Australian governments fund many place‐based collective impact initiatives like Communities for Children (CfC); states fund family support services to keep at‐risk children out of the child protection system; and schools implement many extracurricular programmes. Do these services have a beneficial, cost‐effective collective impact on child well‐being? This paper describes a proof‐of‐concept attempt to address this question by linking for one CfC community individual‐level data on 5‐ to 12‐year‐old participation in programmes delivered through these three sectors with child outcomes. This exercise was unsuccessful despite the investigators’ prior experience; advice from a data linkage expert and our data custodian partners; five ARC reviews; partners’ good will; and ethical safeguards including written, informed parent/carer consent. Obstacles encountered included a lack of data of sufficient quality on children and families’ participation in services, lack of data on children’s outcomes, and prohibitive costs of linkages within government. We offer for others three key lessons: (1) make assumptions explicit; (2) talk to technical experts in data custodian organisations early in the planning process; and (3) undertake, if possible, an initial scoping exercise. We conclude that despite recent legislative and policy reforms many obstacles we encountered will persist in the absence of a national child well‐being strategy.