Module 2: The Business Value of Decision Models

Changing the Game: BPM and BDM

THE NATURE OF A DECISION MODEL’S BUSINESS VALUE
A Decision Model - aims to contribute on the health and well-being of a business.
THESE CONSIDERATIONS INCLUDE THE FOLLOWING:
 How much the Decision Model accelerates business agility
 How the Decision Model proves useful in normal, complicated and threatening times,
and in times of crisis.
 How much profit, reputation, consistency, and quality the Decision Model delivers to
the marketplace.
 How much valuable, expensive and rare business knowledge the Decision Model
secures
 How much the Decision Model corrects shortcomings in today’s best practices, such as
Business Process Management(BPM) and Business Decision Management (BDM)

EXAMPLES OF BUSINESS DECISIONS

In the previous lesson we defined
business decision - as “a conclusion that a business arrives at through business logic and
which the business is interested in managing”

Table 3.1 contains seven examples of business decisions. The first column of the table shows
that each business decision has a name that begins with a “deciding word” (e.g., calculate,
estimate, determine) and the object of that deciding word. Further, the object of that deciding
word is correlated to a conclusion fact type, which is simply the piece of information about which
the conclusion is made. And, third, the business decision and its conclusion fact type belong
within a business context (i.e., a business concept), the latter usually being a business object,
data entity, input form, Web page, etc. The business concepts in this table are person, claim,
insurance policy, loan, inventory item, and vendor. The second row in the table denotes a
business decision that comes to a conclusion about a particular attribute of the business concept
of Claim. This business decision comes to a conclusion about the amount to pay on a claim, and
this judgment is probably based on a complex evaluation of business criteria.

TABLE 3.1
Business Decision Business Conclusion Fact Type
Concept or
Context
Calculate the BMI of person Person Person BMI (Body
Mass Index)
Estimate the payment amount that should be Claim Claim Payment
paid on this claim Amount
Determine payment eligibility of claim Claim Claim Payment
Eligibility
Determine whether the insurance policy renewal Insurance Insurance Policy
method is to be automatic or manual Policy Renewal Method
Determine whether this customer meets the loan Loan Customer Loan
prequalification requirements Prequalification
Status
Assess the minimum stock level for this inventory Inventory Inventory Item
item Item Minimum Stock Level
Calculate the Vendor’s Performance Index Vendor Vendor Performance
Index
Why are business decisions, such as those in the table, so important?
Taylor and Raden - state that “organizations are perceived through the lens of the decisions
they make” (Taylor and Raden, 2007).
However, not all business decisions are equal in value. Some are of high value, others of
less value. Some are automated, others may never be automated. Any one of these business
decisions, taken by itself within the context of one transaction, may seem to have a low value.
However, a new realization is that the high volume of these operational decisions, taken as a
group, has a huge cumulative effect on an organization’s health. Other business decisions of low
volume but of a strategic nature also can have a huge impact on an organization’s health. This
means that there is an emerging business focus on how to better manage and automate
business decisions. This is the practice of BDM. Not only does BDM focus on the business value of
business decisions (and their Decision Models), BDM also advances Business Process
Management (BPM) and Service-Oriented Architecture (SOA) by promoting business
decisions to the status of a visible asset worthy of management.

THE THREE CHARACTERISTICS THAT ARE IMPORTANT TO A BUSINESS

DECISION’S BUSINESS VALUE
Business decisions vary in character a great deal, and are not made in a vacuum. There
are three characteristics of a business decision that are important to its business value as well as
to the applicability of the Decision Model:

CHARACTERISTIC #1: EACH BUSINESS DECISION HAS AN OPERATIVE

CONTEXT.
The operative context is the complexity of the business environment in which the business
decision is made. The complexity of the business environment may range from simple to very
complex. Some business decisions are made in the course of everyday business processes.
Some are made during a business crisis. Some are made when relevant facts are available,
whereas others are made in the “fog of war,” when relevant facts are not available and intuition
must substitute for them. The value of a Decision Model must take into account its operative
context.

CHARACTERISTIC #2: EACH BUSINESS DECISION VARIES IN ITS VOLUME-

BASED ECONOMIC IMPACT.
Each instance of a business decision has an economic impact on the business. An instance of a
business decision that is strategic in nature—to acquire another company, for example—may
have a greater economic impact on the organization than an instance of a business decision to
grant a particular level of discount to a given group of customers. Each business decision also
has a frequency of how often it is made, which is called its volume. For example, a business
decision to acquire another company may only be made on a rare occasion, but business
decisions about customer discounts are likely made many times a day. The collective economic
impact of a business decision, then, is a combination of the economic impact of each individual
instance and its volume. If a business decision granting discounts is made millions of times a
day, the actual financial impact of that business decision may be equivalent to that of a major
acquisition decision. The full value of a Decision Model must take into account both volume and
individual financial value.

CHARACTERISTIC #3: EACH BUSINESS DECISION VARIES IN THE COMPLEXITY

OF ITS BUSINESS LOGIC.
Some business decisions may have one or two Rule Families, each of which may contain only
one or two business logic statements. On the other hand, there are business decisions that
contain many Rule Families, and many hundreds, even thousands of business logic statements.
Some Rule Families may contain as few as no condition columns, or perhaps only a single
condition column, while others may contain many condition columns. The value of a Decision
Model must take into account its complexity and the ease with which its business logic is
understood and managed.

THE IMPACT OF BUSINESS DECISIONS ON A BUSINESS IS A FUNCTION OF ALL

THREE CHARACTERISTICS.
Each characteristic measures a different aspect of a business decision and assists in
maximizing the usefulness of the corresponding Decision Model. The following sections explain
how the three categories determine the business value of business decisions and Decision
Models as follows:
 The operative context - determines how a Decision Model solves a business problem
or opportunity.
  The volume-based economic impact - provides insight into the financial value of the
Decision Model.
 The business logic complexity - assists in understanding the cost of developing and
managing a Decision Model.

CHARACTERISTIC #1: ITS OPERATIVE CONTEXT

The definition of business decision given earlier stated that
business decisions - are conclusions arrived at through business logic, and hence derived
from facts.
 Yet, not all business decisions made in the course of business can be made based on facts
alone, because sometimes the facts are just not available. When events do not follow an
expected course—due to opportunity, crisis, or unforeseen events—decisions must be made
in the absence of desired facts.

 These decisions have to be made by relying on the decision maker’s intuition or

expertise instead of some or all of the facts.
 These types of decisions are referred to as “pattern based” or “event based.” To the
extent that they are not fully fact based, they cannot be fully modeled by a Decision Model.
However, in these circumstances the Decision Model still plays an interesting and important
role.

The concept of operative context - is explained in “The Cynefin Framework”*

(Snowden and Boone, 2007).
 This is a “Leader’s Framework for Decision Making”, formulated by David J.
Snowden and Mary E. Boone.
 The framework defines an operative context in which business decisions are made,
representing these operative contexts as a continuum of “contexts.”

Snowden and Boone define four different operative contexts for business decisions based on
the input to them, as follows:
 The Simple context, where the input consists of “Known knowns.”
- All parties share an understanding of the facts required to make a decision.
 The Complicated context, where the input consists of “Known unknowns.”
- This is the domain of experts, where there may be a need to work in unfamiliar
environments. More than one correct answer may exist for the same input.
 The Complex context, where the input consists of “Unknown unknowns.”
- This is when there is incomplete data, and there may be no correct answer for some or
all inputs.
 The Chaotic context, where the input consists of “Unknowables.”
- Here, the fabric of order is more than disheveled; it is rent. The events of September
11, 2001, is the example given by Snowden and Boone as requiring a series of
decisions in the chaotic context based on unknowables. The financial panic of
September 2008 may be another such case.

The first two contexts, the Simple and Complicated

- are in the Ordered Domain,
whereas the Complex and the Chaotic
- are in the Unordered Domain.

Although the Decision Model is obviously quite appropriate for decisions occurring in the ordered
domain, it also proves useful in the unordered domain.

A BUSINESS DECISION IN THE SIMPLE OPERATIVE CONTEXT

In the Simple context, - the “known knowns” form the natural foundation of a Decision
Model because not only are the fact types known, the fact values* are also known. In
other words, it is possible to identify the column headings for Rule Families, as well as
the condition and conclusion expressions for each business logic statement or row.
In this context, the parties come to agreement on the Decision Model, deploy it, and
measure it against relevant business objectives. Business decisions in this context are
typically those that guide business processes. The sequence of tasks in the business
process is purposely designed to ensure that the facts necessary for making the business
decision are available when the business decision is to be made.
The usefulness of the Decision Model in this context is that it provides a shared
understanding of the logic behind the business decision, ensures a complete and accurate set of
logic, and enhances the agility of the corresponding business processes.
A BUSINESS DECISION IN THE COMPLICATED OPERATIVE CONTEXT
In the Complicated context, - the business decision operates with “known unknowns.” In
this situation, specification of the fact types or fact values requires expert advice
and assistance.

Typical of business decisions in the Complicated context is that the fact types (i.e.,
column headings), are known, but there may not be sufficient information to populate the
condition or conclusion cells for those headings. The usefulness of the Decision Model, then, is
that it highlights the unknowns that help to identify the kind of business expertise needed. When
and if those experts are included in development of the Decision Model, it will represent content
that is shared from expert to nonexpert.

Often a business decision operating in a Complicated context may need to allow for
more than a single solution for the same input. If so, the usefulness of the Decision Model is that
it provides a clear understanding of those multiple solutions and simplifies the comparative
analysis among them. Business decisions in the Complicated context sometimes provide
guidance for business processes, but are more likely to do so in Complex Event Processing
(CEP),* and in other event- driven contexts.

A BUSINESS DECISION IN THE COMPLEX OPERATIVE CONTEXT

In the realm of the “Unknown Unknowns,” the Complex context, - the Decision Model is
more difficult to construct. The fact types are as yet unknown.

There may be no clear solution for some or all inputs. Experimental solutions may be
necessary. The usefulness of the Decision Model is that it is an aid for inventing solutions.

Business decisions in complex contexts are found in fluid circumstances such as

business mergers and acquisitions, disruptions in the marketplace, major changes in regulatory
environments, unexpected litigation, and similar events.

A BUSINESS DECISION IN THE CHAOTIC OPERATIVE CONTEXT

The complexity and unknowns in the Chaotic context - are beyond those even of the
Complex context. In the Chaotic context, the business needs to make decisions in the
realm of “unknowables.” In this operative context, there is little applicability for a
Decision Model that addresses the entire business decision. However, the Decision
Model still plays an important role.

The Chaotic context is improved by finding areas within the chaos that can be
separated from the chaos and developed into an area of relative order. It is in these areas that
the Decision Model proves useful in bringing order to the chaos.

In the Chaotic context, the Decision Model emerges as the primary business construct
by which business leaders carve out islands of order from chaos and develop Decision Models for
them.

Table 3.2 summarizes the usefulness of a Decision Model based on the operative
context. The next characteristic of a business decision that aims to contribute to the health of a
business is its volume-based economic impact.

CHARACTERISTIC #2: VOLUME-BASED ECONOMIC IMPACT.

Its Volume-Based Economic Impact - Taylor and Raden (2007) state that a key factor in
the categorization of a business decision is the number of times it is made in a given
period, such as per day, month, or year.
 For example, a discount decision for customer orders may be made many times a day
based on the quantity of daily customer orders. On the other hand, a business decision to
upgrade the product’s current features may be made as infrequently as once a year or less
often. In general, operational decisions, those guiding core business transactions on a day-to-day
basis, are made in the highest volumes. In contrast, the business decisions made at the strategic
level of the business are typically the lowest volume of frequency.

An example of a strategic-level business decision is a merger decision which, in some

businesses, may be made only once in an entire business lifetime, if ever. An interesting
observation about business decision volume is that it often occurs in inverse proportion to its
individual economic value. For example, although a customer order discount decision may be
made many times a day—perhaps in some businesses thousands of times an hour—the
economic value of each by itself is relatively inconsequential to the enterprise as a whole. The
opposite is also true. A strategic business decision may be made only once in a long period of
time, but a single strategic business decision may have a profound economic value. The overall
economic impact of a business decision to the enterprise is the product of its volume and its
individual economic value. So, in the discount example, the value of the discount decision taken
as a whole is the product of the inconsequential impact of its single execution and the
thousands, perhaps millions, of times that it is invoked in a given time frame.

TABLE 3.2 THE USEFULNESS OF THE DECISION MODEL BASED ON OPERATIVE

CONTEXT

Operative Characteristic Typical Usefulness of the

Context s of Inputs Characteristics of Corresponding Decision
Corresponding Model
Decision Models
Simple Known  Known fact types  Delivers and deploys
knowns  Known fact agreed-upon and shared
values business logic
 Related to  Ensures complete and
business process accurate business logic
models  Is measurable against
business objectives
 Enhances agility in
business processes

Complicated Known  Known fact types  Identifies unknown fact

unknowns (not always) types and values
 Unknown fact  Ascertains areas
values needing specific
 More than one expertise
possible solution  Delivers business logic
for the same shared from experts to
input nonexperts
 Related to  Provides clarity to
complex event available solutions and
processing simplifies comparative
analysis
 May reduce complicated
business decisions to
simple ones
 Expedites changes when
events dramatically
change business
conditions
 Operative Characteristics Typical Characteristics of Usefulness of the
Context of Inputs Corresponding Decision Corresponding Decision
Models Model

Complex Unknown  Unknown fact types  Aids in developing

unknowns  Unknown fact values solutions
 Unclear solutions for  Delivers business logic
some or all inputs in a form amenable to
 Related to fluid iterative development
circumstances  May reduce complex
decisions to
complicated or simple
ones

Chaotic Unknowables  Any of the above  Aids in delivering order

 Related to crisis where order can be
 Applies to islands of envisioned
order within chaos

Thus, a business decision on one side of the volume continuum may be as important to
the economic health of the enterprise as one on the other side. Indeed, Taylor and Raden point
out that while a great deal of support for strategic business decision making is available, very
little is provided for those operational business decisions that are made daily in the business,
and that this is at the peril of the enterprise. The Decision Model provides tangible support in the
management of operational business decisions—an area where such support is lacking.

Taylor and Raden (2007) - highlight the importance of operational decisions and the need to
better manage them in “Smart Enough Systems Manifesto.” They argue as follows:

 OPERATIONAL DECISIONS ARE IMPORTANT.

 Organizations are perceived through the lens of the decisions they make.
 Lots of small decisions add up.
 All decisions an organization makes should be managed as though they are
deliberate.
 OPERATIONAL DECISIONS CAN AND SHOULD BE AUTOMATED.
 High-volume, operational decisions especially can, and should, be automated.
 Traditional technology approaches will not succeed in automating decisions.
 The overall effectiveness of automated decisions must be measured, − tracked,
and improved over time.
 Taking control of operational decisions is increasingly a source of competitive advantage.

As discussed, operational decisions are primarily decisions made in the simple context, and are
—or should be—intrinsic to an organization’s business processes.

Table 3.3 summarizes the value of a Decision Model based on the management level and
volume-based economic impact of its corresponding business decision.

Table 3.3 indicates that the operational business decisions that are the subject of Taylor and
Raden’s book fit into the simple operative context. Hence, the corresponding Decision Models
deliver the same values as any Decision Model operating in that context.

Table 3.3 highlights the fact that strategic and tactical decisions referenced in Taylor and Raden,
although of high economic value, are likely to span operative contexts. Thus, the value of their
corresponding Decision Models is the same as described in Table 3.2 for those contexts.

For operational business decisions, Table 3.3 notes that the Decision Model is traceable
to business metrics. This is especially helpful for those business decisions that, if not done
properly, result in loss of revenue, loss of customers, significant fines, reprocessing of
transactions, or the need to reclaim money incorrectly paid out. All of these have individual
economic value that can add up to significant economic impact.
The third way of categorizing a business decision is by the complexity of its business logic

CHARACTERISTIC #3: COMPLEXITY OF ITS BUSINESS LOGIC

Its Complexity of Business Logic The word
complex - means “a whole made up of complicated or interrelated parts.” (Merriam-Webster
Online Dictionary “complex”)
For a business decision, the whole is its corresponding Decision Model. Its complicated
or interrelated parts are its Rule Families of conditions and conclusions and the relationships
among Rule Families. So, in order to categorize a business decision by the complexity of its
business logic, it is necessary to understand the quantity of Rule Families, conditions,
conclusions, and the relationships among Rule Families.

TABLE 3.3 THE VALUE OF THE DECISION MODEL BASED ON THE

MANAGEMENT LEVEL AND VOLUME-BASED ECONOMIC IMPACT

Manageme Typical Characteristics and Business Value of the Corresponding

nt Level Economic Impact of Decision Model
Corresponding Business
Decisions
Strategic  Low volume  Delivers the same value as its
 High individual economic operative context; see Table 3.2
value
 Spans simple to chaotic
operative context

Tactical  Medium value  Delivers the same value as its

 Medium individual operative context; see Table 3.2
economic value
 Medium collective
economic value
 Spans simple to chaotic
operative context

Operational  High volume  Provides a management approach

 Small individual for the small decisions that add up
economic value to high economic impact
 High collective economic  Provides a starting point for high-
value volume operational decisions to be
 Mostly operates in the automated
simple operative context  Provides traceability to business
objectives and metrics—especially
relating to business decisions that
result in fines, reprocessing, and
reclaiming of money
 Serves to provide competitive
advantage and agility.

The complexity of the business logic of a Decision Model is most often considered
to be high when there is a large quantity of Rule Families (e.g., 10 or more), where each single
Rule Family consists of a very large quantity of fact types (e.g., 10 or more), and many logic
statements in a Rule Family (e.g., 50 to 100s and 1000s). On the other hand, the complexity of
the business logic of a Decision Model is considered low when a Rule Family contains few fact
types (e.g., fewer than 5) or perhaps only a single formula for its conclusion fact type. In
between, these are the Decision Models of medium business logic complexity.
 Some industries operate with very complex Decision Models, where the quantity of Rule
Families and business logic statements is higher than indicated in the previous paragraph and
where the relationships among them are many and complex. This is often true of the insurance
and financial services industries.

Table 3.4 provides a summary of the value of Decision Models when viewed through the
complexity of their business logic. Essentially, for the simplest business logic complexity, the
Decision Model provides a means for documenting, sharing, standardizing, and changing that
logic. As the complexity of the business logic increases, the Decision Model also serves as a
mechanism for discovering that complexity, possibly simplifying it, and positioning parts of the
Decision Model for appropriate automation or for human handling.

TABLE 3.4 THE VALUE OF THE DECISION MODEL BASED ON COMPLEXITY OF

BUSINESS LOGIC
Measure of Characteristics of Business Value of Corresponding
Business Corresponding Decision Decision Model
Logic Model
Complexity

Simple Low quantity of:  Delivers the same value as

complexity  Fact types (columns) Decision Models operating in the
 Business logic (rows) simple context.
 Rule Families (tables)  For Decision Models that are very
 Inferential simple in business logic
relationships complexity, the main value is in
(connections) communicating and standardizing
 Fact types known on the business logic.
 Fact values know
Medium Medium quantity of:  Delivers the same value as
complexity  Fact types (columns) Decision Models operating in the
 Business logic (rows) complicated context.
 Rule Families (tables)  Serves as a mechanism for
 Inferential discovering, gaining clarity, and
relationships standardizing business logic with
(connections) medium complexity, ensuring the
 Fact types known greatest degree of simplicity of
 Fact values known or the business logic as possible.
require human  Serves as a basis for determining
expertise which parts of the Decision Model
can be automated and which
ought to be handled by humans.
High High quantity of:  Delivers the same value as
complexity  Fact types (columns) Decision Models operating in the
 Business logic (rows) complex and chaotic context.
 Rule Families (tables)  Highlights the value and
 Inferential competitive advantage of
relationships strategic human expertise
(connections)  Serves as a mechanism for
 Fact types unknown discovering, gaining clarity, and
 Fact values unknown standardizing business logic with
or require rare human high complexity, ensuring the
expertise greatest degree of simplicity of
the business logic as possible.

CHANGING THE GAME: BPM AND BDM

The previous lesson showed that business decisions made in the simple and
complicated operative contexts are likely to serve as elements of guidance in business
processes. Further, when those business processes are executed in high volume, the business
decisions have a significant economic impact on the enterprise.
HOW THE DECISION MODEL SIMPLIFIES BUSINESS PROCESSES
This lesson introduces the important role of business decisions and the Decision Model
in guiding business processes. First, a common definition of business process is needed.

DEFINITION OF A BUSINESS PROCESS

A business process - is defined as a series of repeatable, defined activities taking place in a
planned sequence by actors (being individuals or systems) within a defined scope of
organization where the tasks add value to a good or a service for a customer.
 Business processes are important; some more than others. A business designs, implements,
manages, monitors, and optimizes them to obtain advantage. The goal in managing
business processes is to provide customers with outstanding products or services, or to
lower costs. In short, improvement in business processes aims to perfect business
performance.
 A business process is wide in scope, an end-to-end chain, rather than a functional narrow
view. So a business process is less concerned with the functional departmentalization
(functional silos) of the organization, than with the breadth of business processes that
deliver value to the customer. So, a business process exists regardless of, and spanning,
the functions of the organization. In this way, the focus is on the value chain of the
organization. The value chain is simply the set of steps by which the business adds value to
the goods or services delivered to customers.

EXAMPLE OF A BUSINESS PROCESS

 For example, within an airline, one business process is the customer’s interaction from the
moment the customer inquires about a flight to that customer’s completion of the journey,
referred to as the “Customer Trip Process.” It involves many steps, including inquiring
about flight times and cost, completing the reservation process, arriving and waiting at the
airport, boarding the plane, and so on until the end of the trip. Of course, this business
process involves many different departments and personnel from the airline. It may also
include the hiring of a car from one of the airline’s rental car partners because wide
business processes may cross the boundaries of the organization to include partner and
customer organizations.

 In addition to spanning multiple departments or functional silos of an organization, business

processes also span a significant period of time. The whole “Customer Trip Process” can
take days, weeks, or months from start to finish. Hence, business processes are “long-
running transactions.”

 To improve business processes, a business designs, implements, possibly automates, and

continually improves them. A whole theory of process improvement has evolved, using
techniques for creating an abstract visual representation called a business process model.
This assists in gaining a shared understanding of the business process among stakeholders.
Today, automated modeling tools for producing business process models can simulate the
business process and serve as a source of building blocks when automating it. There are
many different diagrams that represent parts of business process models, including
flowcharts and activity diagrams, swim lanes and process charts, and process and
functional decomposition diagrams.

 Regardless of which types of diagrams are preferred, most business process models today
do not separate business decisions from the business process. Instead, the business logic of
the business decisions is merged into the visual representation of the business process
flow. This leads to the creation of inefficient processes that are difficult to modify. Such
business processes do not enable optimum business agility even when agility is the main
objective. That’s because such business processes are not decision aware.

DEFINITION OF A DECISION-AWARE BUSINESS PROCESS

This defines a decision-aware business process - as one that is designed to distinguish
between tasks that perform work (i.e., process tasks) and tasks that come to
conclusions based on business logic (i.e., decision tasks). Because a decision- aware
business process makes this distinction, the details behind a decision task are
 separated from the details behind the process task. This separation enables the
details behind a decision task (i.e., business logic) to be represented in a different
kind of model, specific to business logic.

 However, most business processes today are not designed to be decision aware. Taylor and
Raden describe the problem as follows: “Although organizations have automated standard
processes with enterprise software, these operational decisions haven’t been the focus of
investment. They are overwhelmingly made manually or automated poorly, which is a
mistake. Embedding business processes in systems to streamline operations but not
managing and improving these decisions leaves half the opportunities for improvement
untouched” (Taylor and Raden, 2007).

 To manage and improve business decisions, they need to be separated from the business
processes that rely on them. To separate business processes and business decisions, they
must somehow be different from each other in a recognizable way. It turns out that they are
truly different in a very significant way. In fact, the inherent nature of a business process is
very different from that of a business decision. To date, however, this difference has not
been well understood, but the advent of the Decision Model brings this difference to the
forefront.

 Essentially, a business process is procedural in nature, but a business decision is

declarative in nature. However, without a clear understanding of declarative versus
procedural nature, common practice involves creating business process models in which
business decisions are loosely represented as just another part of the business process. In
other words, it is common practice to model business processes and business decisions in a
procedural manner rather than modeling the latter in a declarative manner. This common
practice not only constrains the business decision unnecessarily, it seriously hinders agility
for both the business process and the business decision. Understanding the difference
between a business process and a business decision means distinguishing and preserving
the difference between a procedural versus declarative solution.

DISTINGUISHING A PROCEDURAL TASK FROM A DECLARATIVE DECISION

A procedural solution - specifies how, in a step-by-step manner, something is to be done. So
a business process model is a procedural solution because it prescribes a set of tasks
that are carried out in a particular sequence. The business process model is the
“How” of a unit of work.

A declarative solution,- on the other hand, only specifies what needs to be done, with no
details as to how, in a step-by-step manner, it is to be carried out, because sequence
is irrelevant to arriving at the correct result. A Decision Model is a declarative solution
because it is a set of unordered business logic, not a set of ordered tasks. A Decision
Model is the “What” of a special kind of unit of work. “HOW means saying how, step
by step, the work is to be done; WHAT just means saying what the work to be done
is” (Date, 2000).

 The declarative Decision Model for a business decision should be removed from the
procedural business process so that it can be managed separately in a declarative form.

 Separating business decisions from business process tasks simplifies the business process
model, offers more creativity in organizing the business logic, and delivers the business
logic in a form that transcends technology options. These advantages become clear in the
following examples.
EXAMPLE #1: SEPARATION OF BUSINESS DECISIONS FROM BUSINESS PROCESS
Consider a small piece of a business process model to determine a person’s credit rating.

 Option 1 in the business process models in Figure 4.2 prescribes that first the process
determines a person’s employment history, and then if the result is good, the process next
determines the person’s debt. If the debt is low, the process sets the person’s credit rating
to “A.” However, if the results are bad and high, respectively, the process branches
elsewhere. In this business process flow, the sequence of evaluating the business criteria or
conditions is set in a specific order. The process flow is rigidly specified, not allowing for
alternative sequencing.

 Yet Option 2 prescribes a different sequence that also works, although this process flow
likewise does not allow for alternative sequencing.

 Option 3 offers a significant improvement simply by removing the declarative business

decision from the procedural process flow. It represents a simpler process flow consisting of
only one task. That task combines the whole previously sequential set of tasks into one
task, denoted as a decision task, behind which a business decision executes in a declarative
fashion. Within the process flow, the decision task looks like any other task but contains a
decision shape within the task box. Option 3 also includes the Decision Model diagram,
which puts the Decision Rule Family in context with its related Rule Families, and Option 3
includes a Rule Family table for the Decision Rule Family. Neither the Rule Family table nor
the Decision Model diagram is embedded in the process flow. They are separate
deliverables, anchored to the process flow by the decision shape.
The Rule Family, - by definition, implies no particular sequence among the conditions to be
tested.
 The Rule Family in Figure 4.2 also indicates via the “?” that there are other possible
combinations of conditions to consider. The Rule Family can contain as many rows as are
needed to reach the correct conclusion. For that matter, it can contain additional columns if
other conditions are needed to determine a person’s credit rating. The Rule Family table
also contains business logic for the logic not modeled in the business process models of
Option 1 and Option 2. These include the adjudication of the credit rating for all values of
person’s debt and employment history other than “low” and “good.” Incorporating these
into the business process model rather than in the Decision Model would have enlarged and
added unnecessary complexity and unnecessary sequence information to the business
process model. To change or add conditions in such a business process model is far more
cumbersome than doing so in the corresponding Rule Family.

IMMEDIATE OBSERVATIONS ARE THAT OPTION 3 IS AN IMPROVEMENT OVER

OPTIONS 1 AND 2 BECAUSE IT
 Allows a much simpler business process model
 Easily highlights all possible combinations of conditions
 Permits changes in the Decision Model without changing the business process model
 Permits changes in the business process model without changing the Decision Model
(supporting the principle of separation of concerns, or teasing apart the ball of mud)

TABLE 4.1 IMPORTANT DISTINCTION BETWEEN BUSINESS PROCESS AND

BUSINESS DECISION
Business Process Business Decision
 Procedural in nature  Declarative in nature
 Consists of tasks connected by  Consists of Rule Families
sequence connected by inferential
relationships (all independent of
sequence)
 Is all about how (step-by-step  Is all about what is to be
sequence to carry out work) conducted (the logic leading from
conditions to conclusion)
 Improvements in business  Improvements in a business
process aim for increased work decision aim for smarter business
efficiency logic
 Represented best in a procedural  Represented best in a declarative
business process model Decision Model

 The differences between a business process and business decision is summarized in Table
4.1.
 Table 4.2 summarizes the disadvantages of not separating business decisions from
business processes.

EXAMPLE #2: A BUSINESS PROCESS MODEL NEVER REVEALS ALL BUSINESS

LOGIC
 The previous example illustrated that business process models that do not separate
business decisions from process tasks bury some business logic in the business process
model itself.
 Example #2 now illustrates that, even when this is so, it is usually impossible to resurrect
all business logic from such a business process model. Figure 4.3 shows two business
process models for determining Food Stamp (FS) eligibility. Option 2 depicts a much simpler
business process model than does Option 1. That’s because Option 1 depicts a sequence of
process tasks that are forced to occur in a particular sequence but for which such sequence
is actually not required.
 The business process model is simplified by removing parts of it that can be represented in
a declarative Decision Model. So, the high-level Decision Model in Figure 4.4 represents the
business decision “Determine FS Eligibility.” Although the Decision Model contains business
logic from several of the tasks in Option 1, namely, FS Eligibility and Children Qualification,
it also contains several Rule Families that are not represented by process tasks in the
business process model: Citizenship Status, SSN Validation, Employment Status, and
Income Qualification. So, mingling business decision logic with process flow as in Option 1
does not necessarily expose all of the business logic in that process flow.

TABLE 4.2 DISADVANTAGES TO BURYING DECISIONS (BUSINESS LOGIC) IN

BUSINESS PROCESSES

1 Forces unnecessary sequence and constraints on business logic

2 Makes changes to business process and business logic difficult
3 Adds unmeaningful complexity to business logic and business process
4 Fails to deliver a visual representation of all business logic
5 Makes governance of business process and business logic difficult to manage
6 Results in business logic and business processes that are not reusable
7 Compromises SOA
The Secret of the Missing Business Logic

 But then where is the missing business logic in business process models like those in Option
1 of Figure 4.3? Apparently, because all of the business logic is not directly visible in the
business process model, some of it must be buried in one or more of the tasks. In fact, it
probably is buried in many places, because some of it may be used in several of the tasks
(which is the case in this business process model).

 So, it is likely that some of the business logic is hidden from view in a procedural business
process model such as in Option 1. Further, it is difficult, if not impossible, to resurrect all of
it in one visual artifact—not even by drilling into the detail behind each of the process tasks.
Therefore, the business logic is rendered unmanageable.
 On the other hand, the Decision Model, by definition and purpose, resurrects all of the
business logic in one visual artifact. In a populated Decision Model, all business logic is
clearly visible in one place and assists in rapidly and accurately gauging the impact of
suggested business logic changes without reviewing every task in which some portion of
that business logic may reside.

 In a Decision Model, the business logic in one business decision is a chain of inferential
dependencies. The inferential nature of business logic within a business decision makes it
amenable to having its own model, with distinct boundaries and distinct connections to
business processes as needed. In this way, the Decision Model can be viewed, managed,
and executed as one whole set of business logic, as a black box evaluating conditions and
reaching a conclusion.

EXAMPLE #3: SIMPLICITY, PRODUCTIVITY, AND COST SAVINGS

The business process model in Figure 4.5 is based on a real project and is a typical
representation of a business process model when it is depicted without regard for whole business
decisions. Some of the tasks evaluate one condition, so the sequence of such tasks imposes a
sequence on the evaluation of those conditions. Further, the model contains textual annotations
in red representing other business logic (or business rules) that are not represented as tasks in
the model itself. So, the business logic has two different kinds of representations, neither of
which seems optimal. It is not difficult to imagine that producing such business process models is
time consuming, and they quickly become complex. Management of the business logic and
business rules becomes tedious, if not impossible, because some of it is stated explicitly, some is
buried in the business process model itself, and some is probably missing. In fact, this business
process model became so unmanageable that the client gave up maintaining it, the typical result
for models that mix process and logic, because they are not decision aware.

However, the business process model in Figure 4.6 is a reengineering of the one in
Figure 4.5, but with one subtle and important difference: The business decisions are noted
simply by the decision icon within decision tasks. (The notation style difference between the two
diagrams may be ignored.) The corresponding business logic statements or business rules are
nowhere on this diagram. This solution depicts tasks in a prescribed sequence, differentiating
those tasks representing a conclusion carried out through business logic. The detailed business
logic is captured in corresponding Decision Models.

Simplification became obvious when the quantity of business process models for the
entire project was reduced from approximately 25 to 10. Additionally, at least 5 of those
business process models were reused elsewhere because removing business logic resulted in
more generic business process models. Most of the Decision Models also became candidates for
reuse.
The diagram in Figure 4.7 illustrates one of the project’s business process models and how its
Decision Models represent all of the business logic alluded to in the original deliverable but more
completely, accurately, and visually. The business process model, devoid of business logic in
Figure 4.6, contains visible anchor points (e.g., decision shapes within task boxes) for five
Decision Models. One of those Decision Models is shown in Figure 4.7, complete with its Decision
Rule Family and five other Rule Families. Three of those Rule Families appear as Rule Family
tables.
 With these deliverables, the original set of approximately 250 random groups of
business rules was reduced to approximately 20 business decisions totaling 51 Rule Families in
third normal form. It took much less time to create the revised business process models, new
Decision Models, and populated Rule Families, compared to the original “combined” deliverables
that were ultimately abandoned as unmanageable. Further, the “pure” business process models,
Decision Models, and Rule Families in Figure 4.7 were dramatically easier to implement in the
target technology than the original deliverables.

EXAMPLE #4: WHEN BUSINESS LOGIC REQUIRES DELIBERATE SEQUENCE

 There are, however, certain circumstances when sequence of business logic execution is
relevant to arriving at the correct conclusion. In such circumstances, a single Decision
Model does not suffice. The business logic must be divided into multiple Decision Models,
each linked to a separate process task. In this way, the deliberate sequence of the process
tasks enforces the sequence of Decision Model execution because business process models
enforce sequence and Decision Models do not.

 For example, in Figure 4.3 (Option 2) the process flow has two process tasks, each one a
decision task. The first task calls a decision to Determine FS Eligibility, whereas the second
task calls a decision to Calculate FS. The question is: Can this business logic be combined
into a single business decision, and hence a single Decision Model? If so, then the business
process model needs only one decision task, not two. Therefore, why separate the two
business decisions when it is possible to calculate the value of the food stamps and set the
value to zero for those who are not eligible? At first glance, this appears to be a reasonable
question.

 However, there are business circumstances that require separate business decisions and
Decision Models. First, each business decision may be governed by a different group, hence
having separate Decision Models facilitates separate governing bodies for the business
logic. Second, there may be a process task that needs to take place after one conclusion
based on the first conclusion before arriving at the other conclusion (i.e., sequence!) such
as sending a message based on the first conclusion. These are some of the circumstances
that dictate a required sequence (even if the pure business logic itself does not require a
sequence).

 So, the business process model is simplified and collapsed by removing business decisions
from the business process model when sequence is not required within the business
decision logic. Business decisions are therefore represented, not by spreading or diluting
them across process tasks and imposing unnecessary sequence, but by separating them in
declarative Decision Models that sit behind a decision task. In this way, the business
decision is represented as a separate model from that of the business process model, can
be managed independently, and can share its conclusions with business processes. This
 also implies that a business process model is not the optimum way to visualize the business
logic of a business decision.

 The previous examples dealt with operational business decisions (i.e., guiding operational
business transactions). The next example illustrates that separating business decisions
from business process is also powerful when dealing with strategic business processes and
corresponding business decisions.

EXAMPLE #5: SEPARATION OF BUSINESS DECISION FOR NON-OPERATIONAL BUSINESS

DECISIONS
 In 2000, Pitney Bowes, a $7 billion U.S. corporation, made a strategic business decision to
transform itself from a hardware-oriented provider of mail-room equipment to a software
and service-oriented provider. The tactical approach for achieving this was to do so through
acquisitions.

 So, Pitney Bowes purposely formalized its acquisition business process, identifying places
where important business decisions were needed, and creating rigor around corresponding
fact-based business decisions. Over time the acquisition business process and
corresponding business decisions have been improved based on experience, but the
essentials have remained constant. (Nolop, 2007)

 The Pitney Bowes case not only better defined a strategic business process, but also
developed fact-based business decisions that likely are worth tens or hundreds of millions,
or even billions, of dollars and which are only executed a few times a year. By elevating the
acquisition business process to a formally defined business process, and managing the
business decisions separately from the business process, Pitney Bowes were able to place
these strategic business decisions into the simple operative context. This enabled them to
manage the business process and corresponding decisions in much the same way they
manage operational business processes and business decisions.

 Eight years later, the company had made 83 acquisitions, a rate of over 10 per year, while
investing $2.5 billion. The acquisitions have caused revenue to grow by 25%, substantially
accelerated organic growth, and made a positive contribution to net income and cash flow.

 This example confirms what Chapter 3 pointed out—that the management of business
decisions and Decision Models is valuable for decisions made at various management
levels: strategic, tactical, and operational. Of importance is the Complex operative context
and how much that may be reduced to a Simple context. If all the facts can be known, the
business decision functions in the orderly operative context and can be modeled fully in a
Decision Model.

MORE REAL-WORLD EXAMPLES

A great deal has been written about the use of Business Process Management
(BPM) to achieve business objectives and to transform the enterprise into a customer-focused
value chain. Organizations adopt BPM to achieve operational excellence, the epitome of which is
the lowest possible cost of providing a given level (presumably high) of service. This is
commendable, but is often insufficient for the truly competitive organization. After all, squeezing
inefficiency out of a business process is a tactic available to all competitors in the marketplace.
All other things being equal, the most efficient operator will, over time, inevitably be challenged
by a competitor who achieves or exceeds the same level of efficiency. The key differentiation
then becomes the organizational intelligence that operates behind critical business pro- cesses.
This means that competition is not only about how efficient a business pro- cess is but also how
smart its business decisions are.

EXAMPLE #6: WHEN BUSINESS PROCESS EXCELLENCE ALONE IS

INSUFFICIENT FOR COMPETITIVE ADVANTAGE
An excellent example of business process excellence alone being insufficient for
competitive advantage is Dell. Long the dominant PC company, Dell had achieved leadership
through the creation of an outstanding supply chain, coupled with a razor-sharp focus on their
peerless direct sales channel. With every possible cent squeezed out of their cost, and the ability
to deliver attractive product at the lowest conceivable cost directly to the consumer, they
dominated the PC world for many years.

It appears that Dell’s operational excellence, delivered through its hyper-efficient

business processes, failed to provide the feedback that would warn of the changes in the
marketplace. Even when Dell finally read the signals (in some cases only after outside analysts
had pointed them out publicly), its business processes appear not to have had the ability to
make the necessary adjustments.

Most analysts today agree that Dell’s failure could be attributed to several specifics, chief among
which were the following:
 Dell failed to recognize that the general consumer was transitioning to laptop
computers, and so Dell remained focused on desktop computer sales.
 Hewlett-Packard and other computer manufacturers were able to achieve and exceed
the efficiencies of Dell’s supply chain and direct supply model. Competitors retained
major sales in the retail and other channels, giving them the advantage of cumulative
volume through multiple channels as volume is an important element in the cost
calculus of PC manufacturing.
 Dell remained focused on the U.S. market, and failed to recognize that the international
market was rapidly catching up to the U.S. market in size and importance. Again,
competitors were able to build volume in those markets that positioned them to exceed
Dell.
 Dell’s response to its cost and product disadvantage was to continually ratchet margins
down to meet the competition, which it was unable to off- set by any further saving in
operational expense. Eventually, the point was reached when the gross margin slipped
below operational cost, and the model broke completely. The company was forced, from
a weak position and by 2006—very late in the cycle (some think too late)—to rethink
their strategy and reorganize.
 After the first round of reorganization, by early 2008, Dell’s recovery is faltering. It
appears that the company has not been able to fully embrace the essential
transformation they needed to make, and they remains mired in problems.

The lessons from Dell—and many others—are that key business processes must not only be
efficient and consumer-friendly, they must also be smart and agile. Business processes become
agile when declarative business decisions are separated from procedural business process tasks.
Business processes become smart when the business decisions are governed appropriately by
business leaders. Business decision governance involves monitoring the business decision
performance against objectives and recognizing when events occur that raise the operative
context of a business decision into the complicated, complex, or chaotic realm. When the
business leadership clearly understands the business logic behind the business decisions, the
impact of those decisions can be ascertained, and the business can quickly and easily make
adjustments. Then, those decisions are smart and serve as intelligent business levers.

EXAMPLE #7: THE POWER OF SMART DECISIONS BEHIND BUSINESS PROCESSES

To understand the idea of smart decisions, consider the case of Parker Hannifin, a U.S.-based
Fortune 500 manufacturer of motion and control equipment (Aeppel, 2007). In 2001, a long-time
employee, Donald Washkewicz, became president of the company. He very soon realized that
the company’s pricing policies were flawed: managers would simply pitch their prices to earn the
company a gross margin of 35%, considered an industry standard.

Washkewicz - realized—with the help of outside consultants—that this failed to capture the
value that the company was adding for their customers. He began to under- stand that by being
able to price according to a formula that recognized this value, he could transform the company.
It took a major reorganization of the company’s business processes (and a very significant
change of culture) to engineer this capability. The reengineered business processes separated
price, tying it to a system of intelligence gathering to accurately determine the appropriate price
for a given product.*

This was not easily achieved. The old pricing rules were deeply buried, hard- coded in computer
systems, many of them as recently deployed as the 1990s. But management persevered and
eventually succeeded, despite the impediments created by automation itself.

Today, Parker Hannifin - researches the pricing of each and every product it manufacturers,
and tracks feedback on those products from its customers over time to constantly validate
and adjust the price. It works closely with its customers in helping them understand the
price/performance advantage of its products compared to the competition, and constantly
improves the products in ways that can save the customers money.
At the same time, Parker Hannifin refined its other processes, including its “buy side” value
chain (how it purchases products from suppliers), and adopted lean manufacturing. The
results have been dramatic.
Parker Hannifin today is a leader in its industry, in fact a leader in almost any category of U.S.
business. Between 2002 and 2007, its stock rose 160%, exceeding twofold the rise of the
high-tech NASDAQ, and eightfold the growth of the Dow Jones Industrial index. Other key
numbers:
 Revenue —Just topping the $5 billion mark in 2001; analysts estimate $12 billion for
2008.
 Profits —In 2002, the company achieved a profit of $130 million; in 2008, analysts
estimate profit will be closing in, on $1 billion.
 Fortune 500 Ranking—In 2001, the company was ranked 330; in 2008, it was ranked
247.

Parker Hannifin took several of the major steps necessary to create smart processes.
They separated the business decisions from the business processes, tied specific targets of
performance to the business decisions, and implemented the means to measure that
performance. In their case, the focus was the added value of the product to their customer. Not
only did they determine this value, but actively worked with the customer to understand this
value, and worked with the customer to improve the value by improving the design of the
product.
Today, Parker Hannifin continues to monitor the value of products to ensure that the
price remains competitive while delivering a good return to the company. As a company, it has
achieved some of the highest gross profits in their sector of industry.

HOW THE DECISION MODEL ADVANCES BDM

Business processes that meet the criteria of being smart and agile are those for which the
business decisions have been separated from the business process, are represented in Decision
Models, whose impact on the business is monitored, and whose business logic is adjusted to
remain aligned with business objectives

DEFINITION OF BDM
BDM - The practice of managing smart, agile decisions is called Business Decision Management
(BDM). BDM is often referred to as Enterprise Decision Management—these terms can
be used interchangeably.

Figure 4.10 summarizes the three elements of BDM: business motivation, business logic, and
business metrics. These three elements interact to create “smart” business systems, which is
the goal of BDM:
 Business Motivation: This is the general business plan, and the specific business
objective/s within that business plan, that the business decision is meant to implement.
 Business Metrics: These are the measurements and time periods that are set by the
business objectives, and that must be achieved by the business. These metrics are
arrived at in the business planning phase, and may be supported by predictive
modeling techniques.
  Business Logic: This is the logic underlying the business decision that is implemented to
achieve the business objective. This business logic is formulated to best deliver the
business metrics set by the business objectives. The role of the Decision Model in BDM
is to maintain a stable, normalized and complete representation of that logic.

ESTABLISHING BDM
Early adopters of the Decision Model have begun to adopt best practices for BDM. From an
organizational perspective, doing so requires seven important considerations:
 Consideration #1: Recognize business decisions as key business assets that
drive business processes. Success begins with the recognition that business
decisions and corresponding business logic should be made explicit and agreed upon by
relevant business stakeholders.
 Consideration #2: Adopt the Decision Model as a separate, cohesive
representation for the detailed business logic behind business decisions. The
Decision Model provides a graphic representation of this business logic. The Rule
Families provide a means to explore the complete logic behind the Decision Model in a
clear and simple manner. In other words, the Decision Model can be used to achieve a
shared understanding of the business decisions and reach a consensus around them.
 Consideration #3: Define the active role of business people in creating,
changing, and governing the important Decision Models.
 Consideration #4: Use the Business Decision Management Maturity Model
(BDMM) to develop a road map toward implementation of BDM. This will help to
reduce risk and to maximize the return on investment. (See Chapter 20 for details on
the BDMM.)
 Consideration #5: Establish a center of excellence for BDM to outline
methodology, deliverables, standards, and training.
 Consideration #6: Evolve the current state of business requirements to
include Decision Models.
 Consideration #7: Prepare the technology. Select a modeling tool for creating
and managing important Decision Models. Invest in specific technology for
automating selected Decision Models. Consider technology to provide the necessary
modeling tools to project performance and analytic tools to track the performance of
business decisions. Start to build the architecture to sup- port BDM.