Identifying Goals and Defining Criteria

Key ideas#

The goal mess#

By analogy with the problem mess, Tarasenko calls the set of all stakeholders’ desired futures the “goal mess”. Any move from the problem mess toward the goal mess is an improvement.

Four reasons stated goals differ from real ones#

1. Goal substitution. Experts impose their own view: the operation was a success but the patient died; an award-winning building is unfit to work in. Remedy: involve the stakeholders themselves; do not speak for them.
2. Confusing ends and means. Goals form a hierarchy in which each item is a means for the level above and an end for the level below, so it is easy to pick the wrong level. Ask “What do you need this for?” In one of Tarasenko’s cases, a request to choose a site for a new hospital turned out to serve the goal “improve medical care”, better met by upgrading existing facilities.
3. Incomplete goals. People forget or underrate part of a composite goal. In another case, a merger met its stated efficiency criterion but was blocked because nobody listed the goal of keeping tax revenue for the local budget. A warning sign: goals described in fewer professional “languages” than the stakeholder actually uses.
4. Inability to express a goal. People may not know what they want. Offer a real choice among concrete options, so the goal shows in what they pick, or derive it downward from values they are already aware of.

Idealized redesign#

Ackoff’s technique: imagine the current system vanished last night, and design the one you would want now, ignoring resources and feasibility. Tarasenko reports three effects: it breaks people out of their narrow current view, the play element raises engagement, and a later feasibility check usually shows most of the design is achievable. The main obstacles were self-imposed. The design must still be technologically possible (no science fiction), compatible with its environment and laws, and able to learn and adapt.

Goals of organizations#

Declared goals can differ from actual ones when the organization’s interests conflict with society’s or with its managers’ personal interests; the analyst should notice the gap. The mission sits at the top of the goal tree and should reflect the interests of all stakeholders, not just “grow” or “be more profitable”. Ackoff grades ends as goals (reachable within the planning period), objectives (not reachable within it, but progress is planned) and ideals (never reached, but approachable). Good goal statements are realistic, specific to this problem, and measurable.

Criteria and constraints#

A criterion is a quantitative model of a qualitative goal. You need exactly as many criteria as it takes for the set to model the goal adequately. One criterion is easiest to use but rarely enough: in a city case Tarasenko cites, cost of garbage collection per apartment said nothing about cleanliness and was replaced by measures of disease, garbage fires and rat bites. His practical tip: for each significant attribute, use at least three criteria, covering quality, quantity and time.

Constraints differ from criteria. You trade criteria against each other, but constraints must be met. If constraints make the goal unreachable, raise their relaxation with the decision maker: in another case, an overly strict radar false-alarm limit existed only so that the chief would not be disturbed too often.