Description
Predictive Intelligence A New Frontier In Achieving Market Leadership
BUSINESS MANAGEMENT
154 | BMUS
Predictive intelligence:
a new frontier in achieving market leadership
By Dr Pawan Singh and Dr Mohamed Latib
ooted in rigorous scientific methodologies, predictive intel-
ligence provides on-demand conversion of attitudinal and
likely-behavior data into validated insights. As managers
peer into the future with increased certainty, the results can be truly
transformational.
MODELING BUSINESS INTELLIGENCE
Business intelligence (BI) technology has been advancing in
response to a critical need to devise valid plans of action in the face
of increasing uncertainty in the global business environment. BI,
defined as the ability to draw insights from stores of existing data,
offers a powerful starting point for making better decisions. Nu-
merous case histories indicate that the impact of BI-driven deci-
sions can result in highly significant strategic competitive advan-
tage and cost savings running into millions of dollars. Today the
term business intelligence is applied to a wide range of applica-
tions, many no more than charting and graphic applications. This
does a great disservice to the perceived value of genuine business
intelligence technology. To add clarity to the field, Dr Singh devel-
oped a model for defining business intelligence in 2001, which
divides business intelligence into three classes based on the source
of the underlying information. They are:
INFERENTIAL INTELLIGENCE
Inferential intelligence is the drawing of inferences from dis-
connected pieces of information available either publicly or col-
lected through secondary research. Data gathering and classifica-
tion has been automated to a considerable extent using the internet
and semantic analysis tools. Still, the ability to draw intelligent
inferences resulting in better decisions is primarily dependent on
the diagnostic skill of the analyst.
DERIVED INTELLIGENCE
The ability to analyze and draw conclusions from transactional,
demographic and other types of stored data forms the basis for
derived intelligence. Data mining, business analytics and predictive
analytics are examples of derived intelligence, and these terms are
often used interchangeably with business intelligence. Clearly the
value of derived intelligence applications rests on the quality of the
data and the validity of the algorithms used in its analysis.
Because derived intelligence is based on historical information,
it is a lagging indicator whose value diminishes rapidly with time.
As the business environment has become more volatile, the value
of derived intelligence is increasingly dependent on the quality and
newness of the underlying data. This has resulted in the need to
maintain large and expensive software systems, creating uncer-
tainty that the value of such intelligence will be worth its cost. The
checkered history of customer relationship management (CRM)
systems provides a good example of that uncertainty.
PREDICTIVE INTELLIGENCE
A key driver for adoption of predictive intelligence systems is
the businesses’ need to know the customer as a system of disposi-
tions and behaviors.
Appropriately, rigorous scientific methods can accomplish this
even though people seem unpredictable. A common misconception
about science is that it is all about certainty, when in reality prob-
abilistic scientific applications are far more pervasive than applica-
tions of certainty.
Predictive intelligence (P1) is based on real-time sampling of
attitudinal and likely-behavior data. P1 is rooted in rigorous scien-
tific methodologies and provides on-demand conversion of data
into insights and intelligence through advanced analytics and visu-
alization tools.
Predictive intelligence comes from analyzing responses to
questions asked of the very people that will be affected by it. Its
power springs from this direct correlation. Predictive intelligence is
the most coveted form of business intelligence because it provides
the leading indicators that businesses need.
The underlying data used in PT systems is often collected using
questionnaires or surveys. Our research, however, indicates that a
vast majority of these surveys do not apply the rigor or possess the
validity to qualify as a solid foundation for predictive intelligence.
Data that does not correctly represent the underlying population, or
that does not meet validity criteria, produces results that are inaccu-
rate at best, and often quite harmful. The criticality of scientific
rigor and validity to predictive intelligence shouldn’t be under-
stated.
Advanced predictive intelligence systems convert these insights
into metrics that drive strategic action and measure its effectiveness
R
BUSINESS MANAGEMENT
BMUS | 155
over time. The effect can be truly transformational and can help
businesses achieve sustainable competitive advantage.
THE PREDICTIVE INTELLIGENCE PROCESS
Predictive intelligence is a process. It uses an intellectual
framework that combines thought leadership, deep experience, an
understanding of key scientific, business and behavioral issues,
scientific rigor, and sophisticated predictive modeling. Each com-
ponent is equally important to the generation of high-quality, ac-
tionable, strategy-level insights.
• Intellectual framework: Each PT project begins by defining the
elements critical to the program’s success: specific objectives, pri-
mary response audience(s) and respondent qualifications.
• Questionnaire design: Advanced survey design, including intel-
ligent branching and response validation, is necessary to ease data
collection and protect data from bias or other quality-eroding ef-
fects.
• Deployment: To increase response rates respondents should be
allowed to control the timing and pace of their participation. Multi-
ple levels of user identification and password protection help to
assure the participants that their responses are secure. Third- party
anonymity can also be useful in eliminating respondent bias.
• Data collection and validation: Data collection should include
response validation, checking for such things as mandatory and
appropriate information and duplicate responses.
• Analysis and reporting: The analysis and reporting process
should focus on identifying actionable strategies to address specific
research objectives, turning raw data into reliable information for
decision-making. Results should be delivered in a presentation
format that’s easy to understand and effectively conveys the re-
search findings.
• Metrics development: The development of appropriate metrics
helps organizations to measure and benchmark their current per-
formance and monitor it as it changes over time. To be transforma-
tional these metrics must relate directly to the strategic goals of the
company and measure characteristics that the organization can
influence.
RAISING THE ORGANIZATIONAL
KNOWLEDGE LEVEL
Predictive intelligence allows businesses to quickly understand
current market conditions, evaluate market opportunities and de-
termine the potential for success when entering new markets or
launching new products. It provides top management with reliable
data for strategy implementation.
Often these transformational effects are realized as a highly
valuable outgrowth of intelligence projects that were originally
undertaken for non-transformational reasons. York International
experienced this when it chose PeriscopeSOX, an intelligence sys-
tem from PeriscopeIQ to help the company comply with the new
Sarbanes-Oxley regulations. Ultimately, the system provided much
more. “The primary goal of implementing this solution was to
comply with the law,” says Ian Howells, Director, Corporate Con-
trol at York, “but PeriscopeSOX has provided us with a business
excellence tool that helps us identify and address issues more effi-
ciently.”
Web-based predictive intelligence systems are able to reach out
and collect accurate data from areas once obscured by their separa-
tion from the business. Ingersoll-Rand Waterjet (IRWJ), a leading
global maker of ultra-high pressure waterjet machinery, is able to
collect valid intelligence from their end-users even though they are
separated from them by a network of OEM manufacturers who use
IRWJ components.
“Utilizing the PeriscopeIQ online survey solution, the project
took two weeks from start to finish,” says Greg Mort, Manager,
IRWJ Marketing and Services. “A paper-based method would have
taken us quite a few months to complete, without the quality of
data, response rate and strategic analysis PeriscopeIQ was able to
provide. The solution was cost-effective and helped us to quickly
transform our relationships with our OEMs and significantly in-
crease our customer satisfaction.”
IDEX Corporation, a global manufacturing and maker of the
Hurst Jaws of Life rescue tool, needed to have a clear understand-
ing of the competitive environment and the future of their market,
and to determine how to maximize its relationships with its dis-
tributors and customers.
Owing to the advanced capabilities of predictive intelligence
technology from PeriscopeIQ the original survey was transformed
into a rigorously validated study that covered a wide range of areas,
including critical-to-customers issues, competitive rankings, prod-
uct features, end-user preferences, competitor behavior, market
conditions and sales predictions.
The survey’s analytical results enabled the calculation of stra-
tegic metrics for the top management on critical customer issues
and brand perception, loyalty and equity. A ‘customer satisfaction
study’ became a strategic business tool designed to help achieve
operational excellence through organizational transformation.
Derived and inferential intelligence no longer provide the ad-
vantage that industry leaders demand to reduce uncertainty and risk
in a rapidly changing global marketplace. The need to know is now
– to create new solutions, to predict new markets, and to enhance
profitability and power. Only predictive intelligence can meet the
growing demand to know now. ?
Dr Pawan Singh is co-founder and Intellectual Architect of PeriscopeIQ, and
Dr Mohamed Latib is co-founder and Vice President of PeriscopeIQ.
doc_339382026.pdf
Predictive Intelligence A New Frontier In Achieving Market Leadership
BUSINESS MANAGEMENT
154 | BMUS
Predictive intelligence:
a new frontier in achieving market leadership
By Dr Pawan Singh and Dr Mohamed Latib
ooted in rigorous scientific methodologies, predictive intel-
ligence provides on-demand conversion of attitudinal and
likely-behavior data into validated insights. As managers
peer into the future with increased certainty, the results can be truly
transformational.
MODELING BUSINESS INTELLIGENCE
Business intelligence (BI) technology has been advancing in
response to a critical need to devise valid plans of action in the face
of increasing uncertainty in the global business environment. BI,
defined as the ability to draw insights from stores of existing data,
offers a powerful starting point for making better decisions. Nu-
merous case histories indicate that the impact of BI-driven deci-
sions can result in highly significant strategic competitive advan-
tage and cost savings running into millions of dollars. Today the
term business intelligence is applied to a wide range of applica-
tions, many no more than charting and graphic applications. This
does a great disservice to the perceived value of genuine business
intelligence technology. To add clarity to the field, Dr Singh devel-
oped a model for defining business intelligence in 2001, which
divides business intelligence into three classes based on the source
of the underlying information. They are:
INFERENTIAL INTELLIGENCE
Inferential intelligence is the drawing of inferences from dis-
connected pieces of information available either publicly or col-
lected through secondary research. Data gathering and classifica-
tion has been automated to a considerable extent using the internet
and semantic analysis tools. Still, the ability to draw intelligent
inferences resulting in better decisions is primarily dependent on
the diagnostic skill of the analyst.
DERIVED INTELLIGENCE
The ability to analyze and draw conclusions from transactional,
demographic and other types of stored data forms the basis for
derived intelligence. Data mining, business analytics and predictive
analytics are examples of derived intelligence, and these terms are
often used interchangeably with business intelligence. Clearly the
value of derived intelligence applications rests on the quality of the
data and the validity of the algorithms used in its analysis.
Because derived intelligence is based on historical information,
it is a lagging indicator whose value diminishes rapidly with time.
As the business environment has become more volatile, the value
of derived intelligence is increasingly dependent on the quality and
newness of the underlying data. This has resulted in the need to
maintain large and expensive software systems, creating uncer-
tainty that the value of such intelligence will be worth its cost. The
checkered history of customer relationship management (CRM)
systems provides a good example of that uncertainty.
PREDICTIVE INTELLIGENCE
A key driver for adoption of predictive intelligence systems is
the businesses’ need to know the customer as a system of disposi-
tions and behaviors.
Appropriately, rigorous scientific methods can accomplish this
even though people seem unpredictable. A common misconception
about science is that it is all about certainty, when in reality prob-
abilistic scientific applications are far more pervasive than applica-
tions of certainty.
Predictive intelligence (P1) is based on real-time sampling of
attitudinal and likely-behavior data. P1 is rooted in rigorous scien-
tific methodologies and provides on-demand conversion of data
into insights and intelligence through advanced analytics and visu-
alization tools.
Predictive intelligence comes from analyzing responses to
questions asked of the very people that will be affected by it. Its
power springs from this direct correlation. Predictive intelligence is
the most coveted form of business intelligence because it provides
the leading indicators that businesses need.
The underlying data used in PT systems is often collected using
questionnaires or surveys. Our research, however, indicates that a
vast majority of these surveys do not apply the rigor or possess the
validity to qualify as a solid foundation for predictive intelligence.
Data that does not correctly represent the underlying population, or
that does not meet validity criteria, produces results that are inaccu-
rate at best, and often quite harmful. The criticality of scientific
rigor and validity to predictive intelligence shouldn’t be under-
stated.
Advanced predictive intelligence systems convert these insights
into metrics that drive strategic action and measure its effectiveness
R
BUSINESS MANAGEMENT
BMUS | 155
over time. The effect can be truly transformational and can help
businesses achieve sustainable competitive advantage.
THE PREDICTIVE INTELLIGENCE PROCESS
Predictive intelligence is a process. It uses an intellectual
framework that combines thought leadership, deep experience, an
understanding of key scientific, business and behavioral issues,
scientific rigor, and sophisticated predictive modeling. Each com-
ponent is equally important to the generation of high-quality, ac-
tionable, strategy-level insights.
• Intellectual framework: Each PT project begins by defining the
elements critical to the program’s success: specific objectives, pri-
mary response audience(s) and respondent qualifications.
• Questionnaire design: Advanced survey design, including intel-
ligent branching and response validation, is necessary to ease data
collection and protect data from bias or other quality-eroding ef-
fects.
• Deployment: To increase response rates respondents should be
allowed to control the timing and pace of their participation. Multi-
ple levels of user identification and password protection help to
assure the participants that their responses are secure. Third- party
anonymity can also be useful in eliminating respondent bias.
• Data collection and validation: Data collection should include
response validation, checking for such things as mandatory and
appropriate information and duplicate responses.
• Analysis and reporting: The analysis and reporting process
should focus on identifying actionable strategies to address specific
research objectives, turning raw data into reliable information for
decision-making. Results should be delivered in a presentation
format that’s easy to understand and effectively conveys the re-
search findings.
• Metrics development: The development of appropriate metrics
helps organizations to measure and benchmark their current per-
formance and monitor it as it changes over time. To be transforma-
tional these metrics must relate directly to the strategic goals of the
company and measure characteristics that the organization can
influence.
RAISING THE ORGANIZATIONAL
KNOWLEDGE LEVEL
Predictive intelligence allows businesses to quickly understand
current market conditions, evaluate market opportunities and de-
termine the potential for success when entering new markets or
launching new products. It provides top management with reliable
data for strategy implementation.
Often these transformational effects are realized as a highly
valuable outgrowth of intelligence projects that were originally
undertaken for non-transformational reasons. York International
experienced this when it chose PeriscopeSOX, an intelligence sys-
tem from PeriscopeIQ to help the company comply with the new
Sarbanes-Oxley regulations. Ultimately, the system provided much
more. “The primary goal of implementing this solution was to
comply with the law,” says Ian Howells, Director, Corporate Con-
trol at York, “but PeriscopeSOX has provided us with a business
excellence tool that helps us identify and address issues more effi-
ciently.”
Web-based predictive intelligence systems are able to reach out
and collect accurate data from areas once obscured by their separa-
tion from the business. Ingersoll-Rand Waterjet (IRWJ), a leading
global maker of ultra-high pressure waterjet machinery, is able to
collect valid intelligence from their end-users even though they are
separated from them by a network of OEM manufacturers who use
IRWJ components.
“Utilizing the PeriscopeIQ online survey solution, the project
took two weeks from start to finish,” says Greg Mort, Manager,
IRWJ Marketing and Services. “A paper-based method would have
taken us quite a few months to complete, without the quality of
data, response rate and strategic analysis PeriscopeIQ was able to
provide. The solution was cost-effective and helped us to quickly
transform our relationships with our OEMs and significantly in-
crease our customer satisfaction.”
IDEX Corporation, a global manufacturing and maker of the
Hurst Jaws of Life rescue tool, needed to have a clear understand-
ing of the competitive environment and the future of their market,
and to determine how to maximize its relationships with its dis-
tributors and customers.
Owing to the advanced capabilities of predictive intelligence
technology from PeriscopeIQ the original survey was transformed
into a rigorously validated study that covered a wide range of areas,
including critical-to-customers issues, competitive rankings, prod-
uct features, end-user preferences, competitor behavior, market
conditions and sales predictions.
The survey’s analytical results enabled the calculation of stra-
tegic metrics for the top management on critical customer issues
and brand perception, loyalty and equity. A ‘customer satisfaction
study’ became a strategic business tool designed to help achieve
operational excellence through organizational transformation.
Derived and inferential intelligence no longer provide the ad-
vantage that industry leaders demand to reduce uncertainty and risk
in a rapidly changing global marketplace. The need to know is now
– to create new solutions, to predict new markets, and to enhance
profitability and power. Only predictive intelligence can meet the
growing demand to know now. ?
Dr Pawan Singh is co-founder and Intellectual Architect of PeriscopeIQ, and
Dr Mohamed Latib is co-founder and Vice President of PeriscopeIQ.
doc_339382026.pdf