
to maintain product superiority. The $3 mallion to $5
million annual R&D expenditure cited earlier represents
the minimum that we consider necessary to maintain an
adequate and up-to-date line of products in this field.

If the small manufacturer does not offer a full line,
the product he does offer must be very superior to and
several years in advance of competition. The basic and



5173



applied research cost of such a program would be stag-
gering to the small company, when one considers that
the combined R&D expenditures of the three or four
leaders in the industry are sonnewhere around $200
million a year.

Even if we were to assume that the maverick
could achieve a major technological break-through,
it would not be of lasting significance, since few devel-
opments in electronics can be adequately protected
against infringement or circumvention.

3. Profitable Operation through Geographical,

Industrial, or Product Specialization Also
Appears Improbable

It has been shown that application specialization
is ultimately an unprofitable approach to the computer
business. Geographical concentration appears equally
unattractive for the same reasons: inadequate volume
to handle the high cost of systems development and pro-
gramming support. Industrial or product specialization
also appear to be unlikely avenues of operation, for the
versatility of today's general purpose computers is such
as to make them quite competitive in virtually every in-
dustry and application. There may well be specific
needs that can be better filled with a special purpose
unit than with general purpose equipment. What these
needs are, and whether they exist in sufficient quantity
to warrant the relatively high development costs, is not
known at this time. If they do exist, and can be identified,
the next point should be considered.



(2) Successful Exploitation of Small, Special Purpose Segments
of a Market Is Usually Accomplished by a Company with
Certain Distinguishing Characteristics



The small company that assumes the role of the fringe
supplier to an industry can usually be categorized as small, hard-
hitting, and able to quickly identify, seize, and exploit an oppor-
tunity. Almost without exception, the successful maverick is set
apart from its larger competitors by the following features:



5174



1. The Successful Maverick Is Usually Owner -ope rated

The company that succeeds in this role is usually a
proprietorship or partnership, owned and operated by one
or more individuals that could be characterized as entre-
preneurs. Each of the participants performs the functions
of management, often doubling in brass, and certainly on
a much lower overhead basis than would be found in the
larger companies.

2. Success Is Often Built upon Personal Relationships

In many instances, the maverick company is started
by former employees of one of the dominant corporations
in the field who believe they have identified one or more
segments of the market that are not being supplied by the
larger competitors. The owner-management team is
quick to exploit personal relationships throughout the
industry in an effort to obtain a foothold and keep it.
These contacts are used as a source of sales and infor-
mation on possible market requirements and opportunities.

3. The Fringe Supplier Must Usually Assume a High
Risk for a Low Profit

The segments of the market that are left to the
maverick by the dominant companies in an industry are
normally low volume and /or low profit. The small com-
pany runs a substantial risk that the larger company will
"move in" at almost any time, if its particular segment
appears profitable. Similarly, the small manufacturer
must often depend upon innovation as a source of sales,
which in turn increases his dependence on R&D and his
risk. This high risk, low profit type of operation is
normally characteristic of the maverick - not of the
larger, well-established corporation.



In essence, it would appear that it is quite improbable that the
computer business offers a lasting or significant opportunity to the small,
fringe supplier that finds a place in many other industries. Furthermore,
if such an opportunity were to develop as the computer business matures.



5175



it is questionable whether Bendix, or, for that matter, any other large
corporation, possesses the peculiar characteristics that contribute to
the success of the "hit and run" or maverick company. Even if both of
these conditions were met, it is doubtful that assumption of the maverick
role in an industry of this size and scope represents the best utilization
of the efforts and resources of a corporation with the strength and stature
of Bendix.

10. BENDIX SHOULD DEVELOP A PROGRAM TO REDIRECT THE
EFFORTS OF THE COMPUTER DIVISION

If the conclusions pertaining to Bendix 's opportunities as:

a supplier to the engineering and scientific computer market,

a broad gauge supplier to the computer market in all of its
facets,

and as a maverick, or fringe supplier to small isolated
segments of the market,

are accepted, Bendix is faced with the need to determine a specific
course of action for redirection of the computer division's activity.

This needs to be planned very carefully in order to uphold Bendix 's
reputation as a company that stands behind its products. Furthermore,
it will have to be carried out in a manner that would minimize possible
operating losses stemming from personnel turnover, the return of leased
computers, etc.



5176



A specific program with a fair chance of success is far from
apparent at this time. However, as a starting point, the following
possibilities should be considered.

(1) Combination with Other Bendix Divisions

It is possible that one or more opportunities exist where



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 32 of 140)