Order Flow
Part of Testing a dropshipping business
Measuring delivery reliability before increasing marketing spend
Compare all due orders with the original delivery promise, including overdue unresolved cases, before increasing Australian dropshipping promotion.
Before increasing marketing spend, compare orders already due with the delivery information buyers saw when they paid. Count delivered, late and unresolved outcomes for the same product, supplier route and destination group. Few complaints or a collection of tracking numbers do not show that parcels arrived as promised.
Keep the original promise
For each order, save the delivery range shown before purchase and the payment date. Record supplier acceptance, confirmed carrier handover, delivery or another resolution, with dates and evidence sources. An internal “shipped” label may appear before a carrier has the parcel.
Choose the counting rule before reading the results. For an arrival range, treat a confirmed delivery by its final stated day as on time. Record later customer updates, but retain the original range for judging the original promise.
Count all orders whose deadline has passed
Use orders whose original final delivery day has passed as the due cohort.
Classify each once:
| Outcome | Count when |
|---|---|
| Delivered on time | Delivery is confirmed by the original final day. |
| Delivered late | Delivery is confirmed after that day. |
| Overdue and unresolved | The day has passed without confirmed delivery or another documented resolution. |
| Resolved without delivery | A refund, replacement arrangement or other outcome is documented; describe its timing and status separately. |
For a simple on-time share, divide orders delivered on time by all orders in the due cohort and show every outcome count alongside it. This treats unresolved overdue orders as not confirmed on time.
Keep orders still within their range out of the due cohort. If you exclude a customer-initiated cancellation, state the rule and count excluded orders separately. Do not silently remove late parcels that lack a final scan.
A few due orders produce a volatile percentage. Inspect the cases and compare like routes: a metropolitan parcel from one origin does not establish performance for a regional or overseas route.
Key Delivery Reliability Metrics
- On-time share
- Delivered on time ÷ Total due cohort
- Late deliveries
- Count of deliveries after original deadline
- Unresolved overdue
- Orders past deadline with no confirmation or resolution
- Resolved without delivery
- Refunds or replacements documented
Find where delays began
For each late or unresolved order, check supplier acceptance, processing, carrier handover and transit separately. A carrier event cannot explain a delay before handover. Confirm the cause before changing the advertised delivery range.
Record when the store learned of a problem and whether the buyer was updated. A delay alone does not establish what outcome is appropriate; assess the circumstances and communicate clearly with the buyer.
Make the spend decision
A small successful cohort may justify another bounded test, with the same measurement carried forward. Correct an inaccurate promise before seeking more orders.
These counts describe the observed cohort. They cannot guarantee later supplier or carrier performance.
