Cost per outcome: the number that should replace the overhead ratio.

The overhead ratio asks the wrong question. It measures the split between programme and support, as if support were leakage. Cost per outcome asks the right one: what did it actually cost to produce one real result. Answering it means doing what any mature cost model does, tracing resources through the activities that consume them to the thing they produce. For a non-profit the thing produced is an outcome, and the activities behind it almost always include some of what the overhead ratio would dismiss: the coordinator who made the programme run, the system that tracked the beneficiaries, the training that kept quality high. Those costs belong in the outcome, because without them the outcome would not exist.

Once outcomes carry their full cost, comparison becomes honest. Two programmes that look similar on an overhead ratio can differ enormously in cost per outcome, and the difference points to where capacity is used well and where it is not. This is the whale curve in non-profit form: a few programmes or sites delivering most of the impact per euro, others quietly consuming resources for little result. The response is never simply to cut; it is to understand why, and to direct funding and design where they do most good.

An illustration

An anonymised example. An education charity reports cost per student using only direct programme spend, which makes it look remarkably efficient. When the full activity-based cost is built, including the coordination, follow-up and data work that actually drive results, the cost per successful outcome is higher but real, and it varies widely between cohorts. The cohorts with more support cost more per student and produce far better outcomes, so their true cost per outcome is lower. The naive number rewarded starvation; the full number rewarded what worked. Figures illustrative; framing follows published full-cost research.