A contribution model was built for every property — revenue, less direct operating cost, less vendor cost, less internal labour requirement, less exception burden. Several properties that looked strong on gross revenue were consuming operational resource out of all proportion to what they returned.
Properties were segmented by value and complexity, each tier receiving a different service architecture rather than one standard applied to all.
Vendor management was rebuilt: preferred and backup rosters, response-time commitments, escalation tiers, pricing agreements, performance scorecards. One vendor held nearly 40% of emergency maintenance spend — a concentration risk that had never been named. A second supplier was qualified before it became urgent.
Capacity planning tied property count to complexity to staff capacity to a hiring trigger, so "at what portfolio size do we need another operations manager" had an answer before the strain arrived. Onboarding moved from ad hoc to a structured ten-step workflow.