Outcome-Shaped Organizations
To solve a problem, you can't be built to last.
If you’re a brand called Tide, and you sell laundry detergent, your job is to get a bottle of it onto the shelf at the Walgreens on the corner of 13th and U, and then to get another one there next month, and the month after that, in roughly the same condition, at roughly the same price, with roughly the same brand experience, from now until the end of time.
Almost everything about how you operate flows from those assumptions. You want consistent processes, durable roles, learning that compounds across cycles, an org chart that doesn’t crater when one person leaves, a finance team that can forecast the next quarter, and a brand team that protects a vibe you’ve spent decades building. You measure your performance against sales growth, market share, and net margin. Your whole shebang — your strategic plan, your OKRs, your annual budget, your performance management system — is built around the fact that what you want to do is deliver something, repeatedly, consistently, well.
If you run a soup kitchen, the same sort of thing is true. A very large percentage of people who come in on Wednesday will be hungry on Thursday. To solve for that, you want predictable supply, predictable staffing, a kitchen that runs the same way regardless of who’s volunteering, and a donor base that mails you the same set of checks every fall. You measure your performance against the meals you serve, week over week. The shape of your work and the shape of your organization match each other, and they should.
Now consider what happens if your job is to eliminate a disease.
Almost none of the previous logic pertains. You actively don’t want consistency; you want the willingness to abandon a strategy that isn’t working at month eighteen, and the agility to try something else by month twenty-one. You don’t want durable roles; you want different combinations of people feeding into different dimensions of the work at different stages — surveillance, then strategy, then regulatory coordination, then deployment, then mop-up. You don’t want learning that compounds inside one organization; you want learning that flows out across an entire field of actors (most of whom don’t report to you). You don’t want sales growth as your metric, or for that matter anything rooted in what you yourself are doing; what you want is the absence of cases — a metric that points at a state of the outside world.
The cascade of assumptions built into the commercial firm-form — and into nearly every organization the social sector inherited from it — is correct for delivery work, including a great deal of valuable social-impact delivery work. But it is not correct for outcome work. Running an outcome-accountable organization with delivery-accountable habits keeps you married to the wrong design.
Which is one reason a recent piece by Anastasia Gamick is so exciting.
Anastasia is a co-founder of Convergent Research, the studio that incubates Focused Research Organizations — purpose-built, time-bound nonprofit research outfits designed to attack scientific bottlenecks that fall through the cracks between universities and startups. Nearly a dozen of them now exist. She has been building this kind of organization for years, and the piece reflects that; it reads like someone reporting from inside the work, not theorizing about it from outside the fishbowl.
Her piece, Working on vs. Solving, draws a binary that names the design problem directly. Working on a problem is what delivery-shaped organizations do — funding research, publishing roadmaps, building fields, running programs, coordinating grantees, year over year. Solving is what outcome-shaped organizations do. It means picking a finish line that is a state-of-the-world rather than an activity, and owning the imperative to reach it. The Carter Center’s Guinea Worm Eradication Program reduced Guinea Worm cases from 3.5 million annually to fewer than 20 because Donald Hopkins, with Jimmy Carter, picked a finish line of zero and held himself accountable to it for decades. The Montreal Protocol banned CFCs in 1987 and the ozone layer is on track to recover by the second half of this century because the treaty bound its signatories to a phase-out timeline rather than to a research agenda.
What both examples share is a person or a structure that owns the outcome. Anastasia’s piece proposes a piece of vocabulary for the kind of leader who could play that role — a general manager — and lays out the five tools such a person would need to wield: grants, investments, new organizations, advance market commitments, and policy work, all simultaneously, with the flexibility and capital and time to actually finish.
The point is that — far from simply representing a clever way of doing things — this ‘general manager’ approach is a design solution to a structural problem that makes adopting an outcomes orientation difficult under typical circumstances. The piece is downstream of an earlier essay by Nan Ransohoff, which makes the case that there’s a category of important problems for which no specific person feels accountable for the whole thing, and proposes the general manager as a solution to this state of affairs.
They’re not the only ones to have noticed.
Around the social sector, an entire class of organization has been quietly forming over the past two decades. Their daily operations share very little in common. What they share is a posture. They each reject the central assumption that an organization exists to deliver a defined activity, and instead organize themselves around what it would take to deliver an outcome.
Adam Marblestone — Anastasia’s collaborator at Convergent — has been arguing for and building Focused Research Organizations, time-bound nonprofit research outfits structurally committed to dissolving when the work is done. Michael Kremer’s advance market commitment mechanism — a way of getting suppliers who don’t yet exist to come into being — has been deployed by the Gates Foundation against pneumococcal disease, by the U.S. government during Operation Warp Speed, and by Nan herself at Stripe through Frontier’s $1 billion commitment for durable carbon removal. DARPA’s program-manager model — tenure-limited managers, expiration dates on the badges, three-to-five-year programs that end when they end — has been refining a version of this for sixty-five years. Funder collaboratives like Blue Meridian Partners and the Audacious Project pool capital across philanthropies precisely so no single funder’s institutional constraints set the ceiling.
The institutional shape civil society inherited is mismatched to much of the work civil society is now doing. As this structural problem becomes increasingly explicit, design responses like these keep emerging across contexts. But they haven’t yet fully coalesced as a formal class.
My collaborator Ben Marshall and I wrote about another potential instance of this class in the Stanford Social Innovation Review this past February. We tried to make the case that a ‘creative paradigm’ for social value — organized around time-bound projects, auteurial vision, and discernment-based evaluation — carried proven success in other outcome-oriented contexts, and offered an alternative to an ‘institutional paradigm’ optimized for delivery.
This argument generalizes. What we see over and over again across the sector is that the traditional firm-form serves as a kind of gravity well, pulling organizations into a delivery posture even when their stated mission is something a delivery posture categorically cannot achieve.
—
How does this ‘gravity well’ work?
The tax code is part of it. Treasury Regulation 1.501(c)(3)-1(b) — the “Organizational Test” — requires a nonprofit’s articles of incorporation to “limit the organization’s purposes to one or more exempt purposes” and “not authorize any activity that doesn’t further those exempt purposes except as an insubstantial part of operations.” Note activity as the unit of analysis, and to a lesser extent purpose (rather than e.g. goal or outcome) as the organizational imperative. This then allows IRS Form 990, filed annually, to go and audit actual programming against that stated purpose — creating a scenario wherein substantial mission drift is a paperwork ordeal at minimum and a tax-status risk at worst.
Such legal architecture locks each organization into a purpose-specific ‘groove’ from its first day of existence.
But the much bigger part is everything that follows from the tax code’s premise: that such an organization is the kind of thing that does a defined activity, repeatedly, in a measurable way. Foundation program officers are reviewed against discrete attributable bets. Donor reporting expects activity-level outputs. Strategic plans cover three-to-five years and typically assume stable program areas. Job descriptions describe ongoing functions. Annual reports tell stories about what the organization did, not about what changed in the world. The year-over-year budgeting process assumes steady-state operations. Even the language people use about their own work — we run a program, we deliver services, we maintain a presence in the field — encodes assumptions that fit a delivery operation and don’t fit anything else.
Again, none of this is per se bad. Lots of organizations need to consistently deliver programs. I am glad legal aid clinics deliver legal aid services. The key is just that when you’re trying to scope a new organization or shift an existing organization to mobilize around some kind of outcome, it’s difficult to shrug off all of this inheritance. You can push back against all of it, but you have to do so actively, every day, against every default that surrounds you.
How? There are options.
Lewis Bollard’s farm-animal welfare work at Coefficient Giving has directed over $130 million toward a single outcome — animals affected by improved welfare standards — by treating that number as a portfolio problem rather than a programmatic one, funding corporate cage-free campaigns and welfare science and alt-protein companies and movement-building in over 20 countries, with the weight shifting across these moves as the binding constraint shifts. Frontier, mentioned earlier, doesn’t fund carbon-removal companies the way climate philanthropy ordinarily would; it signs binding offtake agreements that let traditional capital flow into companies that wouldn’t otherwise be fundable, then publishes its contracts open-source so the rest of the market can use them. Cultivarium, a Focused Research Organization at Convergent Research, is a five-year nonprofit startup making 170,000 microorganisms accessible to synthetic biology — recruiting staff for the duration, releasing tools open-source to the field, and dissolving when the work is done. Convergent has launched many others on the same template, including E11 Bio, which is building an imaging system to map brain wiring at a hundredth of current cost.
In each case, the move that mattered was a rejection of the default cascade of assumptions — about what consistency means, what good operations look like, what success is measurable as, what makes a leader effective, what the relationship to time should be, and what the organization is built for — coupled with the construction of new, non-obvious mechanics fitted to whatever is necessary to actually enact the outcome. These aren’t just nice, one-off stories. They’re templates for solving entire classes of problems.
—
So, how to refer to an organization built this way? The phrase that’s been useful in my own thinking is focused mission organization (FMO). An FMO, like an FRO, is a nonprofit organization with a defined finish line — but where an FRO drives toward a research imperative, an FMO drives toward a state-of-the-world outcome. The FMO inherits from the FRO the time-bound posture and the willingness to plan for one’s own obsolescence. It generalizes the FRO beyond science bottlenecks to anything outcome-shaped — policy actuation, market shaping, disease elimination, civic infrastructure. Freedom to Marry, which won federal marriage equality and then dissolved on schedule, is a canonical American example of an outcome-driven project built this way.
But the particular phrase isn’t important. What’s important is that the category exists, and that recognizing the category gives practitioners a place to locate their work and gives funders a class of bet to recognize. Right now, an organization built like Frontier or Convergent or Freedom to Marry finds itself in a bind. It says it’s a nonprofit, and the wrong cascade of assumptions kicks in. It says it’s a “campaign” or a “fund” or an “initiative” and a different, distinct stack of wrong assumptions kicks in. The absence of a category leaves practitioners reinventing their self-conception every time they try to explain their work.
Nan and Anastasia have given the field the general manager as a piece of conceptual infrastructure — a phrase practitioners can pick up, that funders can recognize, that lets a shared vocabulary form. Marblestone gave it the focused research organization. Kremer gave it the advance market commitment. Each idea is a foothold against the default cascade of assumptions, a little flag planted in territory that mostly doesn’t yet exist.
There should be more flags.
—
The thing I most want a reader to take from this cascade of reasoning is not that any particular one of these architectural innovations is the answer, and we should all drop what we’re doing and sprint over to our funders shouting from the hilltops that we are now building one of them. It’s simply that there exists a meaningful new category, and the category is not yet well-populated. As both funders and operators, that gives us an opportunity.
If you work in civil society, at least some of the assumptions built into how your organization runs — or what your organization supports — could be misaligning your work with the outcomes you say you want. The strategic plan, the job descriptions, the performance reviews, the way you talk about what you do at the all-hands — any one of these could have fundamental misalignment baked in. Many of these norms feel necessary because they look like how a serious organization operates. And they feel that way for a reason: they are how a serious delivery organization operates. The sector has had one primary model in front of it for so long that pursuing an alternative can feel more unprecedented than it actually is — almost reckless, an indulgence for the sake of change.
But building this kind of organization isn’t reckless. It can be risky, but that’s different — most real work is risky. The question is just where we embed the risk. The risk of novel strategies and structures is that they’re unproven and may not work. The risk of the default setting is that we know it cannot work for outcome-shaped problems.
Nan and Anastasia are building organizations whose shape matches the outcomes they’re accountable for. So are Marblestone and Kremer and Bollard and the people behind every advance market commitment and every FRO and every patient funder collaborative. The work of the next decade in civil society is to build many more of them, and to build the resourcing infrastructure capable of recognizing them as a class.
The institutional shape we inherited is not the only shape available. It is one shape among many — the one we built defaults around, the one our tax code anticipates, the one our boards know how to govern. Others have to be made. They will not arrive on their own. And the more they integrate themselves into the existing ecosystem and architecture, the less stridently that existing architecture will push back against and ultimately reject them.
So, let’s talk about them. Evangelize them. Fund them. Sit on their boards. Write the contracts that let them work. Build the metrics that let them be evaluated honestly. Treat them, in other words, not as curiosities or novelties or even valuable one-off solutions, but as institutional innovations allowing us to deal with a poison pill baked into the heart of our social problem-solving architecture.
The category exists already. The question is how quickly we open our eyes to it.