How to Review a Fulfilment Exit Without Guessing

An exit interview is a lead, not a root-cause report

A seller may accurately report late customer deliveries, a higher quote, poor communication or weak margin. That statement matters because it describes the experience that changed the relationship. It still does not show where time was lost, whether two quotes covered the same service or whether a later process change prevented recurrence. Calling the feedback a complete diagnosis creates false confidence in whichever lesson sounds most convenient.

Two real exits show why the distinction matters

In one reviewed conversation, a seller who moved fulfilment in-house cited both delayed customer deliveries and higher pricing. The archive supports those stated reasons and the loss of confidence. It does not contain the milestone-level investigation needed to establish the delay root cause, a corrective owner or a verified post-change result. That incident should remain open in an internal review; it should not be marketed as a solved case.

In a separate conversation, another seller confirmed that an alternative fulfilment quote was cheaper, yet the product still appeared unprofitable after advertising spend. That supports a narrower conclusion: the lower quote did not make that particular product test viable on the seller's stated numbers. It does not prove fulfilment price was irrelevant, that every competing quote is incomplete or that no different product strategy could work.

Build the evidence timeline

For a delivery complaint, reconstruct the order journey using the milestones that actually existed: payment, release to fulfilment, purchase or stock allocation, warehouse receipt, packing, carrier acceptance, export, destination handoff and delivery. Record the first point at which the order exceeded its agreed window. If a milestone is missing, mark it unknown rather than assigning the delay to the warehouse, supplier, carrier or customer by instinct.

Compare cost on the same scope

Put both offers on one line-item basis: product specification, inspection, packaging, pick and pack, payment charges, chargeable weight, shipping service, duties or tax arrangement, claims handling and expected delivery window. Then add acquisition cost, refunds and support cost to the seller's contribution margin. A cheaper fulfilment quote can be genuinely better and still leave the product unprofitable; both facts can be true.

Turn a supported cause into corrective action

• State the incident and customer effect without softening it.
• Name the supported cause, or state that root cause is not yet established.
• Assign an owner and a dated action to the failed control.
• Define how the change will be checked: a cohort, sample, milestone report or recurrence window.
• Keep commercial changes separate from operational corrections unless the evidence links them.

Closure requires verification, not better wording

An apology, an internal share or a new classification label is not proof that a delivery problem was corrected. A responsible closure note says what changed, which later orders were checked, over what period and what remains uncertain. If that evidence does not exist, the honest status is unresolved. That is more credible than publishing reassurance the operating record cannot support.

What this guide can and cannot claim

The two exit conversations are authentic and anonymised, and their stated reasons are preserved. They demonstrate why delivery evidence, like-for-like cost and product economics must be reviewed separately. They do not establish a root cause for the delayed parcels, prove a later fix or support a promise that the incident cannot recur. This guide publishes the review method; it does not convert an unresolved departure into a success story.

Evidence and privacy note

This guide is grounded in two separate reviewed operations conversations. No statements are merged between the sellers. Client and store identities, exact products, contact details and shipment records are omitted; broad commercial context is retained only where it explains the review method.