01Define the test question
Choose one primary decision: product acceptance, price response, variation mix, channel fit or repeat-order reliability. Record the quantity, sales window and decision rule before stock arrives. A test with no rule can be interpreted to support any outcome.
02Check received against ordered
Count units and compare product reference, variation mix, visible condition, function, labeling and packaging with the approved record. Separate confirmed defects, transit damage, missing items and units reserved for display or testing.
03Calculate sellable-unit yield
Sellable-unit yield equals units accepted for sale divided by units received. Keep the reasons for non-sellable units beside the result. One percentage without defect, damage or handling context does not show what needs to change.
04Reconcile actual landed cost
Replace estimated goods, freight, duties, tax, inspection, packaging and payment costs with actual invoices where available. Divide the final total by sellable units. Record variance from the pre-order estimate and which assumption created it.
05Track demand in a fixed window
Record units made available, units sold, days on sale, channel, price changes, returns and remaining stock. Keep paid promotion or unusual exposure visible. Do not compare one product's sell-through with another product using a different time window or availability.
06Capture customer evidence
Group recurring questions, objections, returns and requested variations. Preserve privacy and distinguish a direct buyer statement from your interpretation. Customer comments help refine the next product brief; they do not replace measured sales and return data.
07Score supplier execution
Review product match, quantity accuracy, change control, packing, dispatch timing, documents and issue response. Separate supplier-controlled stages from carrier or customs stages. Record how each exception was resolved, not only whether one occurred.
08Choose one next action
Reorder when the defined quality, economics, demand and supplier conditions pass. Revise or retest when one fix can answer the remaining question. Stop when an essential requirement fails or the economics remain outside the business's safe range.