Fuzzy cognitive maps (FCM) are hybrid tools of artificial neural network and fuzzy logic systems. One of their major uses is in decision support systems. To arrive at correct decisions, the fuzzy interconnections between attributes are either trained or assigned by domain experts. The network target may be fixed or unknown. In this paper, the FCM has been illustrated in a strategic domain in which the target is fuzzily defined and unknown. The target value is therefore estimated first by fuzzy inference rules for a collection of imprecisely defined attributes and then the fuzzy gradation to which the value belongs is used as the basis to tune the cognitive map The domain concerns a research institution. The target risk is the time by which the projects run-off beyond the stipulated time of completion. The paper shows that certain instinctively chosen membership functions to tune the cognitive map are able to reproduce the belief surrounding the criticality of the domain.