082 - statistical techniques and international trade

7/27/2019 082 - Statistical Techniques and International Trade

1/7


2/7

STATISTICAL TECHNIQUES AND INTERNATIONAL TRADE*

W are in a new industrial age. Al-most every type of activity in pro-duction and distribution has been re-made several times over, during the pastlo or 12 years, by new uses of statisticaltheory. The same statistical techniquesthat have remade marketing researchhave helped industry to achieve great in-creases in output without capital expend-iture for new machines or for additionalfloor space. They have saved scarce mate-rials, and have brought forth qualityand uniformity not dreamed of a fewyears ago.Everyone knows the impact of newstatistical techniques on marketing re-search, and on production, quality, anduniformity of product, but few peoplehave seen the day ahead when marketingresearch must be considered an impor-tant part of the statistical control of qual-ity. The reason is that good quality andthe right uniformity have no meaningexcept with reference to the consumer'sdemands.International trade is an essential com-ponent of prosperity, political stability,and peace. But international trade de-. xcerpts from an address delivered in Tokyothe 14th January 1952 at a meeting of industrial ex-ecutives sponsored by the Nihon Kezai and the Un-ion of Japanese Scientists and Engineers.

pends not only on a competitive price,but also on dependable quality and uni-formity, which in turn depend to a largeextent on intelligent marketing research.As most of you know, the simple statis-tical techniques of the control chart have

brought forth in recent Japanese experi-ence increases in production that rangefrom 5 per cent to ego per cent, withoutexpansion of plant. Some firms are savinglo per cent of their raw material, com-pared with their performance of a yearago. In some cases, the savings are muchgreater. The Fuji Steel Company re-ported they were able to cut their fuelbill by 29 per cent over their perform-ance a year ago. All these achievements,and more, were reported at the conven-tion held in Osaka the 22d September1951 at the occasion of the DemingAward. In the mill of the Toyo CottonSpinning Company that I visited duringmy work here during the summer I 95I ,three quarters of the girls who had beenengaged in re-work and repair a year be-fore had been moved into production,because of the improved quality of theproduct. Result: increased production,better quality, greater uniformity.

The gain in production and profit, al-though considerable from such achieve-ments, are the least of the gains. The


3/7

MARKET ING NOTES 429

gains from a better competitive posi-tion through improved quality and abil-ity to lower the price are much greater,though more difficult to measure.In these ways, and in many other ways,statistical techniques help greatly to meetthe requirements for increased interna-tional trade.Moreover, I cannot emphasize toostrongly that increased use of statisticaltechniques can contribute vitally towardthe maintenance of private enterprise,which must depend more and more onthe continual improvement of the effi-ciency of production and of distribution,and on the continual improvement of thedesign of product, in respect to both qual-ity and uniformity, to meet the changingneeds and the demands of the consumer,wherever he may be.

In any manufacturing plant, raw mate-rials come in, and product flows out. Theraw materials may indeed be raw, or theymay be sub-assembliesor piece-parts fromanother manufacturer. Eventually, theproduct goes to people who will judgeit. The finished product may indeed befinished, and go to thousands or millionsof consumers, or it may be raw material

&qpliera ofraw materiala

for a relatively small number of othermanufacturers.Anyway, there will be a chain of pro-duction, as illustrated crudely by FigureI . By the statistical control of quality Imean statistical work applied in all stagesof production, from raw material to con-sumer, and back again (see Figure 3).T h e statistical control of quality is theapplica tion o f statistical principles andtech niqu es in all stages of pr odu ctio n,directed toward th e most economic man -ufacture of a product that is maximallyuseful and has a mar ket.Benefits of Quality Control

Through the full use of the statisticalcontrol of quality, from raw material tothe consumer, a manufacturer may ex-pect to achieve in some measure the fol-lowing advantages:1. Increased production, without invest-ment in capital equipment or expan-sion of plant.2. Better quality at lower cost, and bettersuited to the market.3. Better uniformity at lower cost, andbetter suited to the market.4. Savings on raw materials and fuel (a

Conemeratest ofmateriala

R&actione, aseembly, inspection----- \Teetr of processes, \/machinen, methods,rates, oosta

G FIGURE. T he Production LineC


4/7

particularly vital advantage when scar-city of materials threatens production).5. Better operating efficiency: (a) idle timeof machines decreased; fewer rejections;less scrap and re-work; (b) better pre-diction of the market, through con-sumer research, by which the purchaseof materials, and the expansion andcontraction of the plant are carried outrationally, resulting in better economythan would be possible otherwise.6. Decreased inspection, but with in-creased assurance of dependable quality.7. Greater precision of dimensions whenrequired (as when parts are to be inter-changeable).8. Design, better suited to the market,through consumer research, carried outby modern methods of sampling anddesign of experiment.

9. Stronger competitive position, throughability to meet world requirements inprice, quality and uniformity.10. Use of an international language bywhich to furnish statistical proof ofquality and uniformity.

I must now ask you to think with mein broad terms just what good qualityand economic production really mean.Some people think in terms of pricealone. Others think in terms of qualityalone. Some people think of economicproduction as saving lo per cent in thecost of some operation. All these ideasare important, but we must go deeper.

In the first place, price has no meaningexcep t i n term s of the quality of th eproduct. But that is nor enough. "Goodquality" and "un ifo rm quality" have n omeaning except with reference to theconsumer's needs. This is why I mustspeak to you in terms of the entire pro-duction line, which begins with the pro-ducers of your raw materials, and endswith the consumer-the man who usesyour product

Incidentally, the consumer is more im-portant than raw material. It is usually

T H E J O U R N A L 01;M A R K E T I N Geasier to replace a supplier of raw mate-rial with another one than it is to find anew consumer. And a non-consumer, onewho has not yet tried your product, isstill more important to you, because herepresents a possible additional user ofyour product. For this reason I shallspeak to you of consumer research, be-cause I believe in the importance of con-sumer research to Japanese industry, par-ticularly for its export trade.Impo rtance of Co nsu me r Research

Some manufacturers think of con-sumer research as analysis of complaintsfrom purchasers and users. Certainly, noone can deny the great importance of theanalysis of consumers' complaints. Nomatter how silly and unjust a complaintmay be, i t is still important to a manufac-turer, because it shows him where he hasfailed i n pub lic relations-that is, failedto make clear to the public just whatquality they have a right to expect whenthey buy his product. A legitimate com-plaint helps the manufacturer to im-prove his quality, provided he has realquality control, and can be used by themanufacturer in tracing the cause oftrouble all the way back to productionand raw material.

But complaints are only a part of theproblem of public relations. Complaintscome from a very biased sample of con-sumers. Complaints do not provide com-munication with the other consumersnor with the non-consumer.

For reliable and economical communi-cation with the consumers and non-con-sumers of a product, it is necessary tocommunicate with all users and non-users, through the medium of statisticaltests and surveys. The aim of this particu-lar aspect of quality control is re-designof the quality of your product (Figure 3);also adjustment of your plant, and con-traction or expansion of the output of


5/7

M A R K E T I N G N O T E S 431

particular products to meet rat ionallypredicted changes in demand.As the consumer has been left out of

the chain of production in many parts ofthe world where industry needs expan-sion, I shall go into more detail on thesubject.

The main use of consumer researchshould be to feed consumer reactionsback into the design of the product, sothat management can anticipate chang-ing demands and requirements and seteconomical production levels. Consumerresearch takes the pulse of the consumer'sreactions and demands, and seeks ex-planations for the findings.

Consumer research is not merely sell-ing. Real consumer research, geared todesign and production, is an indispensa-ble modern tool for modern problems.

Consumer research is communica t ionbetween the manufacturer and the usersand potential users of his product, likethis:

Manufacturer -{ The userand the non-userThis communication may be carried

out reliably and economically today onlyby sampling procedures and tests de-signed according to modern statisticalprocedures. Through this communica-tion the manufacturer discovers how hisproduct performs i n sewice , what peoplethink of his product, why some peoplewill buy it, why others will not, or willnot buy it again, and he is able to rede-sign his product, to make it better asmeasured by the quality and uniformitythat are best suited to the end-uses of theproduct and to the price that the con-sumer can pay.

Consumer research acts as a governoror servo-mechanism which by probinginto the future market regulates both thedesign of the product and the amount ofproduction.

In the olden days, before the industrialera, the tailor, the carpenter, the shoe-maker, the milkman, the blacksmithknew his customers by name. He knewwhether they were satisfied, and what heshould do to improve appreciation forhis product. With the expansion of in-dustry, this personal touch was lost. T hewholesaler, the jobber, and the retailerhave now stepped in, and in effect haveset up a barrier between the manufac-turer and the ultimate consumer. Butsampling, a new science, steps in andpierces that barrier. The manufacturerof today, but for sampling, would be ou tof touch with the people who use hisproduct, or those who might use it.Manufacturers used to think of manu-facturing in three steps, as shown in Fig-ure 2. Success depended on guess-work-guessing what type and design of productwould sell, how much of it to make. Inthe old way, the three steps of Figure 2are completely independent.

1 2 9-A__Design it Make it Try to sell itFIGURE. The Old Way

I n the new way, management intro-duces, through consumer research, a 4thstep, and runs through the four steps in acycle, over and over as in Figure 3, andnot in the line of Figure 2.

1. Design the product (with appropriatetests);2. Make it; test it in the production lineand in the laboratory;g. Put it on the market;4. Test it in service; through market re-search, find out what the user thinks ofit, and why the non-user has notbought it;5. Re-design the product, in the light ofconsumer reactions to quality andprice;Continue around and around the cycle.


6/7

432 TH E J OURNAL OF MARKET ING

FIGURE. The New Way

This 4th step in Figure 3 was impossi-ble until recently-i.e., it could not becarried out economically or reliably. In-telligent manufacturers have always beeninterested in discovering the needs andthe reactions of the user and of the poten-tial user, but until recently they had noeconomical or reliable way of investigat-ing them.

The 4th step, communication betweenthe manufacturer and the user and thepotential user, gives the public a chance.I t gives the user a better product, bettersuited to his needs, and cheaper. Democ-racy in industry, one might say.A still better way is to begin the manu-facturing and marketing of a product on

a pilot scale, and to build up its produc-tion on a sound economic basis, only asfast as market conditions indicate, re-de-signing the product from time to time inthe light of consumer needs and reac-tions. The cycle is best taken on a spiral,as in Figure 4.

It is not to be supposed that the firstthree steps are the same in the figuresthat display the old and new ways. Con-

sider, for example, design in step 1.Proper design today means not only at-tention to color, shape, size, hardness,strength, and finish, but attention also toa suitable degree of uniformity. Paradox-ically, through the statistical control ofquality, great uniformity often costs lessthan nonuniformity without statisticalmethods. However, with some products,extreme uniformity may be costly, andthe manufacturer must be careful andnot price his product out of the market.In consumer research, the 4th step, themanufacturer studies the requirementsof uniformity, as well as of color, shape,size, hardness, etc. Then , through statisti-cal procedures he achieves the requireduniformity with economy, and his controlcharts furnish proof of th e un ifo rm ityachieved in an international language,known and studied now the world over.Consumer research is a continuousprocess, by which the product is im-proved continually and modified to meetchanging abilities of the manufacturerand changing requirements of the con-sumer. Consumer research, used intelli-gently, enables the manufacturer to runhis factory on an even keel, neither


7/7

M A R K E T I N G N O T E S 433greatly over-producing nor under-pro- it is only one of the indispensable statisti-ducing. He is not likely to let out goo cal tools of production.men one month and then try to recover W. EDWARDSEMINGthem the next. Consumer research usedto smooth out production is a powerful Bureau of the Bu dge t andfactor in economical production. Again, New York University

082 - statistical techniques and international trade

Documents