Metric Definitions — Khumalo & Grain Analysis

These are the precise definitions for every metric computed in this analysis. When AI computes a metric, check the result against these definitions. A number that looks right but was computed differently is wrong.

1. Net Channel Revenue

What it measures: Total sales revenue by channel, with refunds removed.

Why it matters: Comparing channels by gross revenue (including refunds) misrepresents performance. A channel with more refunds looks worse than it is if refunds are subtracted, and better than it is if they are included in a simple sum.

Formula:

SUM(sale_amount) WHERE sale_amount > 0

Grouped by channel.

Important: Revenue EXCLUDES refund rows (negative sale_amount values). Including refunds understates revenue and misrepresents channel performance. Filter to sale_amount > 0 before summing.

2. Channel Profit Margin

What it measures: The percentage of revenue retained after all direct costs, by channel.

Why it matters: Retail has the highest gross revenue but retailers take 35-45% commission. Commission and online channels have no retailer cut but different cost profiles. This metric reveals which channel actually keeps the most money.

Formula:

(net_revenue - retailer_commission - material_costs - delivery_costs) / net_revenue * 100

Grouped by channel.

Where:

• net_revenue = SUM(sale_amount) WHERE sale_amount > 0 for the channel
• retailer_commission = SUM(sale_amount * retailer_commission_rate) WHERE sale_amount > 0 for retail; 0 for commission and online
• material_costs = SUM(material_cost) WHERE sale_amount > 0 for the channel
• delivery_costs = SUM(delivery_cost) WHERE sale_amount > 0 for the channel

Important: Retailer commission applies only to the retail channel. The commission rate varies by retailer (35%, 40%, or 45%). For commission and online channels, retailer commission is zero. Use only positive sale_amount rows.

3. Revenue per Workshop-Week

What it measures: Revenue normalized by production time, by product type.

Why it matters: A dining table that sells for R25,000 but takes four weeks to build generates less revenue per workshop-week than two shelving units at R9,000 each that take one week each. This metric reveals which products use workshop capacity most efficiently — the insight Naledi has never seen.

Formula:

SUM(sale_amount WHERE sale_amount > 0) / (SUM(production_days) / 5)

Grouped by product_type.

Where:

• Workshop week = 5 working days
• Only positive sale_amount rows are included in both the revenue sum and the production days sum

Important: This is the metric that changes Naledi's view of her business. If shelving generates more revenue per workshop-week than dining tables, that means shelving is a more efficient use of her workshop's limited capacity — even though each individual table sells for more.

4. Monthly Net Revenue

What it measures: Net revenue by calendar month.

Why it matters: Twelve monthly data points show whether revenue is growing, flat, or seasonal. Naledi suspects December was slow — the data confirms or refutes this.

Formula:

SUM(sale_amount) WHERE sale_amount > 0

Grouped by calendar month (YYYY-MM).

Important: For trend analysis. Twelve data points spanning January 2024 through December 2024. Exclude refund rows.