Farm Box | Subscription Unit Economics Model

The editable model is ready. All figures are fictional planning assumptions for a made-up produce box service, and every number in the file is a live formula except the light-blue input cells.

Structure (8 sheets)

• Read Me — purpose, sheet guide, how to use, colour legend, definitions, and the model's simplifications.
• Assumptions — global fees and growth, a three-column scenario driver table with a live "Selected" column, and the Active scenario dropdown (B21). Only the light-blue cells are inputs.
• Order Cost Detail — the 36 rows you asked for: 12 weeks × 3 box types, each with price → discount → net revenue → produce, packaging, delivery, labour, payment fees → variable cost → contribution margin → totals.
• Weekly Model — 12-week roll-up with cumulative contribution, cumulative revenue and running margin.
• Unit Economics — per box type and blended: full cost stack, CM per order and %, monthly contribution, lifetime, LTV, CAC, LTV:CAC, payback in weeks and months, 12-week totals, contribution share.
• Scenarios — Baseline / Optimistic / Conservative side by side plus the 9-row per-box-type build-up behind each figure.
• Dashboard — 8 KPI cards plus 6 charts (weekly revenue vs contribution, cumulative contribution, CM per order by box, LTV vs CAC, contribution vs acquisition spend by scenario, order mix).
• QA Checks — 22 formula-driven checks that re-run after any edit.

Baseline results 1,480 orders and $61,095 net revenue over 12 weeks, $22,467 contribution (36.8% blended). Small: CM $8.63/order, LTV $623, CAC $45, payback 5.2 weeks. Standard: $14.84, $1,285, $60, 4.0 weeks. Family: $26.42, $2,859, $80, 3.0 weeks. Blended LTV:CAC 23.5x. Across scenarios LTV:CAC runs 36.4x optimistic / 23.5x baseline / 13.8x conservative, with payback 3.0 / 3.9 / 5.5 weeks — the ordering holds in every case.

Verification I recomputed the model independently in Python from the 36 order rows and the raw input cells: every row identity, all 12 weekly totals and cumulative series, per-box-type and blended unit economics, and all three scenarios match the workbook with zero deviation. All 22 in-file checks read PASS. Cached values are embedded, so numbers display correctly in viewers that don't recalculate. Every sheet was rendered and visually inspected; two rounds of layout fixes removed clipped labels, notes cut mid-sentence, a concatenated summary label and cramped pie labels.

Two things worth knowing

• The 12-week "contribution after acquisition spend" line is negative by design: acquisition spend buys a customer whose contribution accrues over ~20 months, so a 12-week window is not a payback test. Use the payback rows for that question.
• Churn drives customer lifetime and LTV but does not decay the weekly order base — the 12 weeks are modelled as a steady acquisition flow, not a closed cohort. This is stated on Read Me.

Minor residual: the CAC value labels in the LTV-vs-CAC chart are small because the bars are tiny next to LTV; they are legible at normal zoom, and the exact figures sit in the Unit Economics table.