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 31 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 31 of 140)
Font size
QR-code for this ebook

of the industry were not actively competing for the small computer market
until recently and partially because of changes in the characteristics of the
market discussed above.

The G-20 appears to us to be an excellent computer, fully competitive
so far as its technical excellence is concerned; yet it is not unique in this
respect - other equipments are just as good or even better for some applica-
tions. Neither does it, to any significant or lasting degree, enjoy a price
advantage or suffer any price disadvantage. We say to a "lasting degree, "
because it appears probable that the current price-performance disparity
between the new generation equipment and its predecessors will be largely
eliminated through price adjustments, discounting, and lease concessions.

Thus, the factors which govern the probable market penetration of
the Bendix G-2 are totally aside from the equipment itself; they have to
do with the Bendix image in the computer market, the magnitude and
direction of its sales effort, and the service offered both as to the mainte-
nance of equipment and the availability of programming services.



With these significant changes in the nature of the computer busi-
ness and the relative position of Bendix in it in mind, management is
appraising its probable future position in order to arrive at a decision
as to the best course of action for the division. In approaching such an
appraisal, it seems realistic to consider Bendix as having developed a
technically acceptable product (the G-20) and it is now trying for the first
time to break into a business or market characterized by five features of
major importance and pertinence.

The realism of this approach is not significantly altered by the fact
that Bendix has been in the computer business and has established an
excellent reputation with the G-15. The image so created is one of par-
ticipation in a small and special portion of the market, which only slightly
improves its position over a total newcomer attempting to enter the com-
puter business on a broad scale. Likewise, the division's sales force has
had limited prior experience. Those who played the most active part in
the G-15 sales effort are now, for the most part, in administrative posi-
tions rather than active selling. The newer members of the organization
will require a year or more to gain the experience and knowledge neces-
sary to become equally productive.


The five features considered to be of major importance and per-
tinence are:

(1) The Market Is Tremendous.

(2) The Competition Is from Real Industrial Giants.

(3) There Is No Really Significant Product Advantage Held
by Any Manufacturer.

(4) The Product Is Expensive - Thus, Its Purchase Receives
Top Management Attention.

(5) It Is a High -investment Business in Which No Profit Is
Returned until the Equipment Has Been on Lease for
More Than Four Years.

Each of these features is discussed in detail below.

(1) The Market Is Tremendous

There is hardly any question that the opportunities for
worthwhile computer installations are so great that the total
size of the market can accommodate on a profitable basis a
considerable number of competitors if it were equally divided.
Exhibit III, following this page, reflects the annual rental
revenue that has been and probably will be derived from this
market between 1955 and 1965. Ten companies could each
derive over $100 million in revenue apiece in 1965, if the mar-
ket were evenly split. Unfortunately, this is not the case, for


IBM has held, and will probably continue to hold, at least a
70% share, while the second largest contender. Remington
Rand, holds an additional 18%.

Thus, statistics on total market size are not of much
significance in appraising the Bendix problem. What is im-
portant is the probable size of the market after IBM and the
second place contender take their portion. Since this portion
will almost certainly be 85% or more of the market, the revenue
remaining for all others will probably range between $70 million
today and $165 million in 1965. Therefore, it would appear that
the more significant question becomes "what is the minimum
size and revenue needed for profitable operation, in view of the
extensive product development effort and relatively high sales
expense required to participate in this industry? "

(2) The Competition Is from Real Industrial Giants

Of course, as discussed above and shown on Exhibit III,
IBM is the major competitor in this industry, with a historical
share of more than 70% of the total market. In the segments
where IBM has concentrated product and marketing effort, its
share has been closer to 80%. Even in the one identifiable
instance where competition had a significant jump on IBM,


bank automation, the company has captured no less than 50%
to date. Thus, the significance of the fact that Bendix achieved
its success with the G-15 at a time when IBM did not have a
competitive product becomes even greater. It is also important
to note that Bendix's share, though 35% of the small computer
market, represents only a little more than 1% of the total.

Comparatively recently, such companies as RCA,
Minneapolis -Honeywell, GE, and others have entered the
computer field with a major effort. Each has already spent
large sums, upward of $50 million in product and market
development, and appears willing and able to continue a high
rate of expenditure to carve out and maintain a position. Fur-
thermore, all this effort on the part of these relative new-
comers is economically justifiable only on the basis of profits
five to ten years away. In other words, their current expendi-
tures are investments on a long-term basis to establish a posi-
tion from which profits can be earned at some time in the future.
Each of these companies has other profitable and related product
lines to support the development of both equipment and markets
in the computer business.

It would seem that, to be successful in the face of this
type of competition, a major effort is essential, as well as a


willingness to expand aggressively, even though it means fore-
going current profits. The computer business is one in which
large size is essential to success and this means large size in
the computer business.

The only possible alternative would be to find a field of
specialization both as to product and market and establish in
it a preference for Bendix equipment. This seems to be what
Bendix would like to do if it is at all possible.

(3) There Is No Really Significant Product Advantage Held
by Any Manufacturer

The evidence is now strong that there is no essential
difference in an "engineering-scientific" computer and a
"business data processor. " As previously pointed out, real
data processing capacity is a requirement for many engineering-
scientific installations, and the total cost of the more sophisti-
cated equipments is such that the incremental construction cost
of a dual purpose unit is relatively small. In addition, the
customer often requires multiple usage to economically justify
the investment. Thus, by and large, there are now just com-
puters - not E&S computers and data processing computers.

At present, some relatively small differences do exist
in the details of operating features, physical design, et cetera.


which can combine to give a lower cost per unit of work for a
particular application or user than some other possible com-
bination. However, viewed in the sense of the broad problem
now faced, not only by Bendix, but by all computer manufac-
turers, these differences are not now, nor are they likely to
become, such that they provide any one manufacturer with a
lasting product advantage of any significance in the market

The hardware appraisal conducted by the study team in
cooperation with the Bendix marketing department clearly
shows that the Bendix G-20 has no product advantage sufficient
to markedly influence its penetration of the market. Some
features, though ahead of the field at this moment, will un-
doubtedly be matched or even surpassed by competition in
the very near future.

(4) The Product Is Expensive - Thus, Its Purchase Receives
Top Management Attention

The lease or purchase of a modern medium or large scale
computer represents an undertaking of sufficient importance to
receive top management attention in most if not all cases. This
was not always the case with the G-l^, since the smaller expendi-
ture (around $18, 000 a year) represtented by this equipment could


often be authorized by the user himself or middle management.
In the case of the medium and large systems, there is much
more involved than merely the rental cost, which averages
$120, 000 and can easily run to more than one third of a million
dollars a year. The use to which the computer is to be put can
also involve significant additional expenditures for site prepara-
tion, installation, and programming, to say nothing of the major
disruptions and changes in orgainization and procedures that often
accompany a computer installation, especially if it is for data
processing use.

Under these circumstances, decisions as to equipment
selection tend to be made by executives who are not too con-
versant with detailed operating features, but who will place
considerable weight on the experience, reputation, and sub-
stantiality of the computer manufacturer. In this respect, it
can hardly be argued that IBM does not hold a distinct advan-
tage at the moment, because of its well-established image in
the field. This outstanding image stems from several impor-
tant factors. First, some 80% of the installed medium and
large scale systems are IBM products. A number of the more
knowledgeable users were contacted during the course of our
assignment. We found virtually all of them to be satisfied


with the quality of IBM products and service. Although most
were sophisticated enough to recognize that an IBM system
was not necessarily the best in its class, they did indicate
that IBM would have to slip quite drastically in their sales
and service support (which they considered unlikely) or that
a competitive system would have to do much more for less
money than its IBM counterpart (also considered unlikely),
before they would change suppliers.

The second source of IBM's established image stems
from the more than 90% share it holds of the electrical
accounting machine (punched card processing equipment)
market, since the users of this equipment today will, for
the most part, be the new computer users tomorrow.
"Brand loyalty" among this group of potential users, as
well as among the current users of IBM equipment, will,
from all indications, tend to perpetuate its dominant posi-
tion in the market.

This position of distinction can. in all probability, be
overcome in the long-term future, if effort of adequate quality
and magnitude is applied; RCA, GE, Minneapolis -Honeywell,
and others are making a concerted (and, we might add, very


costly) effort and are making some headway. The extent to
which these efforts will ultimately be successful is conjectural.
IBM's volume and position in the market permit them to employ
a number of countermoves that are not available to their com-
petition, including large rental concessions, free programming
and conversion, used conaputers at substantial discounts, and
last, but by no means least, the shelter of a very substantial
profit margin.

In any event, to the extent that other competitors are
successful, it will be even more difficult for Bendix to estab-
lish an equivalent image in the buyer's mind and a significant
place in the market.

(5) It Is a High-investment Business in Which No Profit Is
Returned until the Equipment Has Been on Lease for
More than Four Years

One of the most important features of the computer busi-
ness is its lease orientation. The importance of this and its
effect upon the product line development and marketing proc-
esses were discussed in our 1957-1958 report. It is of suffi-
cient importance to warrant repetition here.

40-927 O - pt. 7-21


It is not enough to simply lease and install a computer.
Once installed, it must, if at all possible, be kept in place until
the investment represented by it is returned and a profit obtained.
Furthermore, if the manufacturer is going to remain in the com-
puter business, he must be prepared to replace the equipment
with another system which can handle the increased work load
or provide the additional capabilities required by the user.

Based upon the present rate of new product introduction
and current industry pricing policies, this cycle takes a little
more than four years. If the installed equipment is returned
to the manufacturer in less than four years because of com-
petitive pressures or a change in the user's requirements,
the manufacturer is either faced with a direct loss, or the
need to "re -lease" the unit, at least doubling the marketing
cost for that imit and further deferring the break -even point.

Thus, the successful (and by this we mean profitable)
computer manufacturer must constantly follow a strategy of
well -planned product development and carry out a closely
coordinated marketing program. This strategy must be
sensitive to two distinct pressures: competitive and the
needs of the users.


Equipment with a competitive technological advantage
may be successfully marketed on the basis of this advantage
until it is equalled or surpassed by one or more competitors.
Once it is surpassed, the market is further divided and the
units already installed come under increasing pressure from
technological or economic obsolescence. In a market where
the product is sold rather than leased, the loss due to obsoles-
cence usually falls to the buyer - in this market, it falls to the

The second source of pressure upon the manufacturer
stems from the almost constantly changing requirements of the
user. Computer usage expands both horizontally and vertically.
Horizontally, in that a large portion of the scientific users are
moving into data processing-type problems. Similarly, machines
previously used only for business data processing are being in-
creasingly employed for the solution of scientific problems or
complex business problems that require extensive computation.

Vertical expansion occurs as a particular user approaches
machine saturation in the course of solving ever -increasing
numbers and types of problems. Pressure then mounts to ex-
pand the capacity of the present system or replace it with a unit
that is more closely geared to the changed requirements. Most


users would prefer to move up within the product lines of their
present suppliers, if their prior experience has been satisfactory,
since it generally minimizes reprogramming, media conversion,
and changes in standing relationships between buyer and supplier.

The net effect of these pressures toward vertical and hor-
izontal product line expansion is increased by the lease orientation
of the business. If a manufacturer fails to offer capable equipment
in all application areas, one or more competitors will use this as
a wedge to drive out his equipment and replace it with their own.
Bendix did this with the G-15 when IBM failed to provide the
IBM 604 scientific user with a suitable computer. Likewise,
the superior data processing capability of the IBM 1620 will un-
doubtedly force a number of G-15's out of the market. Similarly,
the successful manufacturer must also offer larger equipment, so
that his users may grow within his line - not shift to a competitor.
The problems facing Bendix with respect to the G-15 were dis-
cussed earlier. Likewise, similar pressures will eventually
come to bear on the G-20. We do not believe that the problem
of vertical product line expansion can be avoided, for it is one of
the basic characteristics of the market. The manufacturer who
does not offer a full product line will, in all probability, be forced
out of the market by those that do.


The above provides a brief, but succinct, description
of the computer business or market and the current Bendix
position. Given this situation, then, what is the best future
course of action for the computer division?


It seems fairly apparent that Bendix must choose, in the main,
from these five alternative courses of action:

(1) Concentrate on selling to technical centers for engineering-
scientific applications, which may include some subsidiary
data processing applications.

(2) Expand into serving the computer market in all its facets,
that is, selling in all markets for either data processing,
engineering-scientific, or combined applications.

(3) Endeavor to participate in the market on a limited scale,
exploiting small, profitable segments of the market
wherever they might develop.

(4) Redirect the efforts of the computer division to some
market other than general purpose digital computer

(5) Withdraw from the computer business and dispose of
or liquidate the computer division.

The fifth possible course of action is presented at this point in the
discussion only in the interest of completeness. It should be held in the
background to receive serious consideration only if success cannot be


foreseen as a result of any of the four positive approaches. Should it
finally appear to be the course to follow, considerable work would be
needed to develop the details so as not to impair Bendix's position in
other present or future markets.

The first three alternative courses of action are purposefully
phrased in terms of marketing. As previously pointed out, such
differences that may exist between medium to large scale data proc-
essors and scientific computers are relatively unimportant. Certainly,
they are not such that their effect upon the nature, magnitude, or cost
of the required research and development program would be significant
in selecting the correct course of action for Bendix. In either case, the
dimensions of the R&D effort will be governed by the extensiveness of
the product line and its associated peripheral equipment. To the extent
that more accessories and programming support are called for in the
various data processing application areas, it would require more engi-
neering and systems support effort than would be the case if sales were
limited to the purely scientific portion of the market. This extra work
can hardly be a policy determinant at this time.

The question is not, can Bendix design and make computers; it is,
can it lease or sell them in sufficient quantities to be profitable? Some
insight into the answer to this question is provided in the next three



An internal study conducted by the computer division in the spring
of 1960 indicated that annual placement of twenty-four G-20's and one
hundred and two G-15's or the equivalent is required for a profitable
operation. This study is now almost a year old and may well be subject
to question as to some of its details; nevertheless, it does provide a
frame of reference against which to view Bendix sales prospects.

It appears to us unlikely - in fact, almost impossible - for Bendix
to achieve this volume if it concentrates on selling to technical centers
for engineering-scientific applications, even with such incidental carry-
over into data processing applications that would occur. Our reasons are:

(1) To Place Twenty-four G-20's Would Require Bendix
To Capture 90% of the Market after IBM and Its
Closest Rival Had Taken Their Probable Share

Our study, as detailed in Appendix B, indicates that
approximately 750 computer systems in the G-20 category
will enter the market for essentially engineering-scientific
use. Of these, some 450 will be net additions to the exist-
ing population, while an additional 300 will be replacements
for presently installed equipment.


Within the new market, IBM and Remington Rand (or its
successor as second place contender) can be expected to get at
least 80% of the 450 units, leaving approximately 90 for all other
competitors. Within the replacement market, IBM may be ex-
pected to obtain approximately 85%, since most of the units to
be replaced are IBM products. This leaves an additional 45
units for all other suppliers. Thus, the total market available
to Bendix and as many as eight other competitors is in the
vicinity of 135 units, over the next five years. This total is
15 units more than Bendix feels it needs for profitable operation.
Furthermore, Remington Rand, RCA, Minneapolis -Honeywell,
GE, Control Data Corporation, and possibly others are going after
this business at least as aggressively as Bendix.

It seems obvious that, if these figures are anywhere near
correct, it is virtually impossible for Bendix to meet its G-20
volume needs under this restrictive course of action.

With regard to the correctness of the figures, it should be
pointed out that complete and precise data, even on the history
of this market is hard to come by. Furthermore, the market
itself is definitely in an unsettled period where neither the con-
sumer nor the manufacturer is very certain as to what will happen


next. In order to cross-check our forecasts, two completely-
different methods were employed. The basic approach was the
same as that used in our 1957-1958 study and was based upon
historical and probable future expenditures per engineer and
scientist for computers. This estimate was verified by a
correlation analysis and projection based upon federal ex-
penditures for research and development. These two fore-
casts fall within an 8% range of estimate. These estimates
are further detailed in Appendix B.

The one prime source of possible error is in the assign-
ment of 80% of the new market to IBM and one other manufac-
turer. Historically, IBM and Remington Rand have accounted
for 85% or more of the medium computer market and some 96%
of the large systems. Likewise, the assignment of an 85% share
of the replacement market to IBM may be subject to question
(though it was borne out during the field interview phase of our
assignment). Even if we assume that Remington Rand suddenly
withdrew from the market and IBM only obtained 50% of the new
and replacement markets, this still leaves only 375 units for all
other competitors, of which Bendix would have to get 32%. We
do not believe that our estimates are this far off, but even if
they were, it is unlikely that Bendix can, with the sales effort


contemplated in the aforementioned profit analysis, obtain a
32% share with a single modern computer against the other com-
petitors. With a greatly increased sales effort, it is conceivable
that Bendix might achieve this penetration, but at the same time,
the increased sales expense would move it into an unprofitable

(2) The Market Will Absorb Approximately 2, OOP Small
Computers during the Next Five Years; Less Than 75
of These Will Be Bendix G-15's

Analysis of recent market trends and their probable effect
on future new money entering the market leads us to believe
that approximately 1, 200 units will be added to the current
small scientific computer population by 1965. In addition,
approximately 800 units will enter the market as replacements
for currently installed computers and punched card calculators
that have reached economic or technological obsolescence.

Although Bendix has placed approximately 35% of the
small computers currently in use, it seems unlikely that this
.performance will carry into the future. The introduction of
the IBM 1620 second generation computer, will have a much
more significant impact upon the market opportunities for the
G-15 than did the IBM 610 three years ago. This newer machine


offers considerably more capability than the 610 and fits the
IBM product line much better, which will enhance the oppor-
tunity for systems sales. Other new computers that have or
soon will enter the market will also tend to reduce the G-15
market opportunity. As shown in the comparative evaluation
of the G-15 and its current competition in Appendix C, the net
effect of this changed product mix and the superior features of

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 31 of 140)