Featured Member Article
Applied Information Economics (AIE): Not just calculate, scientifically MEASURE ITWritten by Soumen ChatterjeeIntroductionThe economic benefit analysis of an Enterprise Architecture Integration program is directly related to the success of an Enterprise Architecture program. However, the benefits of Enterprise Architecture cannot be measured in a meaningful way. Therefore, the result is that Enterprise Architecture is not as widely adopted, as it otherwise would be. This article explains how Applied Information Economics (AIE), a revolutionary methodology invented by Douglus Hubbard of Hubbard Decision Research, fills in those unanswered gaps that act as a dangerous trap for enterprise investment decision makers. This article explains why AIE is so important in helping organizations realize the actual benefits of an Enterprise Architecture Program and all other subsequent projects executed under the hood of this Enterprise Architecture program. Applied Information Economics: A brief overviewApplied Information Economics (AIE) is the practical application of mathematical and scientific methods to the enterprise decision and valuation process. AIE distinguishes itself from all other decision making methodologies by its synthesis of a variety of scientific and mathematical fields which makes possible everything measurable. AIE uses advanced methods from decision theory, portfolio optimization methods, operations research, statistics, and many other areas to solve even the most difficult IT valuation issues. It also uses the same methods that actuaries, financial analysts or statisticians use. AIE can measure any IT benefit or risk including faster communication, better quality data, or customer satisfaction. The tools of economics, financial theory, and statistics are all major contributors to AIE. In addition to these more familiar fields, AIE includes Decision Theory - the formulation of decisions into a mathematical framework - and Information Theory - the mathematical modeling of the transmission and reception of information. It is important to emphasize, however, that even though AIE is a theoretically well-founded set of techniques, it is a very practical approach. Every proper application of AIE keeps the bottom line squarely in mind. All output from the AIE project is in support of specific practical business objectives. Figure 1: AIE Foundation Blocks. Source: Hubbard Decision Research. Used with permission of Douglas W. Hubbard. AIE significantly deviates from the traditional cost benefit analysis procedures which rarely contain real measurements. Paradigm shifts in AIE include following:
AIE thus eliminates all the executive hypes and pseudo measurement illusions with the introduction of real scientific probabilistic measurement. AIE StepsApplied Information Economics steps are summarized in this following table:Table 1: AIE Steps
AIE suggests that instead of choosing a single point value of any set of variables, rather realistic results could be achieved by ascribing the probabilities on them. For this reason, AIE derives the probability distribution graph of the net benefits of a particular type of investment.
AIE uses “Monte Carlo” simulation to develop a graph of the likelihood of each possible net benefit. AIE further builds on typical Monte Carlo methods by the use of calibrated estimates and information value calculations. A “calibrated estimate” is a subjective probability assessment from a person who has been trained to be neither over- nor under- confident in their probabilities. Decades of research in decision psychology shows that training can compensate for an expert’s tendency to be statistically overconfident in their use of subjective probabilities. An overconfident person may say they are 90% confident in a series of predictions where, historically, they were right much less often than 90% of the time. A well-calibrated person, on the other hand, will be right 90% of the time if they say they are 90% confident, 80% of the time if they are 80% confident, and so on. Calibrated probability assessments allow the decision maker to model the current state of uncertainty about a decision in a way that is neither over nor under confident. Once this is accomplished, AIE uses standard information-value calculations to compute the value of further measurements for each uncertain variable in the model. This allows the decision maker to focus on measurements that matter the most. Finally, once uncertainty is removed to the point where further measurement is not economically justifiable, the decision maker makes a risk vs. return decision. IT Investment as an Investment Portfolio: AIE Risk Return AnalysisAIE Uses the method of Modern Portfolio Theory and treats the set of IT Investment as another investment portfolio. Each Investment is analyzed on a Risk Return basis with regard to its contribution to the portfolio. By using techniques from Modern Portfolio Theory, AIE determines whether the uncertainties inherent in a given IT investment decision are acceptable given Risk return profile of the organization. This figure 3 summarizes the Risk Return Analysis (RRA) steps performed by AIE:
The last step compares the investment to the “investment boundary” of an organization. This is a chart that shows how much risk is acceptable for a given return for a particular investment size. Here is an example graph drawn against an IT investment of $2M-$3M plotted against a client’s investment boundary. In this case, risk is shown as a chance of getting a negative internal rate of return (IRR) from the investment (computed from the Monte Carlo simulations). The objective of an investment is to plot below the investment boundary where the risk is more than acceptable given the return. Such a graph would sometimes show that a lower return investment is preferred to a higher return investment if the risk of the first is much lower. This Graph suggests that the 27% IRR investment is actually preferred to the 83% IRR investment. Figure 4: Example Risk Return Profile. Source: Hubbard Decision Research. Used with permission of Douglas W. Hubbard The net result of this process is to present the uncertain value and risks of any investment (like EA) in a way that would be consistent with more sophisticated actuarial and statistical analysis. Not only does this quantify risk in a meaningful way, but it optimizes the measurement process itself by focusing on the highest payoff measurements. The lesson learnt for the Enterprise is that measurement is the key for any investment decision. MEASURE is a 7 step iterative analysis procedure that refines and simplifies any IT decision making process. Let me explain how to overcome this MEASUREment challenge. MEASURE: A simplified capsule for IT Decision makersMeasurement is most important for any correct IT decision. The Purdue Enterprise Reference Architecture (PERA) Master Plan provides an excellent means of economic justification for any decision, classified into three important categories:
However, “a strategic justification accompanying an economic one may assure justification of a project that might well decide whether the company will become a competitive force in the market or disappear from it. It may be that a project that yields little return is justified from a strategic viewpoint because it is a prerequisite for other profitable follow on projects” [PERA]. Therefore, a strategic viewpoint alone should not drive any EA investment plan under any circumstance. Projects ought to be spawned based on the outcome of the economic benefit analysis process and by prioritizing the payback potential. MEASURE is a simple 7 step analysis procedure introduced by the author. These seven steps are as follow:
Figure 5: What to MEASURE? In our day-to-day business, we follow some of these well-known economic benefit calculation processes such as ABC, ROI, NPV and TCO. On the other hand, Value Analysis, Portfolio Analysis and Risk Analysis techniques are popular analytical approaches. Application Information Economics primarily acts as a strategic decision enabler, showing us the direction alongside strategic, economic and analytical triangle. This below-mentioned table summarizes several important observations against traditional CBA versus AIE: Table 2: Applied Information Economics (AIE) versus traditional CBA
While we have already introduced the reader to the MEASURE magic, it is important to know why enterprise is still far away from the benefit of this MEASURE approach. The main reason behind this impinges upon the spreading of Enterprise Architecture and so the benefits of EA. In our Enterprise, the presence of these antipatterns prevents the realization of the art of enterprise architecture. In a similar way, so does enterprise benefit analysis and finding the right answer for the fundamental question of “Why EA?”. Therefore, it is important to briefly become acquainted with these obstacles, which are weeding out practices in the art of the EA. ConclusionIt is a common inversion symptom across the industry that industry prioritises the analysis sequence as follows: Cost, Benefits, Utilization and Chance of Cancellation. AIE unravels this inversion riddle with the recommendation that measurement sequences should be other way around. This article justifies why this AIE recommended measurement sequence is so important and should be started with Risk-Return profile. This article recommends, “To make the best investment decisions, IT management must learn how to compute the value of information, and start modeling utilization and cancellation in cost-benefit arguments. Once IT Management does, measurements will be properly prioritized and the IT measurement inversion will disappear” [The IT Measurement Inversion by Douglas W Hubbard in CIO Enterprise Magazine dated Apr. 15, 1999]. Terminology Puzzles
About the AuthorSoumen is a TOGAF Certified Practitioner, Sun Certified Enterprise Architect and IBM Certified Specialist in RUP. He is an active contributor and reviewer of industry leading Enterprise Architecture, System Architecture and Reference Framework. He has published several technical papers in international conferences, architect journal and Internet based publication portals. He is a regular columnist of a special enterprise architecture column (30000+ regular readership), called MDA Radar, where he discusses different aspects of Model Driven Enterprise Architecture. Soumen is a member of the IEEE, IEEE Standards, ACM, Information Systems Audit and Control Association (ISACA) and IASA. He holds a Masters in Information Technology. He is currently working as an Enterprise Architect for a global leading consulting company. He could be reached at soumenc(at)acm(dot)orgCopyright DeclarationFigure 1, 2, 4 sources are [HDR]. Used with permission of Douglas W. Hubbard AcknowledgementThe author is sincerely thankful to Douglas W. Hubbard for his valuable review comments, which helped to finish this article. The author is grateful for all his materials and suggestions. Reference[HDR]Hubbard Decision Research |
Navigation
