This paper explores the enormous variation in views, championing behaviours and impacts of liaison people: staff nominated to facilitate, tailor and promote SPIRIT (a research utilisation intervention trial in six Australian health policy agencies). Liaison people made cost/benefit analyses: they weighed the value of participation against its risks and demands in the context of organisational goals, knowledge utilisation norms, epistemology and leadership support. There was a degree of self-fulfilment (organisations got what they put in), but SPIRIT could not always be tailored to address local knowledge needs. We present nine propositions for identifying and supporting liaison people in similar interventions.