464
A Design Tool for Business Process Design and Representation
The abstract class GenericBusinessProces-
sObject is the generalization of the last four
concrete classes in the bullet item (SwimLane,
FlowObject, Artifact and ConncetingObject).
In the GenericBusinessProcessObject class we
GH¿QHWKHGDWDW\SHSURSHUW\VKDUHGE\WKHVHIRXU
classes. The datatype property are:
• Categories.
,Q%301VSHFL¿FDWLRQLWKDV
documentation purpose; in the metamodel
LWLVDGDWDW\SHSURSHUW\RIW\SH³WH[W´
•Documentation.
As categories, it is a
GDWDW\SHSURSHUW\RIW\SH³WH[W´
•Name.
It is a text data type property that
DOORZV XV WR GH¿QHD XQLTXHQDPH LQ WKH
business process diagram for each Generic
Business Process Object.
3URSHUWLHV GH¿QHG LQ WKH $EVWUDFW FODVVHV
cannot contain instances but, thanks to the
class-subclass the relationship property will be
inherited by subclasses until the subclasses will
be concrete.
At this point the main concepts of BPMN are
represented as ontological classes. In order to link
WRJHWKHUWKHPDLQFRQFHSWVZHGH¿QHWKH2EMHFW
Property in the proper classes.
The use of Object Property in the BPMN

ontology is a little different from the traditional
Semantic Web. An example is useful in under-
standing this interesting aspect. Each process may
be composed by different GenericBusinessPro-
FHVV2EMHFWDQGLWLVQRWDPXVWWRGH¿QHLQHDFK
process all the GenericBusinessProcessObject
GH¿QHGLQWKH%301VSHFL¿FDWLRQ,IHDFK*H-
QHULF%XVLQHVV3URFHVV2EMHFW LV GH¿QHG RQO\ E\
LWVQDPHDVROXWLRQPD\EHWRGH¿QHLQWKHFODVV
Process several properties (datatype properties)
each of one of the generic business process. The
generic business process is a more complicated
concept: It has several subclasses and each of them
has its own properties. To solve this problem in
the metamodel that we developed, we adopt an
2EMHFW 3URSHUW\ ³KDV*HQHULF%XVLQHVV3URFHV-
sObject,” which has the class Process as domain
and the class GenericBusinessProcessObject as
range. The OWL code is in Figure 5.
,QWKLVZD\LWLVSRVVLEOHZKHQGH¿QLQJWKH
PRGHO VWDUWLQJ IURP WKH PHWDPRGHO WR GH¿QH
VHYHUDOLQVWDQFHV RIWKHSURSHUW\ ³KDV*HQHULF-
%XVLQHVV3URFHVV2EMHFW´ HDFK RI WKHP GH¿QH
DVSHFL¿FEXVLQHVVSURFHVVREMHFWZLWKLWVRZQ
SURSHUWLHV6WDUWLQJIURPWKLVH[DPSOHZHGH¿QH
LQWKHVDPHZD\WKHSURSHUW\³KDV6ZLPODQH´
This property has the ontological class Business-
ProcessDiagram as domain and the ontological
FODVV6ZLPODQHDVUDQJH)LQDOO\ZHGH¿QHWKH
SURSHU W \³LV0DGH2I /DQH´WRVWDWHWKDWHDFK3RRO

(class Pool is the domain of this property) may
contain one or more Lane (range of property).
Starting from previous consideration the core
classes and main relationship of the ontology
metamodel are represented in Figure 6.
Some special cases have been faced dur-
ing the development of the BPMN ontology
metamodel.
$3RROIROORZLQJ%301VSHFL¿FDWLRQPD\
contain both a Lane and a GenericBusinessPro-
Figure 5. OWL code of has Generic Business Process Diagram Graphical Object
465
A Design Tool for Business Process Design and Representation
cessObject (different from Swimlane) (Figure7).
The problem is how to model this concept: make
two different Object Properties or the same? The
best solution, following the ontology idea, is to
provide the same Object Property because the
semantics of the relationship are the same. We
GH¿QHWKH2EMHFW3URSHUW\³EHORQJV7R3RRO´ZLWK
only one class as range (the ontological class Pool)
and the domain as the union of the other classes:
Flow Object, Artifact, and Connecting Object.
In this way the same property, depending on the
c o n t e x t , i s u s e d t o e x p r e s s b o t h t h e f a c t t h a t a L a n e
belongs to Pool and to lay Flow Object, Artifact,
DQG&RQQHFWLQJ2EMHFWWRWKHVSHFL¿F3RRO
,QWKHVDPHZD\WKH2EMHFW3URSHUW\³EH-
longsToLane” (Figure 8) is used both to model
the fact that one Lane can contain another Lane

DQGWRGH¿QHZKLFK)ORZ2EMHFWDQGRU$UWLIDFW
DUHGH¿QHGLQWKHVDPH/DQH
Additional Classes
In order to cover all the BPMN complexity,
during the BPMN metamodel development, we
GH¿QHFRQFHSWVPRGHOHGDVRQWRORJLFDOFODVVHV
QRWFOHDUO\GH¿QHGLQWKH%301VSHFL¿FDWLRQ
As an example we consider the class Trigger. We
observe that a trigger is the mechanism that allows
Figure 6. Core classes and relationship
GenericBusinessProcessObject
Swimlane
hasSwimlanes
Swimlane
FlowObject
Artifact
ConnectingObject
hasGenericBusinessProcessDiagramGraphicalObjects
Pool
isMadeOfLane
Lane
GH¿QH3URFHVVHV
Processes
AbstractClass
Class
references Generalization
AbstractProcess
PrivateProcess
CollaborationProcess
Figure 7. Ontological property belongsToLane and belongsToPool

GenericBusinessProcessObject
Swimlane
FlowObject
Artifact
ConnectingObject
Lane
Pool
belongsToPool
belongsToLane
isMadeOfLane
466
A Design Tool for Business Process Design and Representation
an event to start, so a trigger is what allows the
HYHQWVWRVWDUW%301VSHFL¿FDWLRQGH¿QHVVRPH
properties for a trigger; for example, if a trigger is
of type Timer, it has the property timeCycle that
GH¿QHVWKHGXUDWLRQRIWKHHYHQWDQGWLPH'DWH
WKDW GH¿QH ZKHQ WKH HYHQW VWDUWV :H OLQN WKH
Ontological class Trigger with the Event by the
SURSHUW\³KDV7ULJJHU´7KHFODVV7ULJJHULVPDGH
up of several subclasses each of them, following
%301 VSHFL¿FDWLRQH[SUHVVLQJDVSHFLDOW\SH
of trigger (Figure 9).
)LQDOO\WRGH¿QHDOOWKH%301SURSHUWLHVRI
HDFK%301SULPLWLYHVZHGH¿QHZKHUHDSSUR-
priate, ontological properties to meet the BPMN
VSHFL¿FDWLRQ
From the BPMN Metamodel to the
Ontological Business Process Model
6WDUWLQJIURPWKHPHWDPRGHOSUHYLRXVO\GH¿QHG

LWLVSRVVLEOHWRGH¿QHDEXVLQHVVSURFHVVPRGHO
G H ¿ Q L Q J L Q V W D Q F H V RI R QW RORJ L F D O F O D V V H V D Q G S U R S -
Figure 8. OWL code of the “belongsToPool” properties
Figure 9. Ontological class trigger
Events
Events
Events
Cancel
Compensation
Error
Link
Message
Multiple
Rule
Terminate
Timer
467
A Design Tool for Business Process Design and Representation
erties (we talk about concrete ontological classes).
6XSSRVHWKDWZHZDQWWRGH¿QHDVLPSOHEXVLQHVV
process made up of one Pool and of a star event,
one task and the end event. The task is linked to
WKHVWDUWDQGHQGHYHQWZLWKVHTXHQFHÀRZ
:H GH¿QH DQ LQVWDQFH RI WKH FODVV 3URFHVV
DQG ZH GH¿QH WKH LQVWDQFHV RIDOO RIVRPHRI
WKHPSURSHUWLHVGH¿QHGE\%301VSHFL¿FDWLRQ
and in the metamodel. Following the property
³LV'H¿QHG,Q3RRO´ZHGH¿QHD3RRODQGDOOLWV
SURS H U W LHV )ROORZ L Q JWKHS URSH U W \³KDV* H QHU LF-
BusinessProcessDiagramObject” (starting from

WKHFODVVSURFHVVZHGH¿QHWKH6WDUW(YHQWWKH
WDVNWKH(QG(YHQWDQGWZRVHTXHQFHÀRZRQH
has the start event as source and the task as target
and another has the task as source and the end
event as target.
Obviously, the BPMN ontological metamodel
development is supported by our editor where it
LV KLGGHQ WKH %301 FRPSOH[LW\ DQG WKH ¿QDO
design (with or without all details) may be saved
in OWL format. The process representation so
obtained follows the metamodel and it is an
instance of it.
FUTURE TRENDS
6WDUWLQJ IURP DQ RQWRORJLFDO GH¿QLWLRQ RI WKH
BPMN metamodel and thus from an ontological
GH¿QLWLRQRIWKHPRGHORXUQH[WVWHSLVRULHQWHG
to two different directions.
One way is to understand the possible key
performance indicator and thus to design and
implement a tool able to make the simulation of
the process in order to provide to the manager the
possibility to understand possible management
and/or design errors and to correct them immedi-
ately. The ontology representation of the process
may help in this work thanks to the possibility to
associate rules to the ontology and thus to think
on a reasoning tool.
6WDUWLQJIURPWKHLGHDWRGH¿QHDOLJKWIUDPH-
work to manage processes and to help in the design
and implementation of Web application based on

a process design, we look to the guidelines that
support the steps from process design to Web
application design. Guidelines will be, as soon as
possible, a Web application design methodology
WKDWVXSSRUWVWKHSURFHVVGH¿QLWLRQ7KLVGHVLJQ
methodology will cover the gap in the traditional
Web application design methodology based on a
³XVHUH[SHULHQFH´SDUDGLJPEXWQRWIRFXVHGRQ
a process.
CONCLUSION
Starting from the necessity to introduce the pro-
cess management in the overall IS architecture,
our work focuses on the description of business
process in a formal way aimed to automate the
EXVLQHVVSURFHVVPDQDJHPHQW7KHIRFXVLV¿UVW
on the notation to adopt: We think that the nota-
tion must be complete and expressive with the
goal to cover the traditional semantic gap be-
tween business and IT experts. Our choice is the
BPMN notation because BPMN goal is nearer to
our goal. We focus also on the formal description
of the business process in a machine–readable
language. Observing that to represent the process
we need a metamodel that is a guideline in the
PRGHOGH¿QLWLRQWKHIRFXVRQWKHOD\HU02)
metamodel is very helpful for our purpose but
we highlight that the UML language used is poor
for several reasons. So we introduce an innova-
tive way to think on ontology different from the
traditional Semantic Web use. OWL provides us

several advantages both in the description of the
metamodel and in the description of the model.
OWL and Semantic Web technologies will help
us in our future works.
REFERENCES
Berners-Lee, T., Hendler, J., & Lassila, O. (2001,
May). The Semantic Web. 6FLHQWL¿F$PHULFDQ
284(5), 34-43.
468
A Design Tool for Business Process Design and Representation
Gruber, T. R. (1993). A translation approach to
SRUWDEOH RQWRORJ\ VSHFL¿FDWLRQV Knowledge
Acquisition, 5, 199-220.
Hammer, M. (1990). Reengineering work: Don’t
automate, obliterate. Harward Business Review,
68, 104-112.
Miers, D., Harmon, P., & Hall, C. (2006). The
2006 BPM suites report Realise 2.0. Retrieved
September 30, 2006, from rends.
com/reports_toc_01.cfm
Retrieved March 15, 2005, from http://www.
bptrends.com
Object Management Group (OMG). (1997, Sep-
tember). 2EMHFWFRQVWUDLQWODQJXDJHVSHFL¿FDWLRQ
(Version 1.1).
Object Management Group (OMG). (2001, Sep-
tember). 8QL¿HGPRGHOLQJODQJXDJHVSHFL¿FDWLRQ
(Version 1.4).
Object Management Group (OMG). (2003, Sep-
tember 8). 80/VXSHUVWUXFWXUHVSHFL¿FDWLRQ

(Version 2.0).
White, S. A. (2004, May 3). Business process mod-
eling notation (BPMN) (Version 1.0 ). Retrieved
May, 2004, from
World Wide Web Consortium (W3C). (2004a,
February 10). OWL Web ontology language
reference.
World Wide Web Consortium (W3C) (2004b,
February 10). RDF vocabulary description lan-
guage 1.0: RDF Schema.
World Wide Web Consortium (W3C) (2004c,
February 10).5');0/V\QWD[VSHFL¿FDWLRQ
ENDNOTE
1
Specialization subclass is also called IS-A,
and it allows us to connect a general concept
ZLWKDPRUHVSHFL¿FFRQFHSW,QRXUH[DPSOH
an AbstractProcess IS-A Process.
This work was previously published in Semantic Web Technologies and E-Business: Toward the Integrated Virtual Organization
and Bu, edited by A. Salam and J. Stevens, pp. 77-100, copyright 2007 by IGI Publishing (an imprint of IGI Global).
469
Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Chapter 2.10
B2C Failures:
Toward an Innovation Theory
Framework
Anil M. Pandya
Northeastern Illinois University, Chicago, USA
Nikhilesh Dholakia
University of Rhode Island, USA

EXECUTIVE SUMMARY
This article uses concepts derived from the in-
vestigation of product and services innovation
failures to develop a strategic market framework
to help understand why so many Internet-based
business-to-consumer (B2C) companies failed to
IXO¿OOWKHLULQLWLDOSURPLVH%&FUDVKHVYLHZHG
collectively, may be seen as representing an initial
wave of failure of an entirely new class of tech-
nology-driven services. Such services sought to
inform, promote, sell, and deliver B2C items in
radically unfamiliar ways. Research shows B2C
¿UPVIDLOHGEHFDXVHWKH\GLGQRWIROORZWLPH
tested business precepts, but does not tell us why.
In addressing this question, this article argues
WKDWX Q VXFFHVVI XO%&¿U PVIDLOHGWRUHDOL]HWKH\
were marketing innovative services. It focuses on
WKHGLI¿FXOW\RIPDUNHWLQJLQQRYDWLYHVHUYLFHVE\
developing an integrated framework using the
continuum of need-solution context, in conjunc-
tion with the notion that seller/buyer perceptions
about the scope of innovations are not necessarily
concordant. Matched perceptions lead to success,
but not always because sellers and buyers can both
misjudge the nature and scope of an innovation.
Using secondary sources, the article illustrates
the explanatory power of the framework and
contributes to e-commerce management issues
by clarifying why, despite resource availability,
PRVW%&¿UPVIDLOHGLQWKHLQLWLDOURXQG

INTRODUCTION
This article proposes an innovation theory-based
framework to help understand why so many
Internet-based business-to-consumer (B2C)
FRPSDQLHVIDLOHGWRIXO¿OOWKHLULQLWLDOSURPLVH
B2C dot.com crashes represent special types of
470
B2C Failures
innovation failures. Our analysis shows that the
SURGXFWLQQRYDWLRQFDOOHG³%&HFRPPHUFH´LQ
LWVLQLWLDOLQFDUQDWLRQZDVÀDZHG
In innovative B2C settings, consumers bal-
ance the cost of time and efforts against services
received, and make judgments about service
quality (Berry, Seiders, & Grewal, 2002). In the
B2C environment, service quality depends on:
(1) the process by which perceptions about the
quality are formed, and (2) the gap between the
perception of the service and the experience of
the delivered service (Brady & Cronin, 2001;
Zeithaml & Bitner, 2003).
Furthermore, in high-tech marketing con-
texts, two factors shape perceived vs. expected
performance. These are the need-solution context
(Leonard-Barton, Wilson, & Doyle, 1995), and the
congruence of perceptions between technology
innovators and technology consumers (Rangan
& Bartus, 1995). In this article, we propose that
in the initial wave of B2C service innovations,
buyers and sellers marched down very divergent

paths. Technology innovators and sellers saw
B2C technologies as being capable of radically
exceeding buyers’ expectations, while buyers
saw B2C innovations as relatively inconvenient
ways of performing familiar shopping tasks.
0DQ\%&¿UPVLQWKH¿UVWZDYHIRFXVHGPRUH
attention on marketing and front-end technol-
ogy, and less on timely delivery and customer
satisfaction. The results were persistently high
FXVWRPHUDFTXLVLWLRQFRVWVZLWKRXWVXI¿FLHQW
revenues (Agarwal, Arjona, & Lemmens, 2001).
5HVHDUFK VKRZVWKDW PRVW %&¿UPV IDLOHG WR
adhere to conventional management principles
(Varianini & Vaturi, 2000). Thus, we ask: Why
GLGVRPDQ\¿UPVZLWKUHVRXUFHVDQGWDOHQWVIDLO
to use time-honored principles? What was it about
this new technology and service delivery method
that these managers misread? We argue that the
¿UPVIDLOHGWRUHDOL]HWKH\ZHUHGHDOLQJZLWKD
new innovative situation, which needed a new
managerial orientation (Achrol & Kotler, 1999).
And as we argue later in the article, the innova-
WLYHWHFKQRORJ\UHTXLUHG%&¿UPVWRDOWHUWKHLU
view of business-customer relations. We develop
a framework to show how innovation context can
lead to problems of concordance of perceptions
between buyers and sellers. Managers have to
¿JXUHRXWKRZWRQDYLJDWHWKLVPD]HRISHUFHS-
tions to be successful.
The article is organized as follows. The next

section examines the scope of the B2C failure
SUREOHP DQG UHÀHFWV RQ VRPH HDUO\ GLDJQRVHV
of it. The section that follows presents the pro-
posed innovation theory-based framework and
provides illustrative evidence. The concluding
section draws together the main arguments,
and makes recommendations for managers and
researchers.
SCOPE OF THE B2C PROBLEM
B2C e-commerce failure was a system-wide
failure. It was not the case where a few manag-
HUVLQDIHZVWDUWXS¿UPVPDGHSRRUMXGJPHQW
FDOOV7KH¿QGLQJWKDWPDQDJHUVRIWKHVH¿UPV
failed to follow well-known management truths
— that costs cannot consistently exceed revenues,
pricing correctly is critical, or customer services
are important for long-term success (Agarwal,
Arjona, & Lemmens, 2001; Marn, 2000; Varianini
& Vaturi, 2000) — is important and necessary,
EXWQRWLQLWVHOIVXI¿FLHQW,QFRPSHWHQFHLVQRWD
VDW LVIDF WRU \DQ VZHUZKH QDODUJHQX PE HURI¿U PV
IDLO7KLVVHFWLRQ¿UVWGHVFULEHVWKHHQRUPLW\RI
the B2C failure, followed by a brief review of the
HPSLULFDO¿QGLQJVWKDWKDYHFDUHIXOO\ORRNHGDW
the B2C failures. This discussion sets the stage
for our contribution and develops our framework
to address the question of why these managers in
DOOWKHVH¿UPVIDLOHG
The Size of the Crash
The B2C market crash was massive and economi-

cally destabilizing. It wiped out billions of dollars
471
B2C Failures
of market capitalization and led to huge loss of
employment. Between 1995 and 2000 a total of 492
Internet-related companies raised $36.3 billion in
capital in the public markets. By 2000, just 11%
of these companies were trading at prices greater
than their offer price. A third of them were trading
below 80% of the offer price. In 1999, 230 Initial
Public Offerings (IPOs) raised $18.2 billion. In
2000, 130 IPOs were offered and raised $12.8
billion, but 133 IPOs representing $10.4 billion
were withdrawn from the market. The market capi-
talization of the Internet sector in 1999 was $881
billion. This plunged to $208 billion by December
2000 (Anderson, 2001). Layoffs in the industry
in the year 2000 were 4,805 in September, 5,677
in October, and by December, the total layoffs
stood at 22,267. Unemployment reached 700,000
for the year 2001 (Corcoran, 2002; Rock, 2000).
7DEOHVKRZVSUR¿OHVRIVRPHRIWKH%&¿UPV
that failed during the 1999-2000 period.
Preliminary Diagnosis
In most B2C debacles, there was persistent dis-
crepancy between high customer acquisition costs
and low revenues. In early 1999, B2C companies
spent more than $1,100 to acquire a typical cus-
tomer, someone who spent only $400. In late 1999,
the average cost of customer acquisition reduced

to $800, but customer spending remained at or
below $400. The average monthly losses per B2C
site were $1 million and $1.1 million, respectively,
LQWKH¿UVWDQGVHFRQGKDOIRIGHVSLWHWKH
reduction in customer acquisition costs (Agarwal,
Arjona, & Lemmens, 2001).
As supplemental revenue sources, B2C Web
sites relied on subscription fees and advertising.
But by the end of 1999, advertising revenue per
visitor had declined from about $1 to 50 cents,
and the rates charged to advertisers had declined
between 3% and 12%. Among the content provid-
ers, fewer than 10% were garnering subscription
revenues in late 1999 (Agarwal, Arjona, & Lem-
mens, 2001).
In the false hope that the Internet was a price
elastic market, many B2C businesses maintained
low margins. However, in reality, instead of
clicking across multiple sites, 80% to 90% of
buyers of books and CDs visited only one site,
even though prices of books and CDs across Web
sites varied by as much as 25% to 30%. Hence it
appears that in this segment of the B2C market,
LQVWHDGRIGHHSGLVFRXQWLQJDVWXWH¿UPVFRXOG
KDYHFKDUJHGKLJKHUSULFHVZLWKRXWVDFUL¿FLQJ
UHYHQXHVDQGSUR¿WV0DUQ
Only a small but solid group of companies
managed to achieve visitor conversion rates of
12%, customer churn rates below 20%, and repeat
purchase rates of around 60% (Agarwal, Arjona,

& Lemmens, 2001). The majority of the B2C
IDLOXUHVZHUHFDXVHGE\³IDWDODWWUDFWLRQ´OXULQJ
visitors to the site but failing to convert them into
FXVWRPHUV 6XFFHVVIXO ¿UPV JHQHUDWHG QHDUO\
three times the gross income from repeat custom-
ers as from one-time buyers. This difference in
performance was the result of superior skills in
acquiring (keeping the customer acquisition cost
low), converting (keeping the purchase process
simple), and retaining customers (making opera-
tional execution satisfactorily reliable, i.e., sites
download quickly, on-time delivery, and ease of
return or exchange of purchases). Additionally, the
successful companies followed the tried-and-true
principles from the brick-and-mortar world:
1. focus on core product or service propositions
WKDW¿WWKHQHHGVRIZHOOGH¿QHGFXVWRPHU
segments;
2. control extensions of product lines and busi
-
ness models, and focus on the core products;
and
 DYRLG³EOHHGLQJHGJHWHFKQRORJLHV´LH
leading but unproven technologies) and fo-
cus on basic product presentation features,
customer service tools, and logistics.
,QVKRUWWREHVXFFHVVIXOD¿UPPXVW¿QGLWV
Q D W X U D O F X V W R P H U V H I ¿ F LH Q W O \ R I I H U W K H P Z K D W W K H \ 
472
B2C Failures

B 2C Firm Internet
Business
Lifespan S tatus Failure Date
Redrocket.com Nickelodeon's
toy sales site
January
1995-0Dy
5, 2000
Acquired by Viacom in
1999
0Dy 2000
BBQ.Com Grilled meats
and sauces
June 15,
1999-0Dy
15, 2000.
Standing as
Barbequemall.com
0Dy 2000
DEN Digital
Entertainment
Network
1996-0Dy
17, 2000
Investment $50 million
$75 million IPO shelved
0Dy 2000
Toysmart.com Smart “good”
(not destructive
or faddish)

toys
encouraged
parents to “click
on your child’s
potential”
January
22, 1999-
0Dy 19,
2000.
Controlling interest by
Disney in August 2000
with $50 million
investment
Disney stopped

support
Pixelon An investment
scam
1996-
August
2000
Bilked $30 million from
investorV promoter
lived in his car, spent
half the capital on
launch
Turned himself

over to the
police

Pop.com Received $50
million from
Paul Allen of
0Lcrosoft
The 80-
employee
content
site never
launched
The sitebacked by
Steven Spielberg,
Jeffrey Katzenberg,
David Geffen, Ron
Howard, and Brian
Grazerburned $18
million
Closed 14 days

prior to the
launch
Pseudo.com Streaming
media
1994-
September
18, 2000
Received $18 million in
1999 and $14 million in
Spring 2000
With 10
channels

chasing 10
different
genres with
streaming rich-
media content
across
broadband,
Pseudo could
not establish
its core
audience
Ingredients.com The company
developed and
manufactured
personal care
products for
Web
distribution
1998-
October
2000
Received $4.5 million in
funding
0 any beauty
siteV
competition
from major
brands and
lack of
originality or

funding saw to
its demise
Table 1. Examples of B2C failures in 1999-2000
Source: Authors’ research based on multiple sources in the reference list
473
B2C Failures
want, and deliver it reliably (Agarwal, Arjona, &
Lemmens, 2001).
B2C Internet businesses were hemorrhaging
money because they failed to follow basic mar-
keting principles. When electronic commerce
was young, it seemed that marketing expertise
could be cast aside. A speedy grab for a large
share of the market with the aim of getting as
many visitors as possible to the site became the
immediate objective. It was assumed, falsely by
PRVW%&¿UPVWKDWDWVRPHVWDJHWKHVHYLVLWRUV
ZRXOGWUDQVODWHLQWRSUR¿WDEOHUHSHDWFXVWRPHUV
(Varianini & Vaturi, 2000).
&ROOHFWLYHO\WKHVH¿QGLQJVFRQ¿UPWKDWPRVW
%&¿UPVIDLOHGWRDGKHUHWRFRQYHQWLRQDOPDQ-
DJHPHQWSULQFLSOHVRIHI¿FLHQF\DQGFRQVXPHU
focus. This raises important questions: Why did
VRPDQ\¿UPVZLWKVRPXFKWDOHQWDQGHDV\DF-
FHVVWRFDSLWDOVRPHSUR¿OHGLQ7DEOHIDLOWRXVH
time-tested management principles? Why did they
fail to realize that managing a B2C e-commerce
required acumen similar to that required in the
conventional brick-and-mortar world? What was
it about this new technology and service delivery

method that most B2C managers misread? Why
could these companies not convert visitors into
SUR¿WDEOH DQG OR\DO FXVWRPHUV" 7KH IROORZLQJ
section addresses these questions.
Toward Innovation Theory-Based
Diagnosis
The Internet implies a revolutionary shift in mar-
keting approach. From being agents of the seller,
%&¿UPVKDYHWREHFRPHDJHQWVRIWKHEX\HU
(Achrol & Kotler, 1999). The Internet allows B2C
¿UPVWRJHWFORVHWRWKHLUFXVWRPHUVLQZD\VQRW
SRVVLEOHEHIRUH)RUWKH¿UVWWLPHWKHVH¿UPVFDQ
represent buyers to the producers of goods rather
WKDQWKHRWKHUZD\DURXQG7KH\FDQ¿JXUHRXW
what their customers want and communicate that
GHPDQGWRWKHSURGXFHUV%&¿UPVIDLOHGWRDS-
preciate this change and did not quite grasp that
their business represented a radical service innova-
tion, not new technology business. Additionally,
they were offering known products to customers
in a new, more convenient, and faster way than
what was possible before. This new service con-
cept emerged of the new technology, but that is
not the whole story. Old retail knowledge about
assessing demand, buying, managing inventory,
developing displays, and improving delivery tech-
nology — all have to be seamlessly integrated to
F U H D W H D V D W L V ¿ H G F X V W R P H U7 K L V V X E W O H E X W V W U D W H J L F 
shift in perspective could have refocused B2C
managerial efforts from being preoccupied with

the new technology and market share dominance,
to providing consistently high-quality customer
service and convenience geared towards foster-
ing buyer loyalty and customer retention. While
DFNQRZOHGJLQJWKDWPDQ\%&¿UPVVWUD\HGIURP
simple and conventional — but nonetheless vital
and relevant — management principles, we present
a theory-based framework to capture some of the
additional underlying complexity of B2C failures,
and to identify ingredients for B2C success.
INNOVATION THEORY-BASED
FRAMEWORK AND EVIDENCE
The classical theory of innovation diffusion
suggests that, to be widely adopted, innovations
must be compatible with existing habits; provide
incentives to change; and pose low physical, social,
economic, and psychological risks to adopters.
Overlaying this set of barriers is the inherent dif-
¿FXOW\RIPDLQWDLQLQJFRQVLVWHQWVHUYLFHTXDOLW\
which becomes important at the back end of a
B2C operation. At the front end, not all customers
are equally savvy about technology, and hence
their involvement can lead to frustration when it
comes to navigating Web sites. Thus B2C man-
agers face serious problems involving customer
perceptions and behavior change when purchas-
ing goods on the Internet. The service quality
model well known in the services marketing
OLWHUDWXUHLGHQWL¿HVIRXUJDSVQRWNQRZLQJ