Analytics-Led HR — Prove the ROI at Vértice Consultoria

A four-round, advanced people-analytics simulation set inside Vértice Consultoria, a São Paulo professional-services firm (4,200 people, R$980 million revenue) whose workforce IS its product. On 16 June 2026 the CFO freezes the R$22 million HR program budget flat and gives the new HR Director 90 days to prove that ONE people-analytics-driven intervention earns its keep in R$ the CFO will recognise — or the budget is cut 15% next cycle and HR loses its seat at the table. High-performer voluntary turnover has climbed to 24% (industry norm ~15%); at R$180,000 per regretted exit and ~400 such exits a year, the firm bleeds roughly R$72 million annually — dwarfing the entire HR budget. Playing the HR Director, you (1) mine five years of people data to find WHERE the firm actually loses high performers, segmenting attrition rather than accepting the loud 'we have an engagement problem' framing; (2) choose the right, decision-relevant analytics question and ONE evidence-based intervention that addresses a cause rather than a correlation, from a firm-wide engagement program (R$6M), a manager-capability program for the weakest 20% of managers (R$3M), a project-staffing/utilisation rebalance (R$2M), and across-the-board retention bonuses (R$8M); (3) allocate within the frozen R$22M envelope, fund a measurement design (control group or staggered rollout) that lets you claim causation rather than assert it, and build a CFO-ready ROI model with a sensitivity range; and (4) commit the plan, design the manager-adoption mechanics that win skeptical line managers, set leading indicators, and defend a single-intervention focus to the CFO. The math rewards the narrow causal question, the cause-addressing intervention, a real measurement design, a focused spend and genuine manager adoption — and punishes the five classic errors: accepting the vague 'fix engagement' question, buying the symptom (retention bonuses), claiming ROI from a correlation with no measurement design, designing an elegant intervention nobody adopts, and diffusing the flat budget across all four programs so none can be proven. Final KPIs track Insight Quality, Causal Confidence, Manager Adoption and the net annual ROI in R$ millions the CFO will accept.