If one asks what expectations outsiders have of organizations, any empirical investigation would probably reveal that promotion or at least appropriate consideration of one's own interests is expected.
This is likely to be true for individuals as applicants or as affected parties, as customers or as patients, as students in schools or as prisoners in jails. But it also applies to the relationship of organizations to organizations, such as firms to their banks or banks to firms seeking loans; or to the relationship of unions to firms or to the state; or of hospitals to health insurers and vice versa. Since consideration of interests is difficult to specify, and since disappointing experiences in individual cases cannot be readily generalized, a kind of trust develops on this basis; or at least mistrust burdens relations in an almost unbearable way, increases the burdens of information and communication, and is probably only sustained if there are no alternatives, i.e., no market.
With this extrapolation to interests, however, one runs into difficulties when the organization, and this is the typical case, has more than one interest to serve. Then the interest must be extrapolated to the interest, so to speak, and find a more general expression. For example, one can replace orientation to "interests" with orientation to "problems." Only in rare cases will the formula be "impartial". Impartiality may be expected from courts, but not from unions. Even the state should not decide "impartially"; after all, to avoid this, the concept of "public interest" was invented. The problem, then, is to find a term that does not ignore interests, for that is the point, but leaves them unmentioned and absorbed in a more formal expression. This is what the concept of rationality does.
Rationality abstracts from the evaluation of interests. From any interest, because it would not be rational: to grant the public interest (or the "common good" or the "public goods") par excellence priority over private interests. If one wants to express this, it is enough to advocate enthusiastically for the public interest. But this would then be an option that already does not apply to the state in the liberal understanding and can a fortiori not be extended to all organizations. If one wants to communicate on the level of the concept of rationality, one must therefore fade out any evaluation of interests, which of course does not exclude that interests can be recognized, acknowledged and evaluated differently. That organizations should decide rationally is thus the formula in which anyone (and now: the organization itself included) can clothe his expectation without identifying with any interests. It is, if you will, a decision-related equivalent of what in epistemology would be called "objectivity". But what can be meant by "rationality"?