Is It Time to Rethink Evaluation as a Relational Practice?

A year or so on from the High Street Heritage Action Zone (HSHAZ) evaluation, one idea keeps returning — an idea that lies not in numbers or methods, but in the possibility of a different approach. One that recognises that place-based heritage programmes are fundamentally relational — shaped by people, contexts and connections — and proposes that evaluation should be too.

Initiatives like these rarely happen in silos. They draw together architects and conservation specialists, cultural practitioners, youth workers, funders, local authorities, property owners, businesses and residents of all ages — each bringing different roles, experiences and ambitions for what change could look like. AMION’s evaluation reflected this relational breadth too: surfacing new collaborations, creative participation, strengthened partnerships and the many ways people worked together to re-energise their high streets.

Yet evaluation practices more broadly still tend to be quite linear — often centred on capturing predefined outputs and intended impacts, even when the work itself is far more fluid and emergent.

• What if evaluation became more relational?
• What if it carried the same curiosity, creativity and collaboration that animate the initiatives it seeks to understand?
• What deeper learning might emerge if we evaluated in ways that echo how the work unfolds?

Why relationships matter in evaluation

One of the clearest lessons from the HSHAZ programme was that impact flowed through relationships long before it appeared in outcomes. Projects progressed — or stalled — because of trust, communication, shared ambition, negotiation and the ability of people to work together across systems. These dynamics were often invisible in the final outputs, yet they shaped almost everything that followed.

If evaluation overlooks these relationships, it risks overlooking the very mechanisms through which change is made.

A relational approach pays attention to the conditions that enable impact:

• the trust that allows ideas to be tested
• the time it takes to build understanding
• and the small adjustments and accommodations that make collaboration possible

It attends not only to what changed, but to how the work unfolded — the motivations, tensions, connections and decisions that influenced the direction of the programme. These are rarely captured by outputs or indicators, yet they are often where the most meaningful learning sits.

In this sense, evaluation could become less about isolating variables and more about tracing how relationships, processes and contexts come together to create — or constrain — change.

Evaluation as a conversation

Despite the best intentions, evaluation can sometimes feel like something separate from the work itself — a process shaped elsewhere, guided by predefined indicators, and experienced as monitoring rather than supporting.

A relational approach offers a different way of thinking. It views evaluation as a conversation: slower, more reciprocal, and grounded in the realities of those involved. It begins by asking simple but generative questions:

• What matters here?
• Who decides?
• What do we need to learn?

Rather than imposing a fixed set of metrics, relational evaluation invites practitioners, partners, communities and funders to co-shape the questions and methods. This does not replace rigour — it reframes it. The parameters remain, but they are developed with a shared purpose, rooted in local context and aligned with the aims of the people delivering the work, not only those funding it.

When evaluation is approached as a dialogue rather than a verdict, the process shifts. Data becomes more meaningful because it is useful to those who generate it, not just those who receive it. Insights are shared, tested and returned to the work as it unfolds, allowing projects to adapt and learn in real time rather than retrospectively.

In this way, evaluation becomes iterative and alive — enabling responsiveness as the work evolves and offering a shared space for learning rather than an output delivered at the end.

What numbers can’t tell you

Quantitative indicators are essential. They tell us about reach, scale and breadth. They help track change over time and can surface patterns that might otherwise be difficult to see. But on their own, they rarely tell the whole story.

Numbers tell us how many, but not how well or why. They rarely speak to the nature of the changes made, the quality of people’s experiences, or what brought them back — or not. Without this, it becomes difficult to understand the true value of a programme, what should be sustained, or what could be improved.

Much of what matters sits beyond numerical indicators — the shifts in trust, connection, and belonging that shape how people experience change. These are the undercurrents: the intangible, relational elements that hold communities together and often determine whether interventions endure. They are the kinds of shifts that tend to take place between the numbers rather than within them.

A relational approach to evaluation creates room for this deeper learning. It asks not only what can be counted, but what matters to the people involved: what they want to notice, what they want to understand, and what they feel is important to track.

When communities feel part of the interventions shaping their places, the outcomes often carry more meaning. A relational approach to evaluation helps us recognise this, making space for the insights that metrics alone cannot hold.

When accountability sits in the approach

Evaluation often centres on accountability — demonstrating that investment has been used responsibly, that activities were delivered, and that intended outputs were achieved. This form of accountability is important. It supports transparency, stewardship of public money and confidence that programmes are making a contribution where it matters.

But when accountability rests only on outputs, evaluation can narrow. It becomes a record of what was delivered rather than an exploration of what was learned. It can encourage organisations to focus on counting what is easiest to evidence, rather than understanding what is most meaningful.

A relational approach proposes a shift: from accountability for outputs to accountability for process. Here, rigour sits in how information is gathered, who is involved, how insights are interpreted, and how openly people can reflect on what worked — and what did not. When accountability is rooted in the approach — in the care taken with participation, method and reflection — the outputs naturally carry that same coherence. A thoughtful process produces thoughtful insight.

What becomes possible through relational evaluation

Relational evaluation opens up possibilities that linear, output-focused approaches often overlook. One of the most valuable is continuity.

Place-based programmes rarely begin or end within the boundaries of a funding window; they have histories, and they have futures. Partnerships form slowly, practices take time to settle, and some effects only become visible long after the formal programme closes. When evaluation feels meaningful to the work itself — rooted in the questions practitioners are already asking — people are more likely to continue reflecting and gathering insight as projects evolve. A relational approach helps carry this learning forward, so it is not lost when a programme ends or a team changes. It creates a thread of understanding that future practitioners, partners and community members can build on, rather than starting again from scratch. It also recognises that progress is rarely linear — that periods of pause, setback or slow growth are often part of how change takes shape — and avoids drawing firm conclusions based solely on the moment a funding cycle ends.

Relational evaluation does not replace established methods; it expands and grounds them. Quantitative analysis, economic modelling and impact frameworks remain vital tools, but they gain clarity when they are not treated as separate strands and instead understood as part of a wider, interconnected picture. Together, these elements create a fuller understanding — one that helps us see not only what changed, but how change came to be and what may continue to emerge.

For place-based initiatives, this matters deeply. A relational approach can offer a more robust and enduring understanding of change.

Looking ahead

There is an opportunity now to broaden how we approach evaluation — not by replacing established methods, but by situating them within the relationships, contexts and processes that give work its meaning. A relational approach moves evaluation closer to the work itself: a way of noticing change as it unfolds, learning with those involved, and understanding what might be emerging beyond the limits of a programme.

But relational evaluation is not easy work. It asks for time, trust and attentiveness — qualities that can be difficult to sustain across programmes that span dozens of places, multiple sectors, and shifting local realities.

Large-scale initiatives like the HSHAZ programme show just how uneven the relational landscape can be; context varies, capacity varies, and relationships vary.

A relational approach doesn’t offer quick fixes or tidy coherence. Instead, it asks us to begin where we are: by creating the conditions for honest conversations, building shared frameworks for learning, and making space for people closest to the work to shape what comes into view. These are small steps, but they are the ones that make room for a more relational form of evaluation to take shape.

Looking ahead, we might also consider whether our language needs to evolve alongside our methods. How we describe value in evaluation often reflects a productivity mindset, which emphasises outputs and endpoints. We might instead borrow a concept from Gujarati: the word ઉત્પાદક [1] (utpādak), often used in a similar way to ‘productive’, but carrying meanings that are far more generative — fertile, rich and unfolding. It invites us to think about conditions rather than conclusions. In many ways, a relational approach to evaluation is ઉત્પાદક (utpādak): it pays attention to what enables change, to the relationships that carry it, and to what continues emerging over time. Perhaps this is where a different kind of evaluation begins.

Acknowledgements

With thanks to The Relationships Project for their ongoing work on relational practice, which has helped shape and deepen my thinking, and to Alex Wardrop at Historic England for the opportunity to share these reflections.

Footnotes

[1] The Gujarati word ઉત્પાદક (utpādak) comes from Sanskrit उत्पादक (utpādaka) — ut- (to arise), pad (to bring forth), and -ak (the agent). It originally described fertile or high-yielding land in agricultural contexts — that which enables growth. In modern Gujarati it can mean “productive,” but it still carries these older senses of being rich, yielding and generative, which is how it is used here.