The animated television show South Park made a business of touching nerves, but even its creators reportedly did not expect the furor that roared forth over their Underpants Gnomes episode satirizing common workplace beliefs and practices. The Underpants Gnomes’ business plan lives on (slightly altered) in web culture as a shorthand for inadequate, failure-prone product or service planning:

1. Do something!
2. ??????
3. Profit!

I spent my entire library career wallowing in Step 2. Open access, campus copyright, research data management, digital preservation—no matter what phenomenon my employer threw at me or I myself decided to tackle, the experience invariably boiled down to:

1. Hire someone! or Form a committee! or Launch a website! or Marketing!
2. ??????
3. Profit!

This is certainly nothing I’m proud of. I’ve spent plenty of time in the three years since my library career ended ruminating on how I might have hauled myself out of the quicksand. Worse still, I was hardly the only librarian so stuck. If I had been, I could simply have learned from more profitable examples, or retreated into a corner wearing a dunce cap to lick my wounds with all my self-blame circuits engaged. I know from talking to colleagues and former students, though, that wallowing in Step 2 is the rule, not the exception, for novel academic library services, particularly those aimed at creating nontrivial change in the external environment. Why is that? Can’t we do something about it? I have long been convinced that vagueness in Step 3 prevents better articulation of Step 2. “Profit!” may appear in many guises, such that planning processes must decide on its desired form lest programs flail directionlessly. Take institutional repositories (please). Probably the best-known formulation of Step 3 for institutional repositories is Cliff Lynch’s “a set of services that a university offers to the members of its community for the management and dissemination of digital materials created by the institution and its community members.” This Step 3 begs vital questions that the rest of Lynch’s discussion never answers:

• Is it the university or the library that offers the service?
• Which community members does it serve? (“All of them” is very likely the wrong answer.)
• Which digital materials count? (Ditto.)
• What exactly does the “institution” create, anyway? (Records, certainly, but using institutional repositories for records is usually a bad idea.)

Lynch might well have had a clear ideal end state in mind that he desired the implementation of institutional repositories to lead to, but his definition did not specify it. How were institutional repositories supposed to get there, as software or as services, if nobody could figure out where exactly “there” was? Raym Crow’s Step 3 supplements Lynch’s by articulating a clear ideal end state, in which institutional repositories “provide a critical component in reforming the system of scholarly communication” and “serve as tangible indicators of a university’s quality.” Unfortunately, Crow’s piece leaves us mired in Step 2, completely at sea about how exactly institutional repositories are supposed to accomplish these profligately ambitious goals. A clearly articulated ideal end state is undeniably helpful—without it, libraries planning new services resemble lemmings rushing one after another off cliffs—but it, too, is not enough. A recent blog post from Wikimedian Sumana Harihareswara introduced me to the “theory of change” framework. Keeping in mind that I am still in the honeymoon phase of excitedly learning everything I can about theory of change so that I can refashion my mental furniture around it, I do believe it to be a planning, assessment, and replanning modality capable of foiling librarianship’s diabolically vague Underpants Gnomes. The central method of theory-of-change analysis is simple: work out the desired end state, then work backward through intermediate states and the actions required to reach them until arriving at the current state. For institutional repositories, then, we might start from either of Crow’s end states (though both require clearer articulation; what exactly does a reformed scholarly communication system look like?) and step backward to the present dismal situation. Steps should not be wild stabs in the dark; they should be plausible and plausibly ordered, feasible, and testable. Requiring feasibility should help libraries prevent the all-too-common staffing antipattern in which monumental change is the sole responsibility of one brand-new hire. Requiring testability organically includes assessment in the planning process. It also acknowledges that what seems plausible in the moment may turn out not to be, avoiding the antipattern in which a program or service that everyone knows is not working is neither rethought nor killed off. Working backward from the goal will surface a lot of workflow, communication, and technological complexity that standard strategic planning methods often ignore. Working backward grounds rootless pundits and overenthusiastic idealists (at any level of the org chart) in workplace realities. Rather than the common strategic planning “so thats”—we’ll have an institutional repository so that we fix scholarly communication!—that gloss over the journey between plan and goal, working backward designs a step-by-step map to the desired destination, a map that can be revisited and changed as the terrain changes or assumptions turn out to be incorrect. Having that map simplifies assessment considerably: Are current efforts leading to the intermediate goals posited on the map? If so, wonderful. If not, it becomes time to rethink the map, the efforts, or both. Best of all, theory of change lets librarianship take back the reins of change management from tired old futurist punditry, the endless frustrating rounds of “Future of X” and “Where is X going?” pontification that library Twitter rightfully delights in mocking. Pondering possible futures should not be separated from actions librarianship can take to bring about desired futures, as though librarians had neither power nor influence. As for good old X, it is going nowhere (for any value of X) unless librarianship actively shepherds it there. Theory of change brings human agency—librarian agency, though not exclusively librarian agency—explicitly back into library planning. No more sitting passively around waiting for “the future” to get here. No more frantic reactions to phenomena we should have anticipated. We will instead have the futures we make. I am sure that theory of change can be abused, like any tool, or honored more in the breach than the observance. I do think it has more built-in safeguards against common planning and implementation pitfalls than other problem-solving methods I have seen and used in libraries. I mean to bring it into my next year’s service work, which as it happens will include at least one gigantic and one smaller-scope planning initiative. Ask me in a year how well they turned out!