In a general sense, a condition in which different elements are present in equal or in the correct proportions.
The term is often used in marketing
materials provided by project portfolio
management (PPM) software vendors. For example, a portfolio might be described as "well-balanced" if it contains
different types of projects or projects with different
characteristics. Unfortunately, few metrics have been proposed for measuring project balance. Because there is no generally
measure for quantifying balance (i.e., no measure for distinguishing good
balance from bad balance), there is no available theory for specifying how
balance should be optimized. For this reason, the
term balance is typically applied in a qualitative rather than quantitative sense.
For example, the term balance might be applied to
argue that a project portfolio
ought to contain a mix of projects with different characteristics or
objectives, such as projects that generate near term revenue along with
projects likely to generate revenue in the longer term. The term is also
sometimes used to argue that individual projects ought to be evaluated
against multiple objectives, and that the most desirable projects
necessarily strike a balance between positive and negative characteristics,
such as risk versus reward. Many PPM tools provide
displays intended to convey the mix or differences in the projects that are
contained within a project portfolio, and these can be useful for
developing understanding of the project mix.
A popular business process developed in the early 1990s
by Robert Kaplan and David Norton for translating an organization's mission
and strategy statements into a quantitative system for measuring
organizational performance. Balanced scorecards collect diverse information
intended to "balance" the traditional, but narrow, financial view of
performance. The balanced scorecard is an excellent tool for helping
managers to understand how the organization is performing and helps
translate strategy into action. However, balanced scorecards are not very
useful for prioritizing and choosing projects, though they may be misused
in this regard.
According to the balanced-scorecard approach,
organizational performance measures should be defined in four areas: (1)
finance, (2) customer satisfaction, (3) internal processes, and (4)
innovation and learning for employees. The selected measures are specific
to the organization and are chosen to reflect the drivers believed to most
important to understanding success.
As examples, measures of organizational financial
performance might include return on investment
(ROI), rate of revenue growth, amount of debt, etc. Customer
satisfaction measures might include number of customer complaints, results
of customer surveys, average time to process phone calls, etc. Internal
process measures might include fraction of projects delivered on schedule,
number of units requiring rework, process yield rates, etc. Learning
measures might include number of employee hours spent in training, numbers
of employee suggestions submitted, etc. The selected measures can be
backward-looking, to monitor how the organization has been performing, or
forward-looking, to assess the future impacts of alternative strategies or courses of
levels of performance may be specified for the measures.
Assessments against the measures are typically arrayed on pages or displays
referred to as "scorecards." Balanced scorecards
are being used in a broad range of activities, from product planning to
incentive compensation, and by federal, state, and local governments.
The major weakness of balanced scorecards is that the
approach does not provide a basis for trading off performance on different
measures. In other words, if an organization improves performance on one
measure without degrading performance on any other measure, that is a good
thing. However, if making a change intended to boost performance in a given
area (e.g., customer satisfaction) threatens performance in some other area
(e.g., finance), a traditional balanced scorecard cannot indicate whether
that change is desirable or should be made.
In an attempt to address the above weakness, balanced
scorecards have sometimes been expanded to include a means for aggregating
individual performance measures into a quantity meant to represent the
overall performance of the organization. For example, most commercially
available tools for project portfolio management allow users to define
equations for combining performance measures—A scorecard is defined
for assessing projects, and the various measures on the scorecard are
mathematically combined to provide an indicator intended to represent the
relative desirability of the project. Typically, the form of the equation
is weight-and-add (sometimes, the performance measures are merely averaged,
which, presumably, implies a desire to weight the measures equally).
Choosing projects so as to maximize a weighted sum of
scorecard measures is almost always incorrect (see preferential independence).
Yet, some try to justify this approach by arguing that organizations should
strive for balance in performance
across various areas and measures. However, maximizing a weighted average
does not necessarily lead to balance (if balance means including projects
that address all measures). In any case, the goal of project selection is
to choose projects that create the most value, not balance (however
balance might be defined).
A project that has a high weighted-average performance
score may or may not be a high-value project. How well the weighted-average
score relates to value depends on, among other things, how the measures are
defined, the number of measures within each area and the degree to which
they overlap or "double count," the organization's current level of
performance, and the basic objectives of the organization. It is typically possible
to define performance measures that can be aggregated into an overall
measure of project value, but this requires a different process for
defining performance measures than that used in the balanced scorecard
approach (see multi-attribute utility
A mathematical formula, originally developed in the 18th
century by the minister and mathematician Thomas Bayes, that shows how
probabilities should be updated based on new information. The theorem is
potentially applicable whenever actions are contemplated that would provide
information relevant to decision making, and some project portfolio management software tools employ Bayes theorem.
In its most basic form, Bayes' theorem expresses the
probability of some hypothesis (or outcome) H given that some event E
occurs in terms of the initial (prior) probability estimate of H and
the likelihood (likelihood function) of E. Specifically:
- P(H) is the prior probability of H,
- P(H|E) is the conditional probability H given that event E is
- P(E|H) and P(E|~H) is the likelihood function—it provides the
conditional probability of E if H is and is not true.
For example, in the context of prioritizing tests for oil
exploration, an important question would be the effectiveness of
alternative tests at resolving uncertainty over the presence of oil. In
this case, the prior probability would be the initial estimate of the
likelihood of striking oil at the location, for example 40%. The likelihood
function describes the accuracy of the test. For example, historically, the
test may have signaled oil correctly 60% of the time when oil is present
and incorrectly signaled oil 20% of the time when oil is not present. If
the test is conducted and if it signals oil, then, according to Bayes'
theorem, the posterior probability of oil would be: 0.6 x 0.4/(0.6 x 0.4 +
0.2 x 0.6) = 0.67. If data showed the test to be a more accurate indicator,
for example, if 80% of the time it predicted oil when there was oil and
only 10% predicted oil when there was no oil, the test would have a more
significant impact on posterior probability. With these alternative
numbers, the probability of oil given that the test predicts oil would be
Bayes' Theorem provides a way to quantitatively describe
the scientific method. If alternative
hypotheses or models are competing for our belief, we can test them by
considering the consequences of each. We can then conduct tests to observe
whether or not those consequences actually occur. If a hypothesis predicts
something should occur, and the something is a good indicator for the
hypothesis, then the result strengthens our belief in the truthfulness of
The process of comparing the performance of one's
organization (in terms of how work is accomplished or with regard to
specified metrics) relative to the best
performance achieved by other organizations of similar type or within the
same industry. Benchmarking is useful because it can help identify
organizational weaknesses and opportunities for improvement.
An increase in the level of achievment for some decision objective, for example, as might result from
conducting a project. For any given decision objective, conducting a
project may cause benefits to increase (desirable), decrease (undesirable),
or remain the same. Also, the benefits derived from a project can depend on
the other projects that are conducted (project interdependencies). To
compare benefits across decision objectives it is necessary to measure
benefits using a common unit, such as dollars (obtained by monetizing benefits). This is a goal of
multi-attribute utility analysis.
benefits as usual scenario
Also called the do nothing scenario a reference scenario which assumes that future evolution is an
extension of the current trends.
Also known as the beta coefficient, a measure of
the degree to which the financial return generated from a stock, or
portfolio or financial securities is correlated or tends to follow the return
generated by the market as a whole. A beta of zero means there is no
correlation, the security moves independently of the market. If the beta is
positive it means the security generally tracks the market, while a
negative beta means that the security generally moves opposite the market
returns. Beta is referred to as an index of the systematic risk due to
general market uncertainties that cannot be eliminated through
Beta can be estimated for individual companies using
regression analysis. In
this method, the company's stock returns (ri) are
regressed against market's returns (rm). The
company stock beta (β) is the slope of the regression line
For example, according to a regression analysis illustrated in the figure below,
the historical beta for the U.S. company Apple is 1.036.
A beta can similarly be calculated for some projects (projects for
which there are historical data on project performance) to measure the degree to which proect returns relate to
financial market conditions. A few specialized project
portfolio management tools use calculations of project betas to
evaluate and compare projects.
A new software product that is nearly fully developed but
not yet thoroughly debugged. Beta versions of commercial application are
often made available to customers for free or at attractive prices,
recognizing that there will likely be numerous problems such as crashes,
errors, inconsistencies, etc.
As it relates to decision making, one of many types of conscious or unconscious errors of judgment that have
been identified that often produce choices not necessarily in the best interest of decision makers.
See joint probability
black swan workshop
A workshop wherein participants attempt to identify
possible events that, if they were to occur, could have a major impact on
success. The term black swan comes from the book The Black Swan by
Nassim Nicholas Taleb, who defines a black swan event as a highly
improbable event with three principal characteristics: It is unpredictable;
it carries a massive impact; and, after the fact, we can concoct an
explanation that makes it appear less random, and more predictable, than it
previously seemed. Black swan workshops are sometimes conducted to support
planning and risk management. Conducting a black swan workshop can help an
organization become better prepared to avoid threats or grasp opportunities
that might not have been previously identified.
An often effective approach for overcoming
reluctance to accept sophisticated models. Initially
a very simple model is developed that provides a crude but understandable analysis is constructed.
After the manager has built up confidence in this model, additional detail
and sophistication can be added, perhaps progressively only a bit at a time.
This process requires an investment of time on the part of the manager and sincere
interest on the part of the analyst in solving the manager's real problem, rather than
in creating and trying to explain sophisticated models. The bootstrapping approach
simplifies otherwise the difficult task of model validating and verification.
A group technique for generating ideas and solving
problems based on encouraging spontaneous contributions from participants.
Brainstorming was popularized by A. F. Osborn in the 1950's by his book
Principles and Practices of Creative Thinking. Although numerous
variations have been proposed, the basic principles of brainstorming are to
encourage the generation of as many ideas as possible, telling participants
to withhold criticisms, using new perspectives to generate unusual ideas,
and combining and improving previously identified ideas. Although there is
limited evidence that brainstorming provides more or better ideas than from
individuals working independently, brainstorming promotes teamwork and can
be an enjoyable experience for participants. Some project portfolio management tools include
features intended to support brainstorming.
A project that
requires or produces a significant or radical change for the business. An
example would be a project aimed at achieving a major competitive advantage
by leapfrogging competitor products. Breakthrough projects may involve new
technologies or processes with the potential of making existing products or
Also called a bubble chart, a graphic display that
uses color coding, shapes, and other visual cues to indicate how items
differ in terms of multiple attributes. Project portfolio management tools often include
capability to display bubble diagrams wherein candidate projects are represented by the bubbles. The x
and y axes represent key project attributes, and bubble size and/or color
indicate other attributes. The displays are used for portfolio mapping, and indicate the
distribution of available projects across various dimensions. Oftentimes,
tool vendors argue that the bubble charts help suggest projects to add or
remove from the portfolio in order to improve portfolio balance, although it is
rare to find explanations for how to measure good versus bad balance on a
Sample Bubble diagram
A weakness of bubble charts is that executives often find
them complex and not very helpful for project selection. Although often claimed to
support decision making, bubble charts are informational displays rather than decision models.
Also called outcome bundle, a shorthand way to refer a specific outcome characterized by
attributes. For example, if N attributes are used to describe an outcome, an outcome
bundle is specified by
specifying the outcomes
x1,x2,...xN for each of the N attributes.
A structured proposal for a project or other business investment intended
to support the decision of whether to undertake that investment. A business
case explains why a project is required for the business and what the new
product, service, or project outcome is going to be. Typically, a business
case will describe the consequences of doing versus not doing the project,
forecast cash flows, and present financial summary metrics, such as the
project's estimated net present value
(NPV) and return on investment (ROI).
It may also present a cost benefit
analysis for the project and identify major project risks and upside
opportunities. A business case model is a model, often implemented
as a spreadsheet, for automating the preparation of business cases. The
model makes it easy to input alternative cash flow scenarios, for example, optimistic, most
likely, and pessimistic scenarios, while showing the impact on project NPV and/or
computing ENPV, along with other project
Also called department, division, or
functional area, an element or segment of a company (such as
R&D;, production, marketing) responsible for a specific business
function. Business units typically are identified on the organizational
chart and are under the domain of a manager. Business units are often
viewed or analyzed in isolation, for example, a project prioritization system may be developed
so as to apply to the projects conducted and
managed by a specific business unit.