Future-First Evaluation

Future-First Evaluation:
A Three Horizons Approach to Systems Change Assessment

The Problem with Current Evaluation Practice

Conventional evaluation is structurally retrospective. Organized around accountability to past decisions — a theory of change, a funded intervention, a baseline — it measures whether a program did what it said it would do. Recommendations for the future tend to be iterations on the existing paradigm rather than imaginative pivots. This is a reasonable approach for stable, linear interventions. It is poorly suited to systems change work, where the goal is not program fidelity but transformation of the conditions that produce outcomes.

This concept note proposes a Future-First Evaluation method that reorients the accountability object — from program performance to system trajectory — and uses the Three Horizons framework as its primary structuring logic.

The Proposition

Future-First Evaluation asks not "did this program work?" but "is the system we are funding producing new and better outcomes — and is it earnestly trying to?" This reframing has two important implications.

The accountability object shifts from the program to the portfolio and the sub-system it operates within. Individual grants are assessed for their contribution to system trajectory, not isolated for attribution.
The temporal orientation shifts from past to future. Imagination — structured and participatory — becomes a legitimate data source alongside conventional evidence.

The Three Horizons Structure

The method is anchored in the Three Horizons (3H) framework, which describes the co-existence of a dominant system in decline (H1), emerging innovations and alternatives (H2), and a transformed future system (H3). The evaluation proceeds in three corresponding moves.

Horizon 3 — Ideation

The evaluation begins with a structured futures process to surface H3 visions from system actors, including beneficiaries, practitioners, and change agents. Crucially, this process is designed to surface competing and divergent futures rather than converging on consensus. The resulting H3 landscape — with its tensions and contradictions intact — serves as the normative anchor for the evaluation.

Horizon 2 — Diagnostic Sorting

Present-day data — innovations, initiatives, policies, coalitions, narrative shifts — is examined through an H2 lens. Each signal is assessed against stakeholder-defined criteria to determine whether it is reinforcing the dominant H1 system or seeding movement toward H3. This sorting is conducted transparently, with decision rules established in advance with commissioners and stakeholders.

Horizon 1 — System Baseline

The dominant system is documented not just as a context statement but as an active force — tracing which H2 elements it is absorbing or neutralizing, and where its structures are creating path dependencies that foreclose H3 futures.

The Accountability Logic

A systems change funder cannot reasonably be held accountable for transforming a system. They can reasonably be held accountable for the quality of their strategic contribution to a system that is attempting to change. Future-First Evaluation therefore holds funders and systems to account on two related but distinct dimensions.

Portfolio contribution: Is the mix of investments consistent with what the Three Horizons analysis suggests is needed — and is there evidence that H2 investments are seeding rather than reinforcing H1?
System seriousness: Is the sub-system displaying the hallmarks of genuine transformation effort — originality of intervention, divergence of outcomes from historical patterns, and willingness to contest H1 rather than accommodate it?

This second dimension is significant. In complex systems, it may be too early to know whether outcomes are better. But a system earnestly attempting change leaves detectable traces: novel approaches that depart from established practice, outcomes that diverge from prior trajectories, and interventions that create productive friction with dominant structures rather than working around them. Divergence and originality — even where definitive outcome evidence is absent — are treated as meaningful proxy indicators of a system in motion.

What This Method means

Future-First Evaluation makes a distinctive contribution by asking evaluative questions that no existing framework asks well:

Where are present trajectories path-dependent in ways that foreclose imagined futures, even when current indicators look positive?
Which H2 investments are being captured by H1 interests over time — and can we detect that drift before it fully forecloses alternatives?
Is the system displaying the kind of originality and divergence consistent with genuine transformation, rather than sophisticated incrementalism?

These questions are not replacements for conventional accountability — they are additions that make evaluation more honest about what systems change work actually is, and more useful to the funders and practitioners trying to do it.