Table of Contents
Go Back
Information Filtering;
A Service by Business Intermediaries
J. Parvizian! and M. Nosouhi?
1. Abstract
Business intermediaries are often blamed for not adding any value to the product.
Therefore, it is always recommended to make direct business connections between
producers and consumers. E-business made this connection more possible and
realistic than ever in a large scale. The core assumption behind this value analysis is
that the intermediaries’ role is limited to the exchange of products and money. The
present work recognizes the information flow through intermediary channels. This
information, that is used in business for market analysis and forecast, advertisement
and so on, like any other information is mixed always with noises, is produced in a
format that may not be suitable for end-users, and reports facts with a delay that may
be too short for decision makers to judge upon or too long to be useful at all. A sharp
increase in demand in a very short time can be misleading for the producer to increase
production capacities. The intermediary inventory can absorb this increase of demand
if it does not survive for long; otherwise will pass it to the producer. Intermediaries, or
any institution with similar effects, can filter the information to eliminate noises, to
present it in a proper format, and to deliver it in critical time steps. This work may
provide a new justification for the collapse of e-based enterprises after a rapid growth
in late 1990s.
2. Introduction
The collapse of electronic companies (.coms) following the rapid boost of late 90s
raised many questions about e-solutions adopted in business. This research is focusing
on one of possible drawbacks of many solutions that are based on the direct purchase
of goods from producers/distributors by consumers. The main idea is to identify the
role of information filtering that is carried out by intermediaries in every business.
Intermediaries, such as distributors and retailers, are business agents that control
customers-producers relations. Emphasizing on the fact that they add no value to the
product, classical views suggest that eliminating them from the business chain can
reduce the price of products and speed up the relationship between the customers and
producers. Rapid development of information technology, easy access to the internet
and to the statistics collected by inteet questionnaires can further support the
elimination the intermediaries. Therefore, producers have access to the information
they need immediately using advanced IT facilities.
' Assistant Professor, Department of Industrial Engineering, Isfahan University of Technology, Isfahan,
84156, Iran. Tel: +311 391 5514, Email: japa@cc.iut.ac.ir
? Isfahan Sience and Technology Town, Isfahan, 84156, Iran.
Wigand and Benjamin [1] describe how the retail price in the high quality shirt market
would be reduced by almost 62% if distributors and retailers were eliminated.
Gellman [2], Gates [3] argue that as “friction-free” electronic marketplace lower the
cost of market transaction, it will become easy to match directly buyers and sellers,
and as a result, the role of intermediaries may be reduced, or even eliminated, leading
to “disintermediation”.
For Lewis [4], given that the intemet encourages direct and immediate contact
between suppliers and end users, together with a simultaneous drop in transaction
costs, there is a strong case for internet-driven ‘disintermediation’- or the elimination
of the intermediary entirely [9].
Bakos [5] summarizes main functions of the market as to matching buyers and sellers,
facilitating the exchange of information, goods, services and payments associated
with market transactions, and providing an institutional infrastructure that enables the
efficient functioning of the market. Intemmet-based electronic market place leverage
information technology to match buyers and sellers with increased effectiveness and
lower transaction costs, leading to more efficient ‘friction-free’ market.
Schmitz [6] identifies three services provided by intermediaries as to hold inventory
to provide (the service of immediacy and) insurance against systematic valuation
tisks, to reduce asymmetric information by establishing a reputation, and to gather,
organize and evaluate information that is dispersed in the society. It is shown that
these three services are nor under threat by the diffusion of electronic commerce.
The emergence of cybermediaries as net-based intermediaries is noted by Sarkar et al.
[7] and Bichler and Segev [8]. Cybermediaries are organizations that operate in
electronic market to facilitate exchange between producers and consumers by meeting
the needs of both producers and consumers. They also increase the efficiency of
electronic markets in a role similar to intermediaries by aggregating transactions to
create economies of scale and scope.
Vandermerwe [9] argue that ‘electronic go-between service provider’ goes one step
further; it recognizes and uses the power and potential of advanced interactive
technologies in order to link individuals who want products and services with those
who can provide them. Then, in this new middle role, it integrates and delivers these
offerings to customers as once superior experience. Cybermediaries, as Caillaud &
Jullien [10] noted, specialize on the pure informational aspects of intermediation, the
physical part being left to sellers’ distribution system.
To summarize, if only transaction costs are considered both consumers and producers
benefit from disintermediation. On contrary, if intermediaries play a role in the
information flow by aggregating and disseminating data to customers, then, in an
electronic market, this service can be provided by cybermediaries [11].
Based on a simple dynamics model, suggested by Forrester [12], to analyze
distribution-production system, this paper investigates how intermediaries can filter
the aggregated information and disseminate them to the third partner. It is suggested
that even if intermediaries are replaced by cybermediaries in the e-market, the
filtering role must be somehow be fulfilled.
When the supply chain is not properly managed even a small fluctuation in demand
can cause large variations in the upper level of the supply chain. This is addressed to
in the literature by the bullwhip effect [13]. In this paper it is also shown that the
filtering role of intermediaries in the supply chain can reduce this effect.
3. Business Model
In the production-distribution model proposed in [12], distributors and retailers are
intermediaries between producer and consumer. They have their own warehouses,
processing delays, shipping etc. This model is very close to the real world even if
delays can be reduced and backlogs can be decreased.
Following the eras of volume and quality, e-biz was now in the scene to provide a fast
way to choose, to order, and to pay. For a short time, all statistics showed that
consumers felt happier to choose in a wider range of products for better prices
spending less time. Producers were also well satisfied of knowing what their
consumers wanted. They could produce, just in time, what there was an order for. The
money transfer was also much improved. It seemed that with only slight modification
to the old models, the emerging version of the business would be understandable. All
models had to be modified only for shorter time delays.
However, a major “why” raised after the recent collapse of dotcoms. Was it just a
quick light in the dark or just the beginning of the end? This research is not to answer
this big question. Rather it tries to illuminate some other aspects of a business that
were never pointed out because they were always present. Aspects like the personal
feeling about the merchandise or service, the person to person contacts involved in
every business, and the information filtering. The concept of filtering signals comes
from electro/mechanical systems in which low-pass or high-pass filters are sometimes
important elements of a system. A low-pass filter, for example, pass only signals with
low frequencies.
The production-distribution model of [12] is examined here assuming that delays are
reduced considerably to make it closer to the e-biz world. For example, the delay in
mail from distributor can be reduced from half a week to half a day or less and so
other delays. Figure 1 shows the schematic of a system from producer to the customer
including all delays involved. To adapt the model all delays are considerably
decreased to examine the effects of deploying an internet-based business model on the
production.
Shipping ;
delay Distributors
Factory warehouse
Inventory
Shipping
delay
Ordering decision
and delay ?
-" Retailers
“CO
Orders
from
customers
Delivery of goods to
customers
Figure 1: Schematic mode!
4. Filtered Information
The role of filtering was present in the main model when the fluctuations in demand
were absorbed by the warehouse of the distributor, or intermediaries. The role of
intermediaries, therefore, is not just transfer of goods and money between the
producers and consumers. Since there is always a level in the intermediary section to
accumulate the orders of the costumers, this level can also play the role of information
filtering by just reflecting only the long-term changes in demand. To model this
filtering, delay function of the intermediary level is replaced, in the modified model,
by an averaging operator. Obviously, any averaging operator is playing the role of a
low-pass filter since the output is insensitive to the fluctuations over small periods of
the input. This element can well play the role of a filter that only passes the long-term
changes of demand and is insensitive to short term periodic or random signals with
high frequencies.
To examine the model, let us first consider the case when there is a medium-term
sinusoidal change in demand. Figure 2 shows the factory production output of the
normal business model for which delays are reduced to match the e-business
solutions, Table 1. In figure 3 this model is examined when delays are accordingly
changed to model the elimination of the intermediary or distribution section. As
shown in this figure, the bullwhip effect is reduced and the maximum amplitude of
the factory output is decreased from 1680 to 1310 units.
Tablel Delays in the original model and modified model.
Delays Original model (weeks) | Modified model (days)
Handling time at Distributor 1 0.001
Due to Unfilled orders at Distributor 06 0.001
Taventory (and pipeline) adjustment at Distributor 4 5000
Clerical at Distributor 2 0.00
Mailing from Distributor 05 0.00
Transportation to Distributor 2 0.00
Factory production output
rest cone enn epee
12266:
‘L00 3075 60.50 ‘9025 120.00
Weeks
Figure 2 Original model with sinusoidal changes in demand;
Input = (100*sin (2*pi*time/52))
Factory production output
aaa?
1083.08
Figure 3 Delays in distribution section are almost zero.
i, ile mara
"100 3075 60.50
Weeks
Figure 4 Delays in the distribution section are very short or zero, but the distributor filters the
information of demand change with PSR' = SMTH*1(PDR°*,4), PSD‘ = SMTH1 (PDD*,).
Now let us our distributor plays his role of filtering still with very short delays to
enjoy the benefits of electronic business. As shown in figure 4 the overshoot has not
changed considerably than the previous case. This is because the sinusoidal change
happened in a very low frequency over one year. To observe the effect of filtering, the
input can be replaced by (100*sin(2*pi*time)). Then the out is as shown in figure 5,
before filtering, and as in figure 6 after filtering. It can be seen that the domain of
variation of factory output is very narrow (curve 1 in figure 5) and a simple
smoothing can even make its band narrower (curve 2). However, with filtering, as
explained earlier, the output converges to the gain of the input that is 1000 and the
factory is ordered to produce enough to respond the average variation of the demand.
It is notable however here that the overshoot is not changing considerably as
expected.
if mm
VA i
Sees
2 1030.00]
roso.0a}
1025.00]
|
900.00] H
1000.00
‘L900 30.75 60.50 90.25 3204
Figure 5 High frequent sinusoidal changes of demand cause variation of the factory output. Delays in
the distribution section are zero and the input function is (100*sin(2*pi*time)).
. Purchase order Sent from Retail
. Smooth function
. Purchasing rate Decision at Retail
- Purchase orders sent from Distributor
. Purchasing rate Decision at Distributor
1; Factory output 2: Smoothed factory output
Pe
1,00 30.75 60.50 90.25 120.00
Figure 6 Filtered information of the demand change prevents rapid fluctuations of the output.
The effects of filtering on the output in response to random input are shown in figures
7-9, Similarly the model can be examined for other inputs such as step or impulse. It
can be seen that the filtered information in the model with zero delays produce a
similar overshoot at the beginning but in later times filtering produce a more evenly
output.
1300
oa
1
Figure 7 Factory production output; smoothed.
Original model with random changes of demand; the random function is: (normal (0,100,175)).
usLay-~-
Units
1022.3
ot
Figure 8 Factory production output; smoothed.
Delays in the distribution section are zero.
Factory production outpt-smoothed
angasy ===>
Units
101360
oa.
1.00
Figure 9 Zero delays with filtered information produce a similar overshoot but in the long-term more
smoothed variations in the output.
5. Conclusions
The main aim of this paper is to appreciate one of the important services of the
business intermediaries. It suggests that intermediaries can play more important roles
than exchange of goods and money. They establish and control information channels
between parties involved in a business. Information is not restricted to the demand
variations; it can be extended to the variation in tastes of customers, information about
competitors and so on. Filtered information can add value to the product by
preventing rushed decisions to invest, or to supply. It can eliminate noises that are not
inherent to the market trend. It can also increase the trust of the consumers by
delivering to them the information that will not be proven to be hasty or misleading.
Filtration can be applied to all information that is exchanged in the value chain. Thus,
even in an electronic based business model in which there is no justification to
consider delays introduced by the intermediaries, there is still enough room for
cybermediaries to play the role of information exchange in the filtered form.
References:
[1] Wigand, R. T., & Benjamin, R. I., “Electronic Markets and Virtual Value Chains on the
Information Superhighway”, Sloan Management Review, Vol. 36, No. 2, pp. 62-72, 1995.
[2] Gellman, R., “Disintermediation and Intemet”, Government Information Quarterly, Vol.
13, No. 1, pp. 1-8, 1996.
[3] Gates, W., “The Road Ahead”, Penguin Books, New Y ork, NY, 1995.
[4] Lewis, T., “The Friction-Free Economy: Marketing Sterategies for a Wired Word”,
Harper Business, New Y ork, 1997.
[5] Bakos, Y., “The Emerging Role of Electronic Marketplace on the Internet”,
Commiunication of the ACM, Vol. 41, No. 8, pp. 35-42, 1998.
[6] Schmitz, S., “The Effects of Electeronic Commerce on the Structure of Intermediation”,
CMIT Working Paper 98-W P-1031, 1999.
[7] Sarkar, B., Butler, B., Steinfield, C., “Cybermediaries in Electronic Marketspace: Toward
Theory Building”, J ournal of Business Research, Vol. 41, pp. 215-221, 1998.
[8] Bichler, B., Segev, A., “A brokerage framework for Intemet commerce”, CMIT Working
Paper 98-WP-1031, 1998.
[9] Vandermerwe, S., “The Electronic ‘Go between Service Provider’: A New ‘Middle’ Role
Taking Centre Stage”, European Management Journal, Vol. 17 No. 6, pp. 598-608, 1999.
[10] Caillaud, B., Jullien, B., “Competing Cybermediaries”, European Economic Review,
Vol. 45, pp. 797-808, 2001.
[11] Parvizian, J. and Nosouhi, M., “Information Filtering by Business Intermediaries: A
System Dynamics Approach”, Proceedings of the International Management Conference,
Sharif University of Technology, Tehran, December 2003.
[12] Forrester, J. W., “Industrial Dynamics”, ninth printing, M.LT. press, New Y ork, 1977.
[13] Stadtler H., Kilgel, C., “Supply Chain Analysis and Advanced Planning, Concepts
models Software and Case Studies”, Berlin, New Y ork, 2000.
Back to the
CC BY-NC-SA 4.0