Outcomes Over Activities: What Boards Are Starting to Ask For

A board member at a program review we sat in on recently asked a question that would have been unusual five years ago and is becoming routine now: not "how many workshops did the program run this quarter," but "of the people who attended, what changed for them, and how do we know." The program team had a good answer to the first question — attendance logs, a dashboard, a clean number. They did not have a good answer to the second, and the silence in that room is a pattern we're seeing repeat across boards that have spent years funding activity and are starting to ask what the activity bought.\n\n## Why the shift is happening now\n\nTwo things are converging. Funders — especially larger foundations that have run their own outcome-measurement pushes for a decade — are pushing outcome reporting requirements down into the organizations they fund, so boards are hearing the same question from above that they're now asking below. And a generation of board members who came up through data-literate sectors is less satisfied with an activity count as evidence of impact, because they've spent careers elsewhere being asked to show causal effect, not just effort.\n\nNeither of these is really new as a value. What's new is boards treating it as a governance question rather than a program-design nicety — asking not just "did the program run well" but "does the data model even let us answer whether it worked," which is a question about infrastructure, not about program quality.\n\n## What activity metrics actually measure\n\nActivity metrics — workshops held, households visited, meals distributed — measure fidelity and reach. They answer "did we do what we said we'd do, and at what scale." That's a legitimate and necessary question, and we're not arguing organizations should stop tracking it. The problem is that it's frequently the only question the data model is built to answer, because activity data is what's easy to collect at the point of service delivery, and outcome data requires a beneficiary to still be reachable, still willing to respond, well after the activity that produced the metric has ended.\n\n## What outcome metrics demand of the data model\n\nAnswering "what changed for beneficiaries" instead of "what did we do" requires three things most activity-oriented systems don't have. A baseline captured before the intervention, on the same indicator that will be measured afterward — not a different, more convenient proxy measured at a different time, which makes before-and-after comparison an approximation rather than a measurement. A beneficiary-level record that persists past the point of service, linked across visits and over time, so "changed" can be measured against that specific person's starting point rather than compared to a different cohort's average. And a follow-up mechanism that reaches people after the program's direct engagement with them has ended — the six, twelve, and twenty-four month check-ins that most activity-tracking systems were never built to trigger, because nothing in an activity-logging workflow prompts anyone to look at a beneficiary again once the service was delivered.\n\nNone of these three is a reporting feature you bolt onto an existing system at year-end. They're data model decisions that have to be made when the program launches, because a baseline you didn't capture in month one can't be reconstructed in month twelve, and a beneficiary record that wasn't designed to persist can't retroactively acquire an ID that ties this year's visit to last year's.\n\n## What we tell boards asking this question\n\nThe honest answer, when a board starts asking for outcome evidence a program's current system can't produce, usually isn't that the program failed. It's that nobody built the system to answer that question, and building it now is a data infrastructure decision, not a program adjustment — one that pays off starting with the next cohort enrolled, not the one that already graduated without a baseline on record.