← Back to cases
CASE A04  /  CREATIVE PRODUCTION · AGENCY

Real-time project P&L from messy timesheets — agency leadership stopped flying blind.

An 80-person NYC creative production studio was running ~40 active projects at any time, but project profitability lived in a quarterly spreadsheet exercise — by which point three projects had already over-served by ~€50k each. We built a daily P&L pipeline that ingests Harvest, Xero, vendor invoices, and a custom rate card, then surfaces live margin per project with alerts at 70% and 90% of budget burn.

+12 ppGROSS MARGIN
DailyP&L vs MONTHLY
−95%REPORT PREP
8 wkTO ROI
Dashboard showing financial charts on a screen
CREATIVE STUDIO · NYC · FEB 2026 · CONFIDENTIAL DETAILS REDACTED
THE PROBLEM

Quarterly P&L reviews caught over-servicing too late. Leadership was flying blind on margin.

The studio's CFO ran a quarterly project P&L exercise — pulling Harvest hours, vendor invoices, and rate cards into a spreadsheet, allocating fixed overhead, and producing a per-project margin. By the time the exercise finished, the quarter was over. The previous quarter's three worst over-servings had already happened, eaten into margin, and were unrecoverable.

Producer-level visibility was even worse: producers eyeballed Harvest weekly but had no clean view of vendor cost burn or fee burn vs. timeline burn. They consistently caught overruns 60-80% in, when there was no time to renegotiate.

THE PROTOTYPE

Daily margin per project. Slack alerts at 70% and 90% of burn.

We piped Harvest timesheets, Xero invoices, vendor PO data and the rate card into a Python aggregation that runs every night. Output: per-project daily P&L with three timelines — fee burn, hours burn, vendor cost burn — overlaid on the contracted scope. Alerts fire into a producer-specific Slack channel at 70% (warning) and 90% (escalation).

Critically, the dashboard rolls up: by client, by service line (production / strategy / motion), by producer, and by month. Leadership sees the same numbers producers see, with no quarterly delay.

THE PILOT

Eight weeks live across 40 active projects. Three over-servings caught mid-flight.

Three weeks in, the system flagged a project at 71% scope burn with 40% of timeline remaining. Producer renegotiated the scope with the client (additional brand assets had crept in). Six weeks in, two more catches — both renegotiated successfully. Total recovered margin in the pilot quarter: ~€140k.

The behavior change was the bigger win: producers started checking the dashboard daily, not weekly. Conversations with clients about scope happened proactively, not at the post-mortem.

First quarter in three years where we knew project margin in real time. We renegotiated three scopes mid-flight instead of eating the loss.
THE OUTCOME

Six months after rollout.

  • Gross margin+12 ppAverage across all projects
  • P&L cadenceDailyWas quarterly
  • Over-servings caught mid-flight3First pilot quarter
  • Recovered margin in pilot quarter~€140k
  • Quarterly report prep time−95%5 days → 4 hours
WHAT'S NEXT

Tying utilization into the model and forecasting next-quarter capacity.

Phase 2 layered in utilization (planned vs. actual hours per role) and a forward-looking capacity forecast. Result: studio leadership can now see 'we are over-sold on motion in Q3' six weeks before the quarter starts, in time to subcontract or push timelines.

YOUR P&L WORKFLOW

Finding out a project lost money two months after it shipped?

20-min audit. We look at how you currently get to project P&L, sample one recent project, and tell you exactly what daily visibility would cost (and save).

Take 2-min assessment