Conversation Oriented Decision Support Systems for Organizations - Enterprise Information Systems Design, Implementation and Management: Organizational Applications - page 121

and effort. The information generated may be
irrelevant to decisions' quality. Participating in
conversations about a decision to be taken does
not guarantee that people will arrive to understand
its why better or that they will comply with it
faithfully. Turning an individual decision into a
collective one may have as consequence to dilute
responsibility.
It follows that the design of a CODSS shall
have in mind to maximize the potential gains and
reduce the potential losses. To maximize gains,
the design must keep open the participation, in
the limit to all people affected by the decision. It
must make available tools for conversations to
be effective and efficient in generating relevant
contributions to decisions' quality. It must have
the capability to register, store and classify issues
and related conversations, so that it augments
organizational memory and fosters organizational
learning.
Simultaneously, the design shall have in mind
to minimize the number of interactions needed,
limit “conversational noise” and lift restrictions
of same time and same space for conversation to
occur. Such minimization of losses or decision
cost must accompany organizational learning. If
a CODSS is successful in fostering organizational
learning, the extent of conversations for the same
type of issue should decrease along time. In a
sense, if the goal of a CODSS is to support con-
versations for decision, its aim in the long term
will be to reduce conversations to the minimum
required, to be a positive factor for coordination
and collaboration to happen with minimal cost
or attrition.
Furthermore, a CODSS must be configurable
in order that those in charge of taking decisions
can tailor it according to their own understand-
ing of the realities of the organizational domain
for which they are taking the responsibility of
deciding.
A key question to be answered is which are
the structural components of a CODSS? From the
viewpoint of conversations being the glue that

permeates an organization's global decision pro-
cess, one is led to require that CODSSs can serve
organizational domains of any dimension from
small to the whole organization in the limit. This
requires independence of scale of the structural
components and an architecture enabling recursive
connections of them, so that, say, a CODSS may
grow up from a given organizational domain, at
any level in the hierarchy of responsibilities, to
the whole organization.
A CODSS should also be able to connect to
other decision support and information systems
of the organization in order to ease and speed the
flow of information among them. In this way, a
CODSS may become a hub organizing decision
and the flows of information to and from decision.
Given the posited characteristics above, I think
the organizations that can benefit first from this
concept of CODSS are those whose leaders, and
desirably the decision-makers and all participants:

1. Recognize that conversations among people
are a necessary and effective component in
a vast class of required or possible decisions
(usually the more important ones).
2. Want to explore and exploit the potential
benefits of supporting such decisions in an
integrated way along the organization based
on information systems.
3. Understand the potential risks implicit in
the implementation of the concept.

In the following, the reader may find proposals
to address several topics on CODSS raised above.

bAckground

Decision support systems are an established field
of information systems science. (Power, 2002) is
a landmark topic on DSSs. (Power, 1995-2002)
is also a good source of general information.
(Burstein & Holsapple, 2008) gives a recent in-
depth treatment of the field. GDSSs are dealt in