Lee Merkhofer Consulting Priority Systems

Technical Terms Used in Project Portfolio Management (Continued)






























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 accepted mathematical 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.

balanced scorecard

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 action. Target 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 analysis .

Bayes' theorem

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:

Bayes formula


  • P(H) is the prior probability of H,
  • P(H|E) is the conditional probability H given that event E is observed; and,
  • 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 0.94.

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 hypothesis.


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 diversification.

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

regression equation

For example, according to a regression analysis illustrated in the figure below, the historical beta for the U.S. company Apple is 1.036.

Computing beta

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.

beta software

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.

bivariate normal distribution

See joint probability distribution.

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.

breakthrough project

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 processes obsolete.

bubble diagram

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 bubble chart.

Sample Bubble Diagram

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 multiple 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.

business case

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 financial metrics.

business unit

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.