Last week’s Value Articulation Intensive (VAI) gathering surfaced a familiar question: when to use existing data to support the case to funders or other stakeholders vs. when to create new evidence for the decision at hand.
The discussion got me thinking about the value of understanding our defaults. Sometimes we default to assuming that we need airtight evidence to support our goals — our heuristics-hungry brains err to the side of caution and assume we need more data or polish than is necessary, and, as a result, the learning we could pursue is squashed before it can even get off the ground. It’s easy to confuse or conflate two very different questions:
“Will this pass peer review?”
“What evidence would help the people involved make a better choice?” This frames research as a range — from minimum viable to comprehensive — that you define with your partners based on the decision at hand.
Question #1 shows up in a variety of ways. Museum professionals don’t have to be publishing to academic journals for this standard to be a kind of unspoken default. It may present itself in your internal monologue as, “Ok, but could I share this project at a conference?” or “But what if Sam tries to sink our ship because we’ve only interviewed 10 people?” We unconsciously migrate academic standards to our efforts because we have egos that need tending or insecurities that we can’t shake, or because a “that’s just how things work around here” approach is pervasive.
Two learning approaches
Let’s assume you’re asking question #2 — “What evidence would help the people involved make a better choice?” — because if you’re reading this it’s probably the only question you should be asking.
There are two ways to answer the question: a) start with existing evidence (could be yours, but doesn’t have to be) or b) create the evidence you need to support a better decision.
It’s important to note that supportive evidence doesn’t necessarily mean positive or affirming evidence. Sometimes “supporting” a purpose or decision means what you learn will disprove what you believe. We can open to that and welcome it as a way to redirect our efforts to other questions or opportunities.
Both these approaches — off-the-shelf and tailored — are valuable. In reality, they’re not binary but live on a continuum, and they’re not mutually exclusive — you can do both — but sometimes working in binaries has the advantage of making ideas memorable, so that’s how I’ll portray these things here.
Incidentally, we’re exploring both these approaches through the Value Articulation Intensive (off-the-shelf) and the Value Realization Collaborative (tailored).
Off-the-shelf (or starting with existing evidence)
This approach works when the evidence already maps to your stakeholders’ decision space. John Falk’s research on museum value translates across contexts. A city councilor in Denver and a provincial minister in Ontario face different political landscapes, but they share similar questions about public investment. The patterns hold, so your job is translation: how does this existing evidence speak to this specific stakeholder’s priorities?
The academic rigor makes sense here. John’s research should meet academic standards. It needs to hold up across contexts, withstand scrutiny, and be defensible at scale. That’s what makes it generalizable.
Tailored (creating purpose-driven evidence)
This approach is better when existing data can’t get you where you need to go — or when the value of new information justifies the effort to gather it.
Museums participating in the Value Realization Collaborative are pursuing this path. When they talk with newcomers to their regions about how they cultivate a sense of connection and belonging with others in their new communities, they learn about well-being strategies and build relationships with the people they hope to serve. The conversation generates both insight and connection.
One signal that you might need to take the ‘tailored’ path is if your stakeholders' mental models don't match the research categories. The way they think about success or impact requires evidence shaped differently than what's available. Or the decision requires understanding something specific about your community that no general study can capture. The question to ask is whether you need ‘proof’ or you need clarity about what would constitute useful information for this specific choice. Too often, we make assumptions about what constitutes proof for others or what would satisfy their needs without decomposing the problem or question with them.
What decomposition looks like in practice
Before you design any learning process, ask your stakeholders: "What would constitute useful information for this specific choice?" Or, better yet, ask, “How have you made decisions like this before?”
Not “What proof do you need?” — that imports academic standards. Not “What would convince you?” — that frames it as persuasion. Decompose the decision instead. What matters most? What assumptions are we operating under? What would change how we think about the options?
This shifts the focus from “build a case” to “illuminate the choice.” Learning becomes a tool for clarity, not ammunition. And the conversation itself — where stakeholders surface competing priorities and clarify what “good enough” means — is often more valuable than the evidence you gather.
Sometimes decomposition reveals you don’t need new evidence at all. Other times, it shows you need experimentation, not study. The default shouldn’t be comprehensive rigor. The default should be: decompose the decision first, unpack conflicting assumptions, then match the learning to what actually moves the choice forward.