We've all heard—and many of us have probably invoked ourselves—the admonition "don't let the perfect be the enemy of the good." It's a concept that has a kind of fraught history in library discourse, because it embodies a tension that exists between two conflicting aspects of library culture: on the one hand, we place a lot of value on accuracy, completeness, and quality in the work that we do; on the other hand, we are painfully aware of the limited resources with which we have to work.

The tension between these two realities is sometimes expressed in the form of conflict between those who invoke the phrase "don't let the perfect be the enemy of the good" and those who hear it. Those who use it may be urging the hearers to get a particular task to the point of "good enough" and then move on to other tasks that also need to be addressed. They also may invoke this concept in order to send the message that we shouldn't let our inability to do something perfectly stop us from doing it at all.

In both of the above contexts, I would suggest that this slogan makes good sense and that it stands in explicit opposition to another familiar saying: "If a thing is worth doing, it's worth doing well." Not all necessary tasks are equally important. This means that, in fact, all things that have to be done are not worth doing equally well—some things are worth doing just well enough to get them off your desk or your to-do list, so that you can turn your attention to more important work. In this sense, "not letting the perfect be the enemy of the good" means recognizing the important principle of opportunity cost: since our time and energy are limited, any task we undertake comes at the cost of other tasks that fall by the wayside. If our time and energy were unlimited, it would be appropriate to aspire to do everything and to do it all perfectly, but since both are strictly limited, we have to make decisions—sometimes wrenching ones—not only about what we will and won't do but also about what we'll take the time to do really well and what we'll do less well so that we can move on to projects of greater value and importance.

On the other hand, "not letting the perfect be the enemy of the good" can also be a cop-out, an excuse used by people who can't be bothered to improve their performance or to deliver exceptional service. (I've seen the phrase "don't let the good be the enemy of the great" used in this context.) If we invoke the slogan in defense of laziness ("don't push me to improve my skills") or organizational inertia ("we're already doing enough to help our patrons and our institution"), then I would suggest that we're misusing it.

So the principle of "not letting the perfect be the enemy of the good" has much to recommend it, but (like virtually every principle) it can be applied both appropriately and inappropriately. I'd like to suggest two rules of thumb for applying it well:

First, apply it where the question is, "Does Proposition X result in a net gain in efficiency or effectiveness?" In other words, we shouldn't forgo a program that would make things better just because it won't make things perfect. Sometimes we experience opposition to new ideas on just this basis. Someone suggests that we create an institutional repository (IR) that will make the scholarship produced on campus more fully available to the general public, and someone else—someone who, perhaps, might not want to undertake the significant work it would require—says, "Look, our establishment of an IR isn't going to solve the problem of access to scholarship." (Of course, it won't, and no one is arguing that it would. The argument is that establishing and maintaining an IR would generate benefits equal to or greater than the cost of doing so.) In other words, that a proposed program doesn't create perfection does not mean that it doesn't achieve enough good to be worth doing.

The principle should also be applied where tasks or practices need to be done to a functional level of quality or completeness in order to achieve an important goal but where levels of quality above the functional return little or no marginal benefit. This is the sense in which the principle is often applied in the context of cataloging: How many subject headings does a record really need to have in order for the item to be findable in the catalog? Catalog records need to be as complete and accurate as necessary, not as complete and accurate as possible.

Second rule of thumb: ignore the principle where the question is, "Does Proposition X get us as close to our goals of efficiency or effectiveness as we need to be?" If the proposition in question is designed to get us closer to perfection in an area where something close to perfection is truly needed, then the principle of "not letting the perfect be the enemy of the good" should go out the window. This begs an important question: Are there areas of library work in which we really should be striving for perfection, or for something as close to it as we can get? Yes, I believe so. The personal service that we provide on reference desks; the salience and clarity of library signage; the accuracy and completeness of our financial accounting; our integrity as employees and leaders—all of these, it seems to me, deserve constant attention, may need constant improvement, and allow for little, if any, compromise.

Of course, no matter what our approach to decision-making and prioritization, the constant danger is that we won't make such decisions carefully and mindfully but will instead fall prey to the temptation to give full attention to the issues and problems that are noisiest, that are brought to us by the people with the strongest personalities, that appeal to our personal biases and preferences, or that otherwise present themselves to us in ways that make giving our time and attention to them the course of least resistance. In such cases, the temptation is to let the noisy be the enemy of the important. That will be the topic of next month's column.