Go To MKT CH 4
Go To MKT CH 4
Go To MKT CH 4
Exhibit 4.2 Factors Affecting Sales Costs per call increases. The agenda for direct sales visits to users of tech-
Load Centers and Air Disk nical producrs might include discussions of producr performance
Circuit Brcakets Compressors Drives characteristics, cost and performance comparisons with competitive
(6 fitms) (4 fitms) (2 fitms) offerings, product modifications to adapt the product to the user's
Average number of sales calls applications, prices, warranties, delivery schedules, operator train-
(monthly) 70 61 40 ing, and product defect claims.
Approximate percc:ntage ofsales In the case of sales visits to distributors, however, such calls
time: spent are often time-consuming for both technical and nontechnical
• ~ on user-customers to producrs. A typical meeting agenda might include distributor stock
- and service existing
products 13% 29% 55% levels, merchandise returns, new product features, competitive ac-
-introduce new products 8% 12% 27% tivity, distributor accounts, credit and collections, distributor busi-
• calling on outside sales ness practices, and product performance problems. In addition,
inBuenccrs and specifiers 4% 11% 0
• calling on distributors 59% 34% 10% manufacturers' sales personnel frequently accompany distributor
• callin~ on others 5% 5% 0 salespersons on customer visits.
• camp eting ~rts, attending Travel costs,.a major component of personal selling costs, are
mtttings; 0 er administrative
duties 11% 8% 8% a function of distances between the sales office and the sales reps'
$280
customers. For example, a sales rep selling oil country tubular
Average cost per sales call $129 $232
goods for Lone Star Steel commented, "If I'm calling on major oil
Total active user accounts called on 220 110 N/A
companies in Houston, I can only manage four calls a day at the
Total number of distributor most. If I'm visiring oil field supply houses in Tulsa, I can get to
locations called on 1,225 100 N/A
six. IfI'm ta1king to independent oil companies in Oklahoma City,
Sales reps' average total annual I can probably see ten a day, all of them on different floors in the
compensation $46,500 $37,000 $58,000
same building."
A sales rep may be assigned to call on all of the existing and
potential accounts in the region or on some subset of customers.
Persorud Selling. The costs of personal selling, excluding For example, maily of IBM's sales reps specialize in particular mar-
sales administration overheads, are a function of call duration, call ket segments, such as banking and finance, manufacturing, distri-
frequency, travel requirements, and time taken for such sales-related bution, transportation, government, education, or medical. Such
activities as reporting, attending sales meetings, and training. As specialization permits anyone salesperson to communicate with his
Exhibit 4.2 indicates, the use of sales personnel time and cost-per- or her customers about specific product applications. It results,
sales-call data vary considerably among three product-market however, in greater travel costs per sales rep than for a nonspecial-
groups we surveyed. ized rep calling on a comparable number of customers in a reduced
The differences in cost per call and number of calls per month geographic area.
reflected in these data may be explained by differences in call con- While a goal in field sales management is to maximize the sales
tent and duration. Generally, the more technical the product, the rep's "time over target," a significant percentage of his or her work-
greater the extent of customization; and the larger the purchase, ing hours is spent on activities such as completing customer call
the greater the number of customer personnel involved in the pur- reports and attending product-training and sales-briefing meetings.
chase decision, and the longer the sales call. Thus, as we move from Tune taken away from customer contact activities is also a major
the least technical product, load cenrers, to the most technical, disk sales expense category. As the data in Exhibit 4.2 suggest, the
drives, a typical salesperson makes fewer calls monthly and the cost amount of nonselling time varies among suppliers.
66 DESIGNING CHANNELS SYSTEMS Distribution Costs 67
Tune spent on sales-related activities will vary with the infor- comparison, field salesforce salaries amounted to over 3% of sales
mation needs of the sales rep, the customer, and management. For in 1985.
example, time spent for product training is influenced by the Nevertheless, industrial trade shows, attended by potential
breadth of the product line, product technicality, and the needs of customers, are a suitable vehicle for introducing new products, de-
customers for product information. The last will be related to the veloping sales leads, and maintaining relationships with existing
importance of the purchase and the extent to which product use is customers. The National Mine Service Company (NMS), for ex-
expected to cteate changes in the customer's operations and proce- ample, a major producer of underground mining equipment and
dures. supplies, reseives about 25% of its promotional budget for 4-6
Further, salesforce needs for nonselling time may be affected trade shows annually and a major exposition, the Ametican Mining
by changes in product or by personnel reassignments. Product- Congress International Coal Show, held every four years. NMS
training costs escalated sharply for the Honeywell Information Sys- spent over $1 million on the latrer event in 1980'>
tems Division (ISD), for example, when it reorganized its field In addition to demonstrating their product lines at trade
salesforce in 1986 into two product-specialized groups, one focus- shows, manufacturers also contribute funding and personnel time
ing on large systems (the DPS-8 series) and the other on small to distributors for demonstrations and displays, as well as paying
systems (the DPS-6 series). ISD management had sensed that its part of the cost of local advertising. The amoWlt that a manufac-
salesforce operated at a disadvantage in having to sell both lines. turer contributes to a distributor's advertising each year is often set
Product line breadth and rapid developments in computer technol- as a percentage of the latter's purchases from the supplier. The
ogy made it exceedingly difficult for the sales rep to represent effec- amount contributed is typically limited to a certain percentage-
tively the full range of Honeywell computers and application sys- say, 50%-of the actual cost of each ad. Other promotional funds
tems.· With the reorganization, the entire salesforce was put provide sales incentives to distributors in the form of awards and
through retraining programs to give sales reps in-depth knowledge prizes for meeting specified sales targets,
of their products, product applications, and markets. Producer expenditures on national and regional advertising are
While the purpose of product training is to aid sales reps in typically the smallest portion of industrial promotion. Ads are
communicating with customers, completing call reports serves in- placed most often in trade magazines to annoWlce new products,
ternal communications needs. Typically, sales call reports provide develop brand familiarity, increase customer· receptivity to sales
information on individual account activity, personnel changes, calls, build goodwill among distributors, and generate sales leads.
competitive developments, customer service needs, and problems Nevertheless, promotional communication in industrial mar-
having to do with delivery, credit, and product performance. The keting seems. to be supplemental to personal selling. In particular,
time required for sales call reporting is related to the needs of the expenditures on advertising and promotion are usually small. In
manufacturer for such information in developing and implement- addition, they seem to be among the first expenses to be cut in
ing marketing strategies. economic downturns.
SlUes Promotion. Compared with consumer goods manu-
facturers, industrial goods producers spend relatively little on me- Taking and Processing Orders
dia advertising, direct mail, and trade shows. While advertising-to- The costs of paperwork-for order receipt, shipping, billing,
sales (NS) ratios of 10-20% are not uncommon for items such as credits and collections, and accoWlt adjustments-are not of the
breakfast cereals, cosmetics, and proprietary dtugs, the six manufac- magnitude ofpersonal selling costs but are often significant. In fact,
turers of load centers and circuit breakers that we surveyed had the costs of order receipt, collections, and adjustments are univer-
average NS ratios of .2%, while their expenditures for trade shows,
demonstration, and sales contests amounted to .5% of sales. In , Thomas V. Ilonoma, "Get MOl< Out of Your Trade Shows," H..--.l B _ Rme1II Jan-
uaty-Fcbtuaty 1983, p. 75.
68 DESIGNING CHANNELS SYSTEMS Distribution Costs 69
sally Wlderstated in the respect that salesforce time spent on these Exhibit 4.3 ffiM Supplies-Direct Sales ACCOWlt Profile, 1986
activities is not accoWlted for separately as order receipt and pro- Purchasing
cessing costs but is lumped in with sales call costs. Locations
Firms Revenuc
In any case, order processing and documentation typically in- --
Cum Number Cum Cum
volve cadres of clerical personnel, which represent a relatively fixed Enterprise Size Number % % $ Mil %
r
(000)
cost. However, the actual cost per order will depend significantly $1 Mil + 15%
{ 29 100
16.4 100 83% 6 100
on the number of items listed, since clerical time spent is likely to $5OOK-IMil of 49 99 9.0 78 of 27 57
$looK-5OOK !inns 220 96 19.0 66 revenue 42 41
vary with the number of lines per order, as is the time spent cor- $ 50-lOOK 195 85 8.7 41 12 17
recting errors. Thus, as a percentage of the value of an individual $ IOK-50K 790 75 16.2 30 14 10
order, paperwork costs will vary greatly because processing costs $ 10K 720 36 6.2 8 4 2
Total 2,003 75.5 $175
are largely a function of the number of items, not dollar amoWlt.
To reduce clerical overhead for order receipt and processing, &nwce: IBM, National Dlstribution Division.
Note: Docs not include sales to IBM authoriud supplies dealcn and wholcsalcrs.
some companies have installed computerized links with their dis-
tributors and large user-customers. A manager in one such firm
commented, mode of order entry to sales expenses, revenues, and product line
You [the authors1 also asked for infonnation on the relative margins. .
costs of handling an electronic order versus a mail or phone ordeL Supplies customers were of three types. First, large corpora-
It's difficult to quantify precisely, but we know that, internally, it is tions, which ffiM supplies salespersons called upon directly, or-
3-4 times easier to handle an electronic order; and that difference dered either from these representatives or by telephone from a cat-
will imptove as we refine our systems. The other major benefit to alog. ffiM also made available to large customers a computerized
us, is that we receive a clean, complete order: the chance for errors, ordering system; although this was just beginning to come into use.
returns, etc., is greatly reduced. From the distributor's viewpoint,
Second, user accoWlts not large enough to warrant direct sales cov-
the greatest savings should be through a reduction in inventory car-
erage had the catalog/telephone ordering option. Third, ffiMsup-
rying costs due to the effective reduction of "lead time." Also, the
efficiencies of receiving the correct product, on time, because the plies distributors ordered by mail and telephone. They could also
manufacturer received a clear order. use an IBM computerized ordering system, but many preferred to
SIIles IkpenseAtuUysis: IBM Supplies Distribution. The use their own computers for generating orders from all their sup-
factors affecting personal selling, order receipt, and processing costs pliers.
are well illustrated in a case history that describes a study made at Although the following data, based on Duffy's study, disguise
ffiM's National Distribution Division (NOD) of the costs telated dollar amoWlts, the relationships among sales volume by size of
to selling ffiM supplies. This product line included supply items customer and by size of order are representative of the actual find-
used with computers, copiers, and word processors. The nature of ings. Exhibit 4.3 shows that 15% of all direct user accoWlts, those
the product line and of customer ordering behavior made order firms over $100,000 in size, produced approximately 83% of sales
processing a substantial sales expense item for NDD. revenue from this class of customer. An analysis of revenue per or-
In 1986, concerned that sales expense-to-revenue ratios had der from direct aCCOWlts (Exhibit 4.4) indicated that 20% of the
risen from 13% in 1968-1972 to 24% in 1986 for the ffiM sup- number of direct orders (120 of 590 orders) accoWlted for 82% of
plies line, William Duffy, manager of ProdUct Planning and Chan- sales revenue.
nel Development at NOD, initiated a study that sought to deter- Duffy also arrayed the number of orders received, sales reve-
mine the relationship of (1) customer size, (2) order size, and (3) nues, and ElR ratios for the three customer categories as shown in
Exhibit 4.5. Exhibit 4.5 data show ElR ratios for orders received
from reseller accoWlts of 10%, from direct accoWlts through IBM
field salespersons of 14%, and through telephone orders, 40%. The
70 DESIGNING CHANNELS SYSIEMS Distribution Costs 71
Exhibit 4.4 IBM Supplies-Revenue pee Order from Direct Exhibit 4.5 IBM Supplies
Accounts, 1986 1986 Revenue and Expense Analysis
Orders R<vcnue
Number of IBM ExpeQ!e
Revenue Number Cum Cum Orders Revenue ($76 M;
Order Size per Order (000) '16 (S Mil) '16 (600 K) ($325 M) 23% EJR)
S500+
$250-499
S2,785
340
20'16 of
number
15565 100
91
82'16 of
r<Ven\lCS
(163
25
100
29 30 K (5%)
-
$14 M
SI50-249 195 ofordcrs 90 80 18 18 "C "C " (14% EJR)
SI00-149 120 80 65 9 10 $100 M
S 50- 99 72 140 51 9 6 (30%)
/
.-
$ -49 32 160 27 ~ 2 Large Direct
User Accounts
// $3OM
Total 590 S230
'k < '" ...I.. . . (40% EJR)
s-r.: IBM, National Distribution Division.
N...: Docs not include ..... to IBM authorized supplies d<alcn and whoIaalen.
.
lower EIR ratios for direct sales and fur sales to large dealers and Other User .
wholesalers could be attributed to the large individual order size Accounts
that typified these two categories. Together these two classcs of $22M
customers generated 7% ofthe total number ofsupplies orders and '" '> ....... I (40% EJR)
60% of IBM supplies sales revenues. The much higher EIR ratio /
II
'"
fur telephone orders is attributable to the small order size: an av- IIII -," ' ,
erage of about $235 pee telephone order, with over 70% of these II $95M ,
orders fur less than $150 each. .. .... " "II (29%) " $10M
Duffy estimated that IBM operators had to complete an aver-
age of two telephone calls per order and to ptepare invoices fur
Reseller
Accounts
I 10K (2%) I____ , I
-
10% EJR)
Stocking, Shipping, and Inventory Risk this risk by buying back obsolete srock or crediting the distributor's
Inventories and accounts receivable generate ongoing financ- account, but the cost must be absorbed at one level or the other.
ing costs for both producers and rcse11ers and can also represent Physical handling costs, as a percentage of sales, are a funetion
significant levels of risk. Our 1986 surveys, for example, indicated of weight-to-value ratios and of distances shipped to stocking l0-
average inventory levels at the distributor and producer levels rang- cations and to customers from the supplier's plant. Oil country tu-
ing between 17% and 22% of annual sales, as follows: bular goods have a high weight-to-value ratio, and physical han-
dling costs are high relative to total selling costs. Disk drives, in
Supplim ~
comparison, have a much lower weight-to-value ratio.
Manufacturers and distributors economize on physical han-
Air comprasors 22% 18%
Load = and circuit breakers 17.5% 22% dling costs in several ways. First, they typically offer lower prices
Diskdrivcs N/A 10.1% for carload and truckload quantities than for lesser amounts. An-
Oil COURtty tubular goods 20.5% N/A other strategy is to srock fast-moving items at numerous field loca-
tions but to srock slow-turnover items at regional centers. In
Inventory costs and risks are associated with the financing of achieving inventory economies through such networking, the man-
stocks, pilferage and obsolescence, and inventory devaluations. The ufacturer must ha1ance higher shipping and storage costs that en-
risks of carrying inventory are related as well to market price and sure rapid delivery for fast-moving items against lower storage costs
supply-demand fluctuations. If product prices decline, the value of and longer delivery times.
goods in srock will decline and may result sooner or later in write-
offs. The inventory devaluations of oil country tobular goods in ReseUer Margins
1980-1985, causing severe financiallosscs for both producers and Not generally accounted for in the producer's revenue stream,
oil field supply houses, are a dramatic example of inventory risk. resale margins comprise a substantial transaction cost component
The level of inventory risk varies with the inventory-to-sales in end-user prices, often 20% or more. This rcsale margin covers
ratio. Generally speaking, the broader the product line relative to the distributor's costs ofdoing business and yields a net profit. The
total sales volume, the higher the ratio of inventories to sales. If distributor's margin is the difference between what it pays its sup-
goods are produced to otder, this principle docs not hold. Never- pliers and what it receives from its customers. Thus, a distributor's
theless, product line simplifications have the effect of lowering margin depends largely on the supplier's price schedule, which usu-
inventory-to-sales ratios. For example, when Siemens-IT'E came ally stipulates different prices for different purchase volumes. In ad-
out with a load center design that could be mounted either on the dition, some suppliers may offer year-end rebates based on annual
wall surface or between studs with the door flush to the wall, it volumes. The- producer's price schedule· typically provides the
enabled its distributors to cut in half the had center srock levels greatest distributor margins on goods purchased in 1arge quantities
needed to fill orders from off the shelf. Previously, IT'E and other and sold to user-eustomers in small quantities. Exhibit 4.6, for ex-
load center suppliers made different models for these two rypcs of ample, is a schedule for grinding wheels that stipulates both func-
installations. This gave IT'E an advantage, short-teno, in building tional (or class-of-trade) diseounts and quantiry discounts. The dif-
its distribution network and gaining market share. In its product ference between the multipliers in the right-hand column and the
redesign, IT'E had reduced inventory risks for distributors by re- one next to it is the funetional diseount.
ducing the amount of srock they had to carry to support a given Such a schedule indicates the percentage factors applied to the
level of sales. list price of the item to determine, by quantity brackets, the net
Another inventory-related risk is product obsolescence, a high prices at which the manufacturer will sell to distributors and to
risk for products with rapidly changing technologies or specifica- user-customers. It suggests that the distributor sell to user-
tions. Suppliers may protect their distributors, fully or in part, from customers at these same prices. However, because of local compe-
74 DESIGNING CHANNELS SYSTEMS Distribution Costs 7S
Exhibit 4.6 User-Customer and Distributor Multipliers Exhibit 4.7 IBM Supplies
Il<IIl Quantity Number of Uscr-Customer Disrriburor Average Discounts, FJR Ratios, and Contribution by Channel
Group Units Multipliers Multipliers
Large Other
FA 1-9 1.027 0.822
A 10-19 0.689 0.551 Direct User
B 20-49 0.585 0.476 Resellers Accounts Accounts
C
D
E
F
50-99
100-249
250-749
750-1,249
0.503
0.447
0.394
0.379
Q.418
0.393
0.355
0.341
Average
Discount 35% 20%
J 10% I
/ /
G 1,250- 0.365 0.328
! ///
/;G'~/
Stnme: Norton Company.
Note: Discounts in this table an: exprosscd" percattages of a liar price. nor shown hct<. 40'/.
10% ,,""/
/ ' ..-//
27%
---
---
-
charge and not to their suppliers' suggested list prices. ..-
While distributor margins are not directly controlled by pro-
ducers, they compensate for distribution costs that the producer
would incur if it elected to perform the services provided by resell·
ers. A relevant consideration in developing modes of distribution, Contribution 35% 40% 45%
Percentage
then, is whether lower total distribution costs would be incurred
by selling through distributors or by going direct.
COST-REVENUE RELATIONSHIPS
Sales expense-to-revenue ratios are measures of CIJSt e.ffidencJ,
with low ratios representing good performance. Cost· ejfeaWentss
measures, on the other hand, also take account of revenues and
profit margins as reflected in market share and produet line profits.
Further, it is unlikely in most cases that maximum revenues and
profits will be realized at the lowest possible FiR ratios. When WI1-
liam Dutfy, for example, wenr on to compare the FiR ratios by
customer order entry mode with the revenues and margins for each
channel, it became apparent that high FiR percentages were actu-
ally associated with high sales margins and vice versa, as Exhibit 4.7
indicates. Thus, while the average FJR ratio for sales to IBM whole-
salers and authorized dealers was the lowest, discounts from list
prices for this market segment were the highest. The net result was
a lower contribution from sales to resellers than to the other two
76 DESIGNING CHANNELS SYSTEMS Distribution Costs 77
account categories. At the other extreme, sales to user accounts of distribution, (3) the number and type of reseUer firms in a dis-
through the catalog/telephone channel had the highest ElR ratio tribution network, and (4) the relative desirability ofthe producer's
(40%) but also generated the greatest contribution (45%) because granting territorial exclusivity and obtaining brand exclusivity.
higher selling costs were more than offset by higher price realiza- The Direct-Indirect Billtmee. The cost of reaching indi-
tion. The discussion of Duffy's analysis will be continued in Chap- vidual customers relative to the account value is a prime factor in
ter 5. choosing between direct sales and reseUers. If, on average, a sales-
Generally, market segments for industrial products differ with person representing a disk drive supplier and earning $50,000 in
regard to prevailing price levels, the distribution costs required to salary can generate $10 million a year in sales from five accounts,
reach customers, and hence the amount of revenue potential and the favored option is clearly direct selling. At the other extreme,
contribution. Firms that focus on higher-margin segments, such as consider a sales rep working for an electrical distributor, paid
Square D, which has built a particularly strong position in the res- $35,000 a year in salary and commissions, making an average of 80
idential housing market, may not have the low ElR ratios typical in calls a month and accounting for annual sales of $700,000, of
selling to lower-margin and more concentrated market segments which 10-20% may be load centers and circuit breakers.' It would
such as manufactured housing. But these firms may realize higher be prohibitively expensive for a supplier such as General Electric to
revenues and margins because of the total segment sales potential reach this sales rep's LC&CB customers directly through sales-
and prevailing market price levels. persons carrying only the GE line.
EIR ratios are also likely to be greater in firms with large Number ufDistributUmLwels. The costs of taking prod-
salesforces providing distributor support and calling on large user- ucts to market increase with the number of distribution levels
customers. However, these producers maintain control over key ac- through which they pass, since each intermediary requires some
count relationships and price negotiations and help to ensure that margin to cover the costs of providing distribution services. At one
large users will buy their brand rather than the competing product extreme is single-tier, or direct, selling; at the other is multitier dis-
lines the distributors may also carry. tribution, with the prodUct moving from the factoty through
agents, master distributors, distributors, subdistributors, and retail
THE INFLUENCE OF DISTRIBUTION COST dealers, or some combination thereof. Whether to add or to elimi-
FACTORS ON MODES OF DISTRIBUTION nate distribution levels at different points in the development of
Distribution costs influence both the stroeture of distribution markets is ofren an issue for sales administrators charged with de-
channels and the locus of distribution functions. Cost e1feetiveness veloping and managing distribution channels. For example, Gen-
depends largely on building and maintaining distribution that, on eral Electric's Component Motor Operation (CMO) served the
the one hand, is low cost relative to competitors' networks and, on consumer aftermarket for motors used in major household appli-
the other, provides a strong end-market presence and brand ances by selling through three master distributors that sold to elec-
strength. To achieve effectiveness, producers make choices, first, be- trical wholesalers that in turn sold to appliance repair shops. This
tween direct sales and intermediary channels. This choice repre- arrangement dated back to the early postwar years when the popu-
sents, in effect, a tradeoff between incurring internal selling costs lation of consumer appliances was considerably srnaller than it is
and "paying" resale margins. Then, in any mixed, or multichannel, today. In the meantime, other motor manufacturers, bypassing
system, there are additional choices about the positioning of distri- master distributors, were selling to wholesalers. Even though prices
bution functions at points in the chain. charged by the master distributors to electrical wholesalers had de-
clined to the point of yielding the former only marginal returns,
Distribution StrUcture competing motor manufacturer prices were even lower on high-
Specifically, cost factor considerations dramatically influence volume motors. A key issue confronting CMO marketing managers
(1) the balance of direct-indirect selling, (2) the number of levels
I These data arc based on our 1986 swvcy of 76 electrical distributors.
78 DESIGNING CHANNELS SYsrEMS Distribution Costs 79
as of 1987 was whether to end rheir master distributor relationships sented by a single reseller in a trading area, rhus granting rhe re-
and sell direct to large wholesalers. [This case is considered in seller "territorial exclusivity." A supplier may also seek to develop a
Chapter 10 on coping wirh change.] network of resellers rhat do not carty competing lines and rhus
In rhis instance, as rhe consumer aftermarket for motors ma- require "brand exclusivity" of its channels.
tared and became more price competitive, distribution cost and rhe Territorial exclusivity is common when rhe reseller must make
need to be price competitive became major considerations. By con- large product-specific investments. Especially in rhe early stages of
trast, in rhe early stages of product-market development, when cus- market development, when rhe returns are uncertain, rhese invest-
tomers are often less price sensitive and have strong brand prefer- ments would be jeopardized by rhe reseller's having direct compet-
ences, channels cost inefficiencies may be tolerated wirhout major itors in rhe trading area. For example, in establishing distribution
losses in revenues and market share. for programmable controllers (PCS),' rhe major manufaeturers-
Number tUUl1jpes of Distribution Firms. A significant Allen-Bradley, General Electric, Westinghouse, and Square DS_all
factor affecting distribution costs may be rhe salaries and expenses opted for territorial exclusivity as an inducement to selected distrib-
of salespersons required to service rhe distribution network and utors to make rhe requisite investments. The investments included
monitor its performance. Square D, for example, employed a sales- rhe cost of demonstration equipment and facilities, rhe salary of a
force of more rhan 650 people, who spent over 60% of rheir time technical expert, and a minimum inventory of PC units and parts.
calling on 1,300-plus Square D distributors, each distributor firm Square D managers estimated rhe first year's cost to each franchisee
operating out of 1-3 sales locations. at $150,000. Such an investment was reasonable only ifrhe distrib-
Control Data's Peripheral Products Company (PPCo), on rhe utor did not have to compete wirh orher distributors of rhe same
orher hand, distributed rhrough two large national distributors, brand for whatever levels of demand could be developed in his or
working wirh rheir headquarters and sales branches rhrough a her trading area.
srna11, distributor-specia1ized salesforce. Very little of PPCo's 72- Brand exclusivity, too, is associated wirh heavy investments in
person direct salesforce's time was spent contacting distributors. a product line. Investments in inventory rhat are large relative to
Two major considerations affect rhe choice among selling rhe distributor's total assets, as well as significant investments in
rhrough a few large national firms, a number of regional firms, or technical skills or facilities tailored to a specific brand, may prevent
many small local operations: rhe costs incurred in building and rhe reseller's making comparable investments in a second brand.
maintaining rhe channels network and rhe level of product-specific
investments that distributors must make relative to product line LxwofDffitributionFooroorn
sales revenues. In some instances, rhe absolute dollar amounts of
certain product-specific investments are large enough to preclude Cost factors strongly influence rhe positioning of such func-
rhe use of srna11, local distributors. For example, large investments tions as selling, physical distribution, and risk assumption along rhe
are required for technically trained personnel in rhe disk drive in-
chain of business units rhat form a distribution system. Under rhe
pressures of competition, cost efficiency and cost effectiveness be-
dustry. This may be one reason that disk drive manufacturers have
often used national and regional distributors, rarher rhan locals. To
• Pes arc industtiallogic controls used in both process and discrctc systems to control ma-
sell disk drives and orher computer-related products, it was essential chining tasks. Pes were initially devdopcd in the rnid-l96Os to replace electromechanical
rhat distributors develop and support staffs of technically trained control devices such as relay SWitches. Pes are generally a more versatile control systan and
computer specialists, representing a fixed cost not likely to be sus- also provide memory storage. Thus, users can assess data such as the number of parts that
pass through a machine, how long a machine has bccu operating, the time bctw= tool
tainable by small resellers. changca, and other usefuI operating information. They can aIao change the sequence of 0p-
BrMUl tUUl TerritIJritU &caMPily. Used wirh reference erations by reprogramming.
51\vo other larac competitors were Modicon, which relied on a direct salcsforcc, and. Texas
to distribution systems, rhe term txdusivity has two meanings, ter- Instruments, wIiieh established a netWork of systems houses with software and engineering
ritorial txdusivity and brand txdusivity. A supplier may be repre- sltills fOr high-tech applications.
80 DESIGNING CHANNELS SYSTEMS Distribution Costs 81
come major considerations in who does what in taking products it in stock? When can I get it? How much does it cost? Often, in
from factory to market. that order.
The demand generation function may be assumed largely by Finally, distribution cost considerations influence the locus of
suppliers, assumed largely by resellers, or shared by both. As we order shipment. Orders received by producers and distributors are
discussed in Chapter 2, where the demand-generation function is not always fulfilled from their respective stocks. Orders received by
positioned in the channel depends primarily on customer size and resellers ror large quantities, especially of high weight-to-value
location and the nature of the appropriate mode of communica- products, may be shipped from the producer's inventories to the
tion. The more concentrated the market and the larger the individ- user-customer. These "drop-shipped" items pass, not through the
ual account or potential order, the more likely it is that the sup- distributor's warehouse, but through the distributor's books ror re-
plier's salesforce will take an active selling role at the user-customer cording the sale, billing, and collection. For example, McKesson
level. Thus, according to a sales manager in an electrical manufac- Chemical, the largest distributor of chemicals in the United States,
turing company, "The larger the customer, the more 'out-front' fills 80% of its orders from stock, while another 20% is shipped
selling we do to help our distributors get the order." direetly to McKesson's customers from its suppliers' stock.
At the same time, suppliers are loath to spend $200-300 per GESCO, the third largest distributor of electrical eqnipment and a
sales call if their distributors can be as effective at $80 per call. That captive distribution arm of the General Electric Company, fills from
depends on the sales task. The initial sale of a new product to an stock about 95% of its orders from the residential construction
existing customer or of existing products to a new account may call market but ouly 50% of orders received from nonresidential con-
for multiple messages, high in product application content, to the tractors. The remainder is shipped from GE product departments
several members of the buying decision-making unit. In contrast, and GESCO's outside suppliers. The difference between the two
repeat purchases by satisfied users may be generated through short segments is the size of orders placed and the type of product pur-
and simple interchanges of infortnation having to do largely with chased. Nonresidential contractors buy and install large pieces of
price and delivery, as calculated from our 1986 surveys. electrical equipment not used in residential construction. 6
Compare, ror example, the average costs per sales call made by Alternatively, orders taken by producers may be delivered from
producer and distributor sales representatives in three product- local distributor stocks when immediate availability and delivery in
market groups. small quantities are important to the customer.
Customer communication costs vary significantly among the several Physical distribution functions also tend to be carried out at that
modes of communication and even within modes. To a 1arge extent, the point in the chain where they can be performed at least cost, commensur-
form of communication should depend on the infurmation needs of the ate with customer, satisfaction. As nored above, product weight-to-value
customer and on where in the channels the particular mode of communi- ratios affect this choice, with large, bullry itemS going direct to uscrs from
cation may be performed in a cost-dlicient way. However, ifsales revenues the producer's factory. The need of customers to have local stock avai1a-
and margins are considered satisfactory, suppliers may clect to retain con- bility and to order intermittendy in small amounts will also determine
trol over selling functions even though it may mean higher costs. shipping strategies. For example, orders negotiated direcdy by producers
Processing costs per order vary with the number of line itemS or- may be filled by srocks held in distributor warehouses.
dered and not with the dollar amount of each order. Thus, the leverage Responsibility for financing sales credit and carrying inventories to
points in controlling order-processing costs lIS IJ perrmtRge uf silks are es- serve end-customers may be taken by producers or rescllers, depending on
tablishing minimum order points and providing customer incentives to the extent of risk. Traditionally, suppliers going through exterual inter-
reduce order frequency and increase individual order amounts. There is mediaries have perceived the costs and risks of eatrying inventories and
often also significant cost-saving potential in computerized order-entry collecting receivables to be in the rescller's domain. However, in some
systems. cases (e.g., oil country tubular goods) such risks may not be sustainable at
Inventory eatrying costs, typically a large cost component, depend the distributor leVel and some suppliers have stepped in to assume this
primarily on the amount ofinventory needed for a given level ofsales, that burden. In addition, suppliers often finance distributor inventories
is, the inventory-to-sales ratio, and on the tisks of inventory devaluation through providing extended credit, or "dating plans," offeting up to six-
due to technical obsolescence or price deretioration. Freight and storage month terms on receivables. Dating plans not only recognize the financial
COSts vary with product weight-to-value ratios. Ways of optimizing physi- limits of distributor businesses but also appeal to suppliers as a way of
cal distribution costs include the netWorking of field srocks, bypassing competing for sales by loading up distributor shelves.
distributor warehouses with drop shipments direct from producer to user, Distribution cost factors also influence whether suppliers give resell-
and providing price incentives for carload and truckload quantities. (Phys- ers and agents territorial exclUSivity. Simply stated, if the investments re-
ical distribution cost optimization may be thought of as balancing COSt sellers have to make in order to stock and sell a particular supplier's prod-
and speed of delivery to satisfy customer requirements.) uct line are large relative to the sales potential in a trading area, then the
How do cost factors affect the design of distribution systemS? Dis- supplier is prone to franchise only one agent or distributor in the area to
tribution functions tend to be performed at different levels in the system, enhance the attractiveness of the investment. As for brand exclusivity, dis-
depending on whether they can be supporred by avai1able sales revenues tributors may be more willing to carry only one single brand if heavy
and margins. Thus, distribution expense-to-revenue ratios lead suppliers investments and costs must be incurred simply to carry the chosen line
to move often from produet-specializc:d-or customer-specialized-direct and if the returns on the investments in additional lines promise only mar-
sclIing to full-line direct sclIing to rescllers as they go from highly concen- ginal returns.
trated, large volume markets to relatively thin ones. A supplier's relative The need for certain high product-specific investments at the reseller
reliance on selling direct versus going through rescllers may be explained level also limits the number of qualified distributors. Under this condition,
largely by (1) the different EJR ratios associared with reaching customers suppliers will tend to build their distribution around the larger regional
of widely varying purchase potentials in markets that range great1y in den- and natioual distributor chains that have sales volume bases large enough
siry and (2) differences in the nature and cost of sales ca1Is made by direct to support investments in specialized sales resources.
salespersons and cescller sales reps, respectively.
All these cost factors influence the number of levcls in the distribu-
tion chain and the positioning of such functions as sclIing and physical
distribution. Even in distribution systems that rely heavily on rescllers,
suppliers may play very active roles in demand generation with large in-
dividual user accounts. Distributors may then perform an account-
servicing role at a significandy lower cost per call than would be incurred
by the supplier's salesperson.
Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.
Alternative Proxies: