What type of DSS is it?
by Dan Power
Editor, DSSResources.com
Managers and developers are recognizing that products and systems advertized as business intelligence, decision support systems or knowledge management systems are not solving the same problem. The question of "what type of DSS is a specific decision making relevant software product or system?" is common and recurring. The message is getting out that all DSS are not the same. DSS researchers have observed and recognized the differences
for many years, but now managers, developers and vendors are realizing that
differentiation or categorization of DSS is important and "real". Classification of
objects, things, and artifacts is partly systematic analysis and partly
the application of specific criteria from a rubric that is linked to a
classification scheme. Many of us played the game 20 questions during
our youth or as parents with our children. The game is fun, low cost
and educational and it teaches classification skills. Often times the
player begins with a question like "Is it a plant, animal or mineral?"
The 20 questions elicit information about the object and help determine
what it is. When classifying specific DSS or DSS generators this approach can be applied.
Let's use a revenue management application as an example. Some revenue management
applications are model-driven DSS and some are decision automation
systems and hence not DSS at all. One vendor is IDeaS (Integrated Decisions & Systems, Inc.). It is a privately held venture financed company (cf., http://ideas.com)founded in 1989 by Dr. Ravi Mehrotra and some colleagues. The
company initially developed revenue optimization and yield management
software for the Airline Industry and began marketing a solution for the
hospitality industry in 1996.
What type of DSS software development package does the company market?
The company claims it markets "automated
decision execution technology". This phrase indicate IDeaS may not be a DSS at all. Also, "IDeaS technology is fundamentally differentiated as a true 'decision' technology, combining analytics with scientific decision modeling." Analytics is often synonymous with model-driven decision support. If pricing is "automated" and the computer software makes the decision rather than a pricer or a manager, then the software is better classified as a decision automation system. DSS keep a human decision maker in the loop.
In 1999, Joan Marsan had a review of Revenue Management Systems in
Hotels Magazine. Joan noted "In the earliest days of hotel revenue
management systems (RMSs), forecasting and optimization focused on
overall demand. If seasonal demand was predicted to be high, lower rate
categories were closed. If demand forecasts were grim, lower rate
categories opened. ... In the hotel market, advanced RMSs that handle
multiple properties with many room options can determine the cross-price
elasticity of demand when room options are opened and closed."
Technology advances have enabled software companies to develop products
to support and even make revenue management decisions. Joan notes "The
greatest enabler of RMS advancements has been the improvement of
interfaces between property management (PMS) and central reservation
systems (CRS)." RMS decision support subsystems or automated decision
systems can now use "live" data for forecasting and optimization
calculations.
According to the optims.com website "Based
on real-time demand forecasting by market micro-segment and an
optimization model, Yield Management (also known as 'Revenue
Management' or 'Real-time pricing') is an economic technique to
calculate the best pricing policy for optimizing profits generated by
the sale of a product or service, based on real-time modeling and
forecasting of demand behavior per market micro-segment." Revenue management systems are not always decision support systems.
How does one know what type of DSS is being used in a specific
situation? Why is it important to be able to differentiate or
discriminate among types of DSS? Differentiating the type of DSS IS
IMPORTANT because it helps us communicate and understand! So what
questions can differentiate types of DSS? After you have gathered as
much information as possible on the vendor's product or the system that
is being used in a company that someone calls a DSS, then you can ask the following questions.
1. Is the computerized system intended to support decision making? If
YES, then possibly a BI, DSS or KMS.
2. Does the computerized system have multiple identifiable subsystems?
If so, focus on only one subsystem at a time. Put a specific boundary
around a possible decision support subsystem.
3. Is the focal system a Decision Support System, a decision automation
system or used for a special decision support study? Check the 7 DSS evaluation criteria:
Facilitation, Interaction, Ancillary, Repeated Use, Task-oriented,
Identifiable, Decision Impact. For more information read Ask Dan! "What
are the characteristics of a Decision Support System?", March 30, 2003.
4. Does electronic communication provide the dominant decision support
functionality?
4.a. Does use of the DSS involve synchronous or asynchronous
communication and collaboration?
4.b. Are tools provided to facilitate communications about a specific
decision situation?
If all answers in Q4 group are YES, conclude --> Communications-driven
DSS.
5. Does the DSS include a large, structured store (database) of
historical data?
5.a. If so, can users query and interact with the data store?
5.b. Are real-time data updates an important component of the
application?
5.c. Are predefined screens (reports) available to users?
5.d. Is data displayed on a map or geographic representation?
5.e. Does the dominant functionality of the DSS come from rapid access
to and analysis of the data store?
If the answers to 5 and 5.e. are YES, and to some of 5.a. - 5.d.,
conclude --> Data-Driven DSS. The system may provide operational business intelligence, corporate performance management, or business activity monitoring.
6. Does the DSS include a large database of unstructured documents?
6.a. If so, can users search, retrieve, summarize and sort documents for
decision support?
6.b. Are documents used in a work flow or decision process to present
information and record evaluations?
6.c. Do documents and document retrieval and analysis provide the
dominant functionality for the DSS?
If all answers in Q6 group are YES, conclude --> Document-Driven DSS. Some may classify the system as a knowledge management system.
7. Does the DSS store and codify knowledge or expertise?
7.a. Is human expertise stored using AI technologies?
7.b. Does the DSS provide recommendations or advice?
7.c. Does the provision of expert support provide the dominant
functionality for the DSS?
If all answers in Q7 group are YES, conclude --> Knowledge-Driven DSS. Some may classify the system as a knowledge management system.
8. Does the DSS include one or more quantitative models?
8.a. Can users manipulate the model and perform "What if?" and
sensitivity analysis?
8.b. Does the interactive model analysis provide the dominant
functionality for the DSS?
If YES, conclude --> Model-Driven DSS.
When in doubt collect more information and ask more questions. An organization benefits when a shared vocabulary is used, especially for strategic issues like decision support.
*************
References
Marsan, Joan, "Smarter Revenue Management Systems," Hotels Magazine,
March 1999, http://www.hotelsmag.com/0399/0399tech.html
Mehrotra, R., "Yield Management System Measurement," IdeaS, Inc., June
1999, http://www.hotel-online.com/Trends/IDeaS/YieldMeasurement.html
Peregrine Systems, "Eye-on Yield [Revenue] Management: Revenue Revival,
Hospitality Technology, May/June 2000,
http://www.htmagazine.com/archive/MayJune2000/MayJune2000_8.shtml.
Power, D., "What are the characteristics of a Decision Support
System?" DSS News, Vol. 4, No. 7, March 30, 2003.
Modified from Power, D., "What type of DSS is X? or What type of DSS is a revenue
management application?" DSS News, Vol. 5, No. 7, March 28, 2004.
Last update: 2009-03-21 09:05
Author: Daniel Power
Print this record
Show this as PDF file
You cannot comment on this entry
