Refining advocacy assessment: reflections from practice

• This paper revisits how existing advocacy evaluation frameworks classify people and activities, and define and assess outcomes. We discuss how assessments could be more specific and propose bounding the scope of inquiry in one of four ways: strategy specific, outcome oriented, actor-centric or system-wide. • In classifying activities, the same action or event may be situated at different phases of the change pathway – in some cases used as a tactic to influence a policy outcome, and in others an intended outcome itself. • Because advocacy is more relational than other types of more technical development interventions, there will be fewer sources of directly observable data, and the direction of potential bias may be unknown. • In terms of learning, advocacy initiatives are contextually dependent, therefore lessons may be less directly transferable to subsequent phases of an initiative or to other settings. Organisations have bounded repertoires and the transferability of skill sets is limited, so advocates adapt how and with whom they engage more than what they do