A velocity-based workforce allocation model that transformed capacity planning from reactive to predictive across six global delivery regions.
Resource allocation across six delivery regions was driven by historical headcount models with no relationship to actual project velocity or demand patterns. Bottlenecks were diagnosed only after they caused delays. Contractor utilisation was routinely below 60% while critical path work was under-resourced. There was no shared language between project managers and resourcing teams, producing a chronic mismatch between supply and demand.
Developed a velocity-based allocation model mapping each project's demand profile against a regional capacity baseline updated weekly. Introduced sprint-level velocity metrics — story points, cycle time, and throughput — as the primary language of resourcing conversations. Built a demand forecasting model using trailing 12-week velocity data to surface bottlenecks 3 weeks in advance. Ran a structured enablement programme to shift resourcing managers from headcount-think to throughput-think.
Resource utilisation improved from 61% to 83% within two quarters. Bottleneck detection moved from reactive to proactive, with 3-week advance notice achieved in 78% of cases. Contractor over-allocation incidents fell by 65%. The model was rolled out to four additional regions within 12 months of initial deployment.