weekly_hiring_approvals.xlsx · as of Aug 31, 2026 · illustrative data

Key insight

I'd approve the ten requests that clear the checks and send back both Challenge requests. The Data Engineer and Sales Coordinator are unbudgeted adds, and each one puts a department that's already over budget further over; Sales Support alone is $310K over before the hire. Holding them keeps $406K of FY27 room instead of $159K.

The bigger fix is the benchmark. "Under budget" is a weak test when the budget itself is a blanket raise on last year. I'd work with the head of HR and the CFO to set headcount KPIs by department, built from historicals: headcount and payroll per $1M of revenue, cost per FTE vs. plan, vacancy rate and time to fill, and backfill vs. net-new mix. That gives every hiring decision a trend line to be judged against, not just a budget line.

Architecture

Three checks, four decisions

Each request runs three checks: is the offer more than 5% above its reference pay (the last person's pay for a backfill, budget pay for a budgeted seat, market midpoint otherwise); is it a new seat outside the budget; and, counting the easier requests ahead of it, does it push the department over budget. Backfills only get the pay check, since the seat is already budgeted.

The tool does the math; people still make the call. The meeting spends its time on the Review and Challenge rows instead of re-deriving the same numbers for every request.

Temporary vs. ongoing savings

The year-to-date number hides the real story. Open-seat savings stop the day those seats fill, while pay above budget and unbudgeted hires keep going and grow to a full year. Splitting the two is what turns "$526K under budget" into "$542K over on a run-rate basis."

blouse_capital_dashboard · power bi

Key insight

Cumulative IRR sits at 35% (7% still in the pipeline). The returns are outsized by design: as a minority co-investor alongside a capital partner, Blouse Capital earns a promote on top of stacked acquisition, development, asset-management, and property-management fees. So when a deal executes to plan, the equity return compounds. On current validated cap rates, data centers are driving performance; office is the drag.

Architecture

Pooled cash flows, not averaged returns

IRR cannot be averaged across assets without distortion, each property has its own timing and magnitude of cash flow. The model pools cash flows across the portfolio before computing IRR at the corporate level, and separately at the per-deal level, so the headline return is mathematically defensible rather than a shortcut.

Power BI's DAX has no native XIRR function, so the iterative IRR calculation is verified independently in Python before being trusted in the dashboard.

Corporate model assumptions

The corporate layer is built bottom-up. Corporate compensation runs at 24% of fee revenue, and direct operating costs (field staff and regional operations) at 52% of fee revenue. Together they drive a total corporate EBITDA margin of roughly 12%.