rajesh sirji

8/6/2019 rajesh sirji

http://slidepdf.com/reader/full/rajesh-sirji 1/14

CASE STUDIES IN BUSINESS

REACHING THE CONSUMER THROUGHDIRECT PERSONAL SELLING

A competition becomes keener, theproblem of increasing sales vol-ume so as to retain a normal pro-

portion of the market, or to make thisproportion larger, becomesa perplexingone to many manufacturers. One solu-tion that has proved effective in the case

of several nationally known manufac-turers has been that of selling direct tothe consumer through canvassers. Sosuccessful have a few of these firms beenthat numerous concerns engaged inother lines are considering the possi-bility of selling their own products inthis way.

It should be borne in mind by thosewho propose to substitute this method

of selling for another now in effect thatone hears only of the successes in thisfield. The instances in which firms havetried direct personal selling and havefound it inapplicable or ineffective arenot brought to the attention of thebusiness world beyond the few personswho are immediately concerned. Care,therefore, must be taken in drawingfrom the experience of a few firms theconclusion that this methodis the solu-tion of the distribution problem of allwho sell to the general public. It is thepurpose of this discussion to bring outsome of its uses and limitations.

' It should be understood that so-called wagonretailers, merchants who have no other place ofbusiness than their wagons and who make regulardeliveries at the householder's door, are not in-cluded in the above, although they handle con-venience goods. The handling of necessities forwhich there is a renewed need daily, the frequency

with which calls are made, and the lack of neces-sity of making a sale at each call are points of

Obviously, the type of goods to besold will have a bearing on whether ornot sales may be made through can-vassers. Certain specialties may requirethis method of selling, at least duringthe time they are being introduced to themarket, while goods havinga shopping

appeal, or those bought on a conve-nience basis, may find it a poor plan. Ineach of these fields there are successfulexamples. The question that arises iswhether concerns which are not pio-neers, and which, therefore, cannotenjoy a practical monopoly of theexceptions to ordinary buying habits,can secure a sufficient volume of sales inthis way to make it a profitable meansof selling their products.

Convenience goods usually do notlend themselves to this manner of dis-tribution .' Th is type of goods com-monly is bought at the nearest store asthe need arises. The purchaser doesnotwish to postpone his purchase; if hecannot get what he wants at the neareststore, he will go to the next nearest.Because it is impossible, except in asmall number of instances, for the can-

vasser to be at the purchaser's doorwhen the need for his product arises,many sales undoubtedly are lost. Thesuccess of the Fuller Brush Com-

difference. The function of the milk dealer, forexample, is more that of delivery on • prear-ranged contract than it is selling.

Missionary selling direct to consumers througha staff of canvassers, for the purpose of ttimulatinga market which is regularly served by retailers, isalso excluded from this discussion, because this isessentially a tales promotion rather than a lellinc

scheme. It is not intended to replace the estab-lished methods of distribution.



CASE STUDIES IN BUSINESS 95

pany in this field is believed to have re-sulted largely from the fact that there

are a sufficient number of exceptions tothe usual buying habits to furnish a suf-ficient market for one firm. The lineof high-grade specialty brushes handledin addition to the convenience line hasalso probably had some bearing on thema tter. It will be interesting to notewhether competition using the samemethod of selling increases the marketfor the part of this product which issold direct to the consumer, or merelydivides it among the firms in the field.

Unless there is a radical change inthe buying habits of consum ers, andthere seems to be no basis for assum-ing that such a change is to occur in thenear future, it appears unlikely that alarge part, or even an appreciable part,of the convenience-goods market willbe served by direct personal selling.The problem of properly timing calls in

such a way as not to lose sales is re-garded as a serious one when severalfirms serve the same market. T hi s,added to an already high sales cost,brought about by conditions which willbe explained later, would not pen nitoperation in competition with estab-lished convenience-goods retailers.Then, there is also the difficulty that theunit of sale in this class of goods isusually so small that the canvasser'scommission is insufficient to interesthim. This increases the sales-forceproblem, which at best is a troublesomeone under any forni of canvassing.

In the cases following, the questionsof direct personti selling of shoppinggoods and specialty goods under differ-ent conditions are taken up.

Shopping Goods

The case of the Antrim Hosiery Millsill f h diffi l i

countered in selling shopping goodsdirectly to consumers through can-

vassers.The Antrim Hosiery Mills, locatednear Cleveland, in 1922 had been manu-facturing and selling hosiery for men,women, and children, for more than ageneration. Un til 1922 Antrim hosierywas of a low-priced, durable grade, andwas designed to retail for from 25 to75 cents a pair. A t that time the execu-

. tives of the company decided to producemore stylish hosiery as well as to im-prove the quality. Th is decision wasmade as the result of the change in con-sumers' buying motives during theprevious 10 years. Various styles ofwomen's hosiery were designed to bemade of mercerized cotton, fiber silk,fiber silk and pure silk mixed, pure silk,and silk and wo ol. A few styles ofhalf-

hose for men were produced in thesematerials. The principal buying motives

of the consumer to which the companyappealed were quality of the product,price, and satisfaction of personal ap-pearance. Th e hosiery was designed toretail for from 25 cents to $1.50 a pair.The largest portion of the improvedhosiery to be sold under a separatebrand. Regal Antrim, retailed for $1a pair.

The Antrim Hosiery Mills was

distributing the major portion of itsproduct through wholesalers, and the re-mainder, about 5%, to chain-stores andretail buying synd icates. Sixty per centof the product was sold under the An-trim brand, and the remainder underwholesalers' private brands. From 1915to 1919 an advertising campaign hadbeen conducted by the company in twonational weekly ma gazines. From 1919to the early part of 1923 the companyh»d restricted itself to advertising inlocal new spapers. In 1923 the Antrim



HARVARD BUSINESS REVIEW

mence national advertising in threenational periodicals.

At the same time the decision hadbeen made to produce more stylish ho-siery, the decision also was made to selldirectly to department stores. Distri-bution through wholesalers was notproving successful, caused, in part, bythe change of hosiery from a "conve-nience" to a "shopping" article, and thegrowth of department stores, and, inpart, by increased competition amongmanufacturers endeavoring to distribute

through wholesalers. During 1923 thecompany succeeded in securing approx-imately 75 retail accounts, but only asmall number of these retailers pur-chased more than a few hundred dollars'worth of hosiery from the Antrim Ho-siery M ills. The transition from whole-sale to retail distribution was a slow anddifficult proceeding. The re was a con-stant risk that the policy of sellingdirectly to department stores would soantagonize the company's wholesalecustomers that the latter at any timemight refuse to buy from the AntrimHosiery M ills. In a few territories thisalready had become a fact.

During 1920 and 1921 the companyhad undergone severe financial losses,with the result that it was essential forthe mills to increase sales as rapidly aspossible. In the president's opinion75 %of all the mills attempting to distributethrough wholesalers was sufficient tosupply the demands of the latt er. H estated, however, that the productionof 95% of the mills attemptingto sell directly to department stores andimportan t retaile rs was necessary to fillthe needs of these two classes of pur-chasers. Fo r this reason, the executivesof the company were convinced that theAntrim Hosiery Mills had a greater op-portunity to increase sales by sellingdirectly to department stores than by

continuing to distribute only throughwholesalers. In December, 1923, thenormal production capacity of the millswas about 32,000 dozen pairs a week,although the sales department was ableto sell only about 16,000 dozen pairs.At this same time, the president of thecompany became interested in the suc-cessful experiment of the RichterHosiery Company, which was sellingdirectly to consumers; the phenomenalsuccess of this company suggested tothe president that the Antrim Hosiery

Mills might adopt a similar form ofdistribution in addition to those chan-nels already in use.

In 1920 the Richter Hosiery Com-pany had been selling about8,000 pairsof hose a day. The company had hadno brand of its own, and sold only towholesalers. Distribution under thismethod had not been satisfactory; hencethe company had decided to sell directlyto consumers. Four styles of hosierywere produced: pure silk, and silk-and-wool mixtures for wom en; pure silk, andsilk-and-wool half-hose for men. Thesesold at retail for from $1.30 to $1.50.This small number of styles had beenmade possible by selling directly to con-sumers. For a manufacturer sellingdirectly to the retail trade, it was neces-sary to have from 20 to 30 styles. Therewas a production saving, therefore, re-

sulting from this method of distribution.In 1923 the Richter Hosiery Companywas carrying on its sales efforts in thefollowing m ann er: Th e United Stateswas divided into 275 sales territories,each of which was in charge of a salesmanager whose expenses and remunera-tion were met by a small commission onthe number of pairs of hosiery sold inhis territo ry. Each te rrito ry was di-vided into districts small enough for anagent to cover three times each year bya house-to-house canvass. The agent



CASE STUDIES IN BUSINESS 97

received a commission of approximately18% of the sales price. Thi s commis-sion was paid in cash by the consumerwhen the order was given, and theremainder was paid c.o.d. when thehosiery was delivered by mail. TheRichter Hosiery Company conducted anational advertising campaign at a costof about $500,000 a yea r. Tw o fullpages a month were taken in a nationalweekly magazine and one full page amonth in five monthly magazines. In1923, as a result of this sales plan, the

sales of the company had been increasedto 125,000 pairs a day. Gross sales for1922 amounted to approximately $18,-000,000, and a profit of $600,000 wasshown for six months in 1923. Thesuccess of this company had led thepresident of the Antrim Hosiery Millsto investigate the plan of selling directlyto consumers. If selling directly toconsumers were to become the most

profitable method of distribution, itmight be the ultimate solution of theproblem of the Antrim Hosiery Mills.

Three Plans Suggested

There were three plans suggestedwhich included the employment ofagents to solicit orders directly fromconsumers. The first two were lessradical than employing an organizedstaff of agents under a plan similar tothat used by the Richter Hosiery Com-pany. Th e last method suggested wassimilar in all essentials to that employedby this concern. The first method in-volved giving a commission of 20% toindividuals selling stockings directly toconsumers, because of a personal inter-est which they had in the AntrimHosiery Mills. At that time, many of

the executives and salesmen and wives ofthe executives and salesmen of the com-pany secured orders directly from their

friends, which they forwarded to themill. In addition, there were friends ofthe executives who also solicited ordersdirectly from their friends. Some ofthese orders were filled direct from themill and others from the Antrim Ho-siery Shop, a retail shop operated at themill. The efforts of all these individ-uals resulted in valuable advertising forthe mill's product, and introduced theRegal Antrim brand to many desirableconsum ers. If all individuals sending inthese ord ers w ere given a commission of

20%, their interest would be stimulatedmore actively and a greater number ofdirect orders might result. Th ese indi-viduals, acting as agents, could collecttheir commission in cash when the orderwas taken and the mill would receivethe remaining 80% of the sales pricec.o.d. when the order was filled. Ifthis system were placed under the juris-diction of the Antrim Hosiery Shop,the mill itself would be relieved of theexpense and inconvenience of fillingthese ord ers . Such a plan, however,probably would not result in a largeenough increase in sales to warrant risk-ing the possibility of antagonizing thecompany's distributors. Furthe rm ore,the company would have little controlover a sales force of the type that wouldbe interested in this plan.

The second method was to employagents to make a house-to-house canvassof customers only in those cities inwhich the sales department desired tosecure retail customers. If the salesmancould state that a certain number ofconsumers in the town already had pur-chased Antrim hosiery, which they hadsecured directly from the mill, and thatthese consumers might become custom-ers of the retailer, there was more

likelihood of securing the desired retailaccount. In addition to this advantage,orders would be received by the hosiery



HARVARD BUSINESS REVIEW

shop. The canvassing would serve asadvertising for the Antrim product.This plan differed in one essential fromthat used by the Richter Hosiery Com-pany, because the canvassing was to beonly temporary. Consequently, it wouldbe extremely difficult to secure compe-tent agents who lived in the city se-lected. It was likely th at a staff ofcanvassers might have to be employedto travel from city to city in advance ofthe retail salesmen. Th is would add thecost of traveling expenses to the liberalcommission of 20% . Because it wouldtake one agent at least a week to solicitthe desired number of consumers ineach town, while a retail salesman couldcover the same city in one day, it wouldbe necessary to have six agents for everysalesman. It seemed doubtful wh etherthe heavy expense incurred in such aplan would be justified by the resultssecured.

The last method wasto adopt a plansimilar in all essentials to th at employedby the Richter Ho siery Company. Inthe experience of this latter company,the proportion of the selling price to theconsumer secured by the mill was littlemore under this system of distributionthan under the established methods em-ployed by the Antrim Hosiery Mills.The greatest proportion of saving re-sulted from the fact that it was possible

to reduce the number of styles to threeor four and concentrate on these. Thesuccess of the Richter Hosiery Com-pany had shown this method to be suc-cessful. This means of distributionprobably would destroy the existing dis-tribution of the Antrim Hosiery Millsthrough wholesalers and retailers fortwo reasons. Firs t, the retailers andwholesalers might be so much antag-onized by such a policy as to refuse tobuy Antrim hosiery; second, it was prob-able that canvassing agents couldnot

be employed successfully without theguaranty that Antrim hosiery could bepurchased only from them.' There wasfurther objection that if a competinghosiery mill should enter the field ofdirect selling, the method would becomeless effective. If consumers were solic-ited frequently by agents from severalmills, they might become disgusted withsuch agents and refuse to receive them.The plan involved the difficulty ofbuilding up an organization of agents.In the experience of the Richter HosieryCompany, it had been proved necessary,in order to retain its agents, to enablethem to earn commissions amounting toabout $35 a week. Assuming this to betrue, it was necessary for agents of theAntrim Hosiery Mills to make sales of$175 a week. If the stockings weresold at $1 a pair, each agent should sell175 pairs a week. The norm al produc-tion capacity of the mill was 65,000dozen, or 780,000 pairs a week. About

4,450 agents would be necessary to sellthe entire production of the mill.It was suggested that if a separate

corporation were used to distributestockings on this plan under a differentbrand name, it might be possible to ex-periment with the system of distributionwithout risking the loss of the existingsales to wholesalers and retailers. TheAntrim Hosiery Mills could control

such a corporation and supply the ho-siery without allowing the wholesaletrade to know that the Antrim HosieryMills was connected with the undertak-ing. The re were the disadvantages thatthe value of the Antrim brand would belost, and the consumer would not havethe satisfaction of buying directly fromthe manufacturer. Th is method, how-ever, would enable the Antrim Hosiery

'Where manufacturers sell through canvassersand also through retail dealers, it is customary tomake a distinction between the items sold by each.



CASE STUDIES IN BUSINESS 9 9

Mills to try the system and to securethe benefits of it if successful, and, ifunsuccessful, the company was insured

against the loss of the established dis-tribution.A few of the directors of the Antrim

Hosiery Mills were of the opinion thatany method of selling to consumers di-rectly could not be expected to sell animportant proportion of the mill's pro-duction. Th ey opposed, therefore, anyefforts to sell directly to consumers be-yond those already used. Th ey statedthat this plan would divert the attentionof the executives and salesmen from theattempt to increase sales to retailers.

It is believed that the opinion of thedirectors was a sound one, and that theAntrim Hosiery Mills should not haveattempted to distribute to consumersthrough canvassers. It is true tha t oneof its competitors had built up a largevolume of business in this manner, butdespite the opinion of the sales man-

ager of one of the largest firms engagedin this method of distribution that themore firms which ado pt it as a policy themore accustomed consumers will be toreceiving salesmen at their doors, andcc)nsequently, the less sales resistancewill be encountered, it is thought thatas soon as the visits of the house-to-house salesmen become frequent, eachsalesman will experience increased diffi-culty in getting a sufficient amount ofattention directed to his particularwares. Shopping goods particularly donot lend themselves to thia type of dis-tribution, for on such goods the cus-tomer wishes to make comparison as toquality, style, and price. T he succewof the competitor may be attributed tothe fact that he was in the field first,and not to the idea that goods of thiitype are sold readily in this way. T he

great bulk of such goods mustbt told

through shopping instttutioM, Mch «•

department stores; unless there is aradical change in the buying habits ofconsumers, the Antrim Hosiery Mills

would run a serious risk of reasoningthat, because the Richter Hosiery Com-pany succeeded in increasing its salesvolume by using an unusual method ofmarketing, the buying habits of con-sumers could be changed readily. Itmust be remembered that the large vol-ume of sales is still made through

, retailers.It would be difficult for the Antrim

Hosiery Mills to establish a separatecorporation for the distribution of ho-siery on a house-to-house basis withouthaving the fact known. Com petitorswould be quick to make capital of sucha ventu re. It is therefo re believed tha tit would be inadvisable for this firm toattempt direct distribution to consumersthrough personal salesmanship.

Selling Shoes Directly

T he case of the Arno t Shoe Companyconsiders specialty goods, as well as adifferent type of shopping goods. T heArnot Shoe Company, which was lo-cated in a mid-western city, had manu-factured a line of men's and women'smedium-priced shoes for approximately25 years. Its product had hecomeknown nationally and enjoyed a good

repu tation for quality. T he line wasvery complete, comparing favorablywith that of competitors in variety ofstylet offered. T he company sold itsshoet directly to retailers by meansof traveling talesmen. T he re was nohranch organization. T he shoes weredelivered from the factory, and the talesforce was controlled from a centralsales office located at the factory. Inthe past, retail dealers had been grantedexclutive agendet for Arnot shoet.

The manigement had allowed the



IOO HARVARD BUSINESS REVIEW

equipment' and processing methods tobecome gradually obsolete so that thecost of production had been increasing

at a more rapid rate than that of com-petitors, who had adopted more modernmethods of opera tion. W hile the man-agement realized the desirability of im-proving both machinery and processes,the financial position of the company atthat time would not permit an addi-tional investment to be made.

As a result of the higher cost of pro-duction, the price at which Arnot shoeswere quoted to retailers was higherthan that quoted by competitors on com-parable goods. Com petitive factors,therefore, forced the Arnot Shoe Com-pany to accept the less desirable re-tailers. The credit risk with this typeof dealer was high, and, in general,their selling ability and location werenot conducive to greater sales volumefor the Arnot Shoe Company. Th iscondition in the latter months of 1924

forced the company to appeal to itsbankers for financial aid.After a careful analysis of the finan-

cial needs of the company, the banking-houses decided that conditions weresuch that a reorganization should beeffected. In lieu of this decision, Mr.Harvey was appointed controller ofthe company to represent the bankers'interests.

Mr. Harvey realized that the diffi-culty in the past had been inability ofthe company to secure the desirable typeof retailers because of high productioncosts. The expense involved in remod-eling the plant, however, would begreat, and he desired to avoid the ad-ditional outlay at this time if some othermeans of saving the company could be

' Th e equipment leased from the U nited Shoe

Machinery Company had been kept up to date, butauxiliary equipment and processes had beentllowed to fall into obsolescence.

devised. H e believed that there was apossibility of changing the method ofdistribution so that the higher cost of

production could be offset. In the courseof his investigation of distributive chan-nels, the success of several large com-panies which were selling directly to theconsumer by a canvassing method cameto his attention. A house-to-housemethod of distribution would offer theadvantage of more intensive selling, andwould enable the company to overcomethe present difficulty of distributingthrough undesirable retailers. M r.Harvey, therefore, became convincedthat this method of distribution would,in a large measure, solve the company'sproblem. H e thought it desirable, how-ever, to consult several prominent shoemanufacturers, sales managers, andsales-management experts before pro-ceeding with the new plan. These con-ferences brought out many points thathe had not considered previously, andfinally it was decided that direct sellingwould not prove advantageous in thesolution of the company's problem.

The decision of the Arnot Shoe Com-pany not to use a house-to-housemethod of selling apparently was basedon two considerations. Th e first of theseconcerns the cost of distribution.House-to-house selling is not a means oflowering distributive costs, as is some-times believed. T he number of sales-men required for house-to-house sellingwould be many times the present salesforce which sells only to reta ilers. Asalesman selling from house to housecannot cover a large amount of terri-tory . His customers are many withina comparatively small area , and callsmust be made more frequently than inthe case of a salesman calling on re-tailers. Fu rthe rm ore , the inherent diffi-

culties in house-to-house selling makethe task of securing, training, and re-



CASE STUDIES IN BUSINESS IO I

taining salesmen a very perplexingproblem. The result usually is thatsalesmen of doubtful ability are ac-cepted. All of these factors tend tomake the personnel turnover high, and,in point of fact, some of the highlyorganized companies that sell throughcanvassers have a labor turnover ofover 150% annually.* The supervisionof such a large and scattered force ofcomparatively low-grade salesmen alsoconsiderably increases expense.

Because salesmen for house-to-house

canvassing are hard to secure, compa-nies using this method of distributionare compelled to offer relatively highrate s of commission on all sales. Thetotal sales-force expense, therefore,would be many times greater than theamount now paid to salesmen sellingonly to retailers. An additional expensealso is incurred in providing the in-creased number of salesmen with sam-ples. The line carried by house-to-housecanvassers, however, undoubtedly wouldbe smaller than that carried by sales-men selling to retail dealers.

When the manufacturer assumes theretailers' functions, he also assumes hisrisks. If credit were extended, the riskinvolved would be increased over thatat present. The number of accountswould be greatly increased and the typeof risk would be even poorer than that

of the retailers to whom the companynow sells. Th e Arno t Shoe Companyprobably would sell only on a cash basis,although such a policy would limit dis-tribution. Retailers ordinarily offercredit service, and, in the case of pur-chases of relatively high unit value, sucha service is utilized extensively. In ad-dition to an increased credit risk or alimited market, there is increased ship-

*See Advcrlising Fortnightly, January 16, 1924,p. 15.' IhlA.

ping expense. W hen shoes are deliv-ered in packages of one or two, thecosts will be greater than when it ispossible to make shipments in large rquantities. Fina lly, it would be neces-sary for the company to maintain acomplete in-stock dep artm ent. Shoescould not be manufactured after thecustomer's order was received, as is or-dinarily the case when a company sellsto retailers. Aga in, this is assuming afunction of the retaile r. T he manufac-turer enjoys some advantage in having

the stock centralized, but the quantitywhich must be carried in each style isgreater, and the chance for style loss,especially in women's shoes, conse-quently is increased.

It should be remembered that it is thefunction which costs and not the agencywhich perform s it. W hen a manufac-turer assumes the functions of the re-tailer, his costs also must be assumed.Consequently, the costs of selling di-rectly to the consumer by a canvassingmethod does not result in lowering theexpenses of distribution.'^

Buying Habits

The second point relevant to thedecision of the Arnot Shoe Companyinvolves a consideration of consumers'buying hab its. M en 's shoes are pur-

chased largely on the basis of depen-dability in quality, comfort, and appear-ance. A consumer can determine thequality of the shoes as readily in thehome as in the store. If a shoe whichhas been purchased from a canvasserproves defective, however, the consumermust return it to the factory for adjust-ment. Th is involves correspondence,delay in receiving the adjustment, andthe probability that it will not be satis-factory when it is received. The per-sonal element, which plays a large part



1 0 2 HARVARD BUSINESS REVIEW

in making satisfactory adjustments, islacking in the house-to-house method ofselling. In this respect, ther efor e, theretailer has an adva ntage. In order todetermine properly the comfort of ashoe and the way it will appear on thefoot, it is necessary that the shoe befitted and tried. Th is, howev er, is im-possible in house-to-house selling ofshoes, for, obviously, a salesman cannotcarry a shoe of every style and size. A sa result, customers must be fitted bymeans of measuring sticks and sizes ap-pearing in old shoe s. Such m ethods, at

best, are haphazard, and will result inmany misfits, with consequent discon-tent on the part of the purchaser. Re-turns also will be numerous because ofmisfits. Furtherm ore, as stated, a sales-man cannot carry a sample of everystyle; consequently, it is often necessaryto buy on the basis of photographicprints. Such me thods do not prove sat-isfactory in the case of a product whichto a large extent sells on the basis ofappearance. It seems obvious, there-fore, that the retailer has an insuperableadvantage over the house-to-housemethod in the merchandising of men'sshoes. His larger stocks make possiblemore correct fitting in a number of dif-ferent styles. Co mfo rt and appearancecan be carefully ascertained by the con-sumer before the purchase is made.

In the case of women's shoes, the out-standing buying motives are emulation,which involves the ttyle factor, comfort,and, to tome extent, dependability inquality. T he strength of these mo tivesvaries according to the personal desiresof each consumer. The motives of com-fort and dependability in quality havebeen discussed above in connection withmen's shoes, and they do not appear tobe different in the case of women's

shoes. The third motive, that whichinvolves the style factor, however, in-

troduces a new consideration. W om en'sshoes, in general, may be classified asshopping goods, for they are purchased

after a comparison of style, quality, andprice has been ma de. Fo r this reason,as brought out in the case of the AntrimHosiery Mills, house-to-house selling isundesirable.

There also are further difficulties inrecognizing consumers' buying habitsin a house-to-house me thod of sellingwhich do not fall within the classifica-tion of buying m otives. A s a generalrule, shoes are not purchased far in ad-vance of the time when they are needed.When shoes are purchased from a re-tailer, they are delivered to the customerimmediately or sent soon thereafter.When shoes are bought from a can-vasser, however, the order must befilled at the factory. Th is means de-lay; the delay varying in length as thedistance from the factory to the cus-tom er. Ad ditional delay occurs when

shoes do not fit or are not the stylewhich wa s ordered. T hi s not only iscontrary to consumers' buying habits,but also allows time for a change ofmind between the time the order istaken and the shoes delivered. A greaternumber of returns, with consequenthigher costs of selling, is the result.

The fact that shoes are purchased atrelatively infrequent intervals makesthe task of the salesman in properlytiming his visits one of great difficulty.If the salesman does not appear at theright time, the consumer normally m akesthe purchase at a nearby retailer's,for, as a general rule, the demand can-not be delayed greatly. T he salesmanalso experiences the difficulty of select-ing the best time of the day in whichto make his visit. T he A rno t line in-cluded both men's and women's shoes.

But men ordinarily are absent from thehome during the day, and it is imprac-



CASE STUDIES IN BUSINESS 1 0 3

ticabic to attempt to see them duringbusiness hours. When they return

home at night, however, they usuallydo not desire to be bothered by a can-vasser. It is practically impossible,therefore, for a canvasser to time hisvisits in a manner that will yield themaximum volume of sales.

Specialty Goods

The following case deals withspcicialty goods which have been sold toa large extent through canvassers. T heAllister Electric Company, in an effortto produce a vacuum-cleaner whichwould compete with the high-qualitycleaners on the market, had designed acleaner without regard to cost. Th ismachine,, ca lled Jthc A ll is te riiUiaxuir,. jRasbelieved to be superior to others inquality and efficiency of ope ratio n. Itcould be sold to the consumer for

$59.50, which was approximately thesame price as that asked for the highestgrade of competing machines. It wasdecided that the Allister cleaner couldnot be distributed successfully throughwholesalers, because wholesalers, byreason of the multiplicity of their prod-ucts, were unable to render the sales-development service necessary to aid theretailers in selling a new device.

In January, 1923, a method of dis-tributing directly to retailers was begunwhich was similar to that practiced byseveral other manufacturers of vacuum-cleaners. Sales wer e made to retailersunder two different plans, the first ofwhich provided for an outright pur-chase of machines at a discount of 40%.Und er the second m ethod, called the"resale plan," the retailer purchasedmachines at a discount of 20%, with the

privilege of returning all unsold clean-ers. The company agreed to placewith the latter retailers all orders which

it secured directly from consumers.Approximately one-third of all orders

obtained by the Allister Electric Com-pany were filled on the resale plan. Be-cause of returns, however, those orderscomprised only about 25% of net sales.

A sales force was organized, and thiswas controlled by 10 division managerslocated throughout the United States.They supervised district managers whosecured orders from retailers in theirdistricts and organized corps of can-vassers. Th e division managers weregiven commissions of 3% on sales intheir territories, and district managers,commissions of 6% . Canvassers dem-onstrated the machines in homes andsolicited orders from housewives; theywere paid a commission of 20% on all

retailers who purchased under the re-sale plan.

During 1923 the company expended

about $100,000 in advertising in na-tional mediums in order to establish thename of the Allister cleaner with thepublic. Ad vertisin g in the SaturdayEvening Post consisted of four doublepages in color and two single pages.Sales for the year amounted only toabout $350,000, but the advertising ex-pense was regarded as a necessary initialinvestm ent. It wa s difficult to increasethe volume of sales because of the lackof a sufficient force of canv asser s. T h edistrict managers were unable to obtainan adequate number, and found it diffi-cult to retain the services of those whowere successful. If a canvasser wer eemployed and failed to obtain an orderduring the first three or four days, hebecame discouraged and sought employ-ment elsewhere. It wa s estimated thatthe force of canvassers approximated

about one-fourth of the desired number.In July, 1924, a distribution plan was

id d hi h id d f h di



1 0 4 HARVARD BUSINESS REVIEW

continuance of the canvassing organiza-tion. Sales to retai lers under the resaleplan were to be discontinued, and thecompany was to secure contracts fromretailers for the outright purchase of aspecified number of machines for a year,at a discount of 40% . Division and dis-trict sales managers were to continueunder the arrangements already inforce. Fo r each machine sold to a re-tailer, the company agreed to spend $5in local new spaper advertising whichdisplayed the retailer's name. All na-

tional advertising was to be discon-tinued.It was expected that newspaper ad-

vertising of Allister cleaners in the salesdistricts, which would be more extensivethan any campaign previously under-taken by a manufacturer of vacuum-cleaners in the United States, wouldstimulate consumers to go to the storeswhere Allister cleaners were sold and

purchase those machines. The expecta-tion was based on the opinion tha tvacuum-cleaners no longer were spe-cialty articles which necessitated demon-stration and high-pressure selling meth-ods to secure orders. The company wasinformed that vacuum-cleaners were inuse in from 40% to 50% of the wiredhouses in the United S tates. The con-clusion was drawn that an increasingvolume of replacement orders could besecured, and that the use of vacuum-cleaners in almost half of the wiredhomes has established the machine as apractical household necessity. If thisopinion were sound, consumers might beinduced to purchase the article by meansof newspaper advertising, which wasused successfully to stimulate the saleof other staple products.

If consumers failed to make purchases

as a result of the newspaper advertising,and the retailers were unable to disposeof machines in stock the sales volume

of the company would be curtailed, andits reputation with retailers would beinjured. No other manufacturer of high-grade vacuum-cleaners had dependedwholly on local newspaper advertisingto stimulate sales of the product.The success of the plan dependedon the soundness of the company's opin-ion that newspaper advertising couldeffectively supplant canvassers in induc-ing consumers to purchase vacuum-cleaners.

The company decided to adopt the

proposed plan. It has now been inoperation for several months, and thesales volume of the company has notsuffered because of the discontinuance ofthe canvassing organization.

This case brings out three significantpoints concerning the distribution ofmerchandise by means of a canvassingorganiza tion. The experience of thiscompany, which is similar to that in thecases discussed above, seems to betypical in regard to the difficulty ofobtaining and holding canvassers. M ostsalesmen do not desire this type of work.The difficulty of approach and the manyrebuffs which are received in house-to-house selling prove to be a great ob-stacle both to new men entering thefield and also to older salesmen in theorganization. As a result, canvassersavailable to a company, in general, are

of limited ability; a type that finds itdifficult to obtain permanent sales workin other fields. The combination of thistype of salesman with the inherent diffi-culty of the work leads to a very highrate of turnover in personnel.

A second point of interest in this caseconcerns the use of retailers in connec-tion with a canvassing program. Wherethe unit of sale is relatively large, credit

often becomes an important factor inmaking sales. In such cases, the ser-vices of the retailer to the manufacturer



CASE STUDIES IN BUSINESS 1 0 5

are valuable. The reta iler is in closertouch with consumers in his locality and

can handle credit relations with less lossthan would be possible for the manu-facturer through his canvassers. Usingthe retailer in this manner seems par-ticularly advantageous for a companyintroducing a specialty product that willeventually find more efficient distribu-tion entirely through retail agencies.Contacts which will prove valuable inthe future distributive policy can bemade while the product still is in the de-velopmental stage.

The third, and perhaps the most sig-nificant of the points brought out, indi-cates the value of house-to-house sellingfor introducing a specialty articlequickly. House-to-house selling may beused most effectively when articles re-quire demonstration and educationalwork to induce the consumer to usethem. Home demonstration of specialty

articles that are not generally under-stood proves more convincing than storedem onstration. Fu rtherm ore, canvass-ing, which is an intensive method ofselling, acquaints a greater number ofpeople with the product in a short spaceof time than can be accomplishedthrough retail-store distribution. Oncean article has become well known andis found in general use, however, theneed for high-pressure selling disap-pears, and distribution through regularchannels becomes more effective.

General Significance

Methods of distribution should bebased on a careful consideration of con-sumers' buying hab its. In analyzing thesuccess of manufacturers who now em-ploy a house-to-house method of selling,one should determine whether or notthe success is based upon fundam ental

k i i i l i

In many instances, there are enoughvariations from normal buying habits

to insure the success of the limited num-ber who take advantage of such excep-tions. This explains the progress madeby some of the concerns which were firstin the field. As the number increases,however, it is probable that the numberof exceptions to usual methods of buy-ing will not increase to such an extenttha t all will be profitable. On the con-trary, it is likely that the existing mar-ket will be divided among the greaternumber of canvassers. It is believed,furthermore, that as the number of can-vassers increases, sales resistance alsowill tend to increase.

In addition to these general princi-ples, the following points of significanceare brought out by the above cases:

1. House-to-house selling ordinarilyis not a less expensive method of distri-bution than selling through the usual

channels.2. Convenience goods, that is, those

goods which are commonly purchasedimmediately after the desire for themis felt, usually should not be distributedthrough house-to-house salesmen.

3. Goods in whose selection the ele-ment of shopping plays a large part arenot readily adaptable to this type ofdistribution.

4. Specialty goods which are techni-cal in nature, and whose sale is depen-dent largely upon demonstration of per-formance to prospective customers,often require house-to-house selling.This is particularly true of productswhich are just being introduced. Be-cause house-to-house selling implies in-tensive selling, it is in this field that it ismost effective.

5. Specialty products which are well-known ordinarily are sold at lower salescost through other channels than the



io6 HARVARD BUSINESS REVIEW

sold on a house-to-house basis afterbecoming known may change to a lessexpensive method of distribution.

6. It is unlikely that house-to-houseselling and distribution through the or-dinary channels can be used effectivelyby a firm for the sale of identical mer-chandise.

7. W here good s are so bulky thatsamples cannot be carried, or where a

large number of samples is necessary,the plan becomes impractical. The samething is true, to a large extent, in in-stances where the product must be madeto fit the purchaser.

8. Eve n in cases wh ere direct sellingappears to be advisable, the difficultyof building up and maintaining a corpsof efficient canvassers is so great that itfrequently renders the plan inexpedient.

CERTAIN LIMITATIONS IN THE APPLICATIONOF SCIENT IFIC MANAGEMENT

period used by other manufacturers inproducing cheaper pianos, results in anextremely high inventory of raw mate-rials and goods in process.

No effort is made to speed up pro-duction because it is thought that sucha move would influence the quality ofthe work. Even during the war period,when orders were on hand for five

times the number of pianos the factorycould produce, no effort was made toincrease the output, for fear that thequality and tone might be impaired.The primary interest is to maintain areputation for a superior product.

One of the chief difficulties encoun-tered by the company is the maintenanceof a force of highly skilled and loyalworkers. The factory superintendenthires workmen, and the foremen areresponsible for training them to therequired degr ee of skill. T h e foremandoes this either by giving them instruc-tions himself or by having the new em-ployees work with the best men in thedepartm ent. N o written instructionsfrom the office are given to the work-men, the best method of doing eachoperation being left to the judgment ofthe individual workman.

In the Walker factory, methods em-ployed to give control are not as ex-

ti g th ll f d d

Walker Piano Company,which manufactures pianos of the

highest quality, has an output of 1,500units a year. Th ree types of grandpianos and two types of uprights com-prise the comp lete line. T he com panyemploys about 400 men in the factory,practically all of whom are working onday-rates. On ly a few of the unskilled

operations, such as the sawing of lum-ber, are on a piece-work basis.The company distributes its product

directly to 60 dealers located all overthe world. N o aggressive sales meth-ods are employed. Althou gh the W alk-er piano sells for about $300 morethan the next most expensive piano, be-cause of the superior qualityot its prod-ucts the company never has had anydifficulty in selling all the pianos it couldmake, even during years of depression.

In the process of manufacture, there-fore, no costs are spared to insure thebest quality of work. Grea t care istaken in the purchase of all materials,especially lumber. Individual trees areselected, and the lumber is seasoned inthe company's own yards for a periodof 3 years before it is put into the proc-ess of manufacture. Th is, togetherwith the fact that it takes the WalkerPiano Company 18 months to make a

i i t t t th 2 th



Harvard Business Review Notice of Use Restrictions, May 2009

Harvard Business Review and Harvard Business Publishing Newsletter content on EBSCOhost is licensed for

the private individual use of authorized EBSCOhost users. It is not intended for use as assigned course material

in academic institutions nor as corporate learning or training materials in businesses. Academic licensees may

not use this content in electronic reserves, electronic course packs, persistent linking from syllabi or by any

other means of incorporating the content into course resources. Business licensees may not host this content on

learning management systems or use persistent linking or other means to incorporate the content into learning

management systems. Harvard Business Publishing will be pleased to grant permission to make this content

available through such means. For rates and permission, contact [email protected].

rajesh sirji

Documents