Wednesday, April 25, 2012

Principle #5. Information Integrity

Unfortunately, the management in most organizations today is based on implicit goals. Managers with a vague idea and understanding of their goals pass information on to their subordinates in a trimmed and distorted form. The subordinates then take these vague and trimmed goals and try to work them out for their subordinates. Therefore, after the information has gone through two or three management levels, it is nearly impossible to match the lower-level goals and their outcomes to the original upper-level goals. Also, while time passes, situations change and goals also change. Because of this, the discrepancies between goals grow larger and larger. Such non-transparent management is highly ineffective and could lead to unpredictable, unbelievable, and unrelated results. The links between the results and upper-level decisions are missed, and it takes a lot of effort and deep understanding of the situation to figure out what happened and why. Such a style of management is called “Management and Analysis Islands,” and it’s a big mess.

Goal-oriented management offers an entirely different approach. In C2/GOMA goals are defined explicitly. Higher-level goals are detailed, broken down into subgoals, and propagated to executors. When goals are defined the information about their causes and effects, and the relations between higher and lower levels goals, is preserved. Achieved results are also linked to the goals and actions that led to those results. All of this information makes it possible to clearly see why certain decisions are made, how those decisions are executed by lower management levels, and how results relate to decisions at any level.

In order for goal-oriented management to do this, one of the most important and needed C2/GOMA functions is the support of decision reason and consequence links in the information flows. This requirement is addressed by the 5th automation principle--the principle of Information Integrity.

The principle of Information Integrity allows us to eliminate gaps in management and analytical information, and then link that information into continuous flows. Having an absence of information links between management decisions, between reasons and their consequences, makes management non-transparent and significantly complicates management in multi-level systems. The inability to clearly and easily link results to their original decisions breaks the feedback loop. It puts results analysis on the shoulders of human-managers and becomes a serious barrier to increasing the level of automation. Thankfully, the principle of Information Integrity takes care of this problem.

Definition: “Any decision in an organization has reasons and consequences. An analysis, based on the outcomes of previous goals, is used to drive new decisions. The management automation system, in order to help its users make logical and correct decisions, must provide an infrastructure to keep the information whole and complete.”

Key points of the definition:

  • Every management decision has reasons and consequences.
  • Achieved results are linked to the decisions that led to their execution.
  • Execution results are used in making new decisions, changing goals, or setting new objectives (feedback loop).
  • Finally, goals in C2/GOMA have to be defined explicitly. The management automation system has to maintain all formal links between goals, and between goals and their results.

When using the principle of Information Integrity:

  • Look for goals in a management system, even if they are set implicitly. Usually they are hidden under different names and abstractions such as projects, tasks, orders, actions, control signals, etc. Capture these goals in an explicit way by defining them with the 5W principle (who, what, where, when, why).
  • Use the 5th element of the 5W principle--“why?”--to link goals with their super-goals. By doing that at all management levels, you will form the necessary reason-consequence links in the management information flows.
  • In order to improve analysis it is good to have a history of management decisions, i.e. a history of goal changes/manipulations. This will help to relate goals to their original origins.
  • Besides hard reason-consequence links in goals and information, there are soft links related to synergy, conflicts, or the competition for resources. It is important to differentiate between these two types. Hard links exist all the time, while soft links are temporary and depend on a current situation. When goals or situations change, soft links may also change, appear, or disappear, but hard links will stay no matter what. It is important, however, to record all links in order to keep the integrity of the information.
  • Quite often, execution results are just dumped into a big database and used to generate reports. By looking at those reports, a human has to recall a situation in the past and then mentally link the results to goals set at that time. Obviously, such an approach has a lot of issues and becomes a serious barrier toward further automation. Because of that, when implementing a new C2/GOMA system, you have to maintain formal links between results and their original decisions, goals, and the actions that led to them.
  • There is also a simple mechanism using goal vectors in order to implement formal reason-consequence links. When using this procedure:
    • Goals are uniquely identified in a system.
    • When a new goal is formed, higher-level goal identifiers--tags--are stored with it.
    • When results are produced, identifiers of the goals that caused the execution are taken and stored together with the results data.
    • Stored goal identifiers are then used to aggregate the results (SUM, MIN, MAX, AVERAGE) and link them to goals at any management level.
If the management automation system has a common goal repository, which allows it to quickly and easily restore the entire goal chain using reason-consequence links, it is possible to use a truncated goal vector which contains only the identifiers of the lowest-level goals. Otherwise, when goals are spread across multiple components and it is not possible to quickly restore the chain of goals, a complete goal vector shall be stored with the identifiers of all sub- and super-goals, i.e. the entire chain of goals.
  • Finally, it is also important to mention escalations. Having an escalation is a notification that a set goal could not be achieved as defined or expected. Consequently, escalations, like results, have to be linked to their corresponding goals.

By using only explicit goals and keeping information whole and complete--by using the principle of Information Integrity--all the decisions and related results of a management system can be easily seen and understood.

No comments:

Post a Comment