What’s inside: monitoring · output checking
The question this file answersWhat happens when a chargeback lands on a Friday and the evidence is spread across the OMS, the carrier and the help desk?
Fits: retailers where 15–30% of orders come back.
Typical pattern, not a measured desk — the e-commerce playbook gives a returns rate (15–30% of orders in fashion and general retail) but no chargeback volume; none is invented. Notices arrive; a specialist rebuilds proof of delivery, order and messages by hand while the scheme's response window, days long, runs.
On the day a notice arrives, the specialist opens a draft with tracking, authorisation and the customer thread already attached, and a list of what the scheme's reason code still needs. The decision to fight or accept, and the fight itself, remain theirs; the gathering is gone.
Typical, not a measured client result. Every figure here comes from the playbook source named below.
the response drafted with the evidence attached — the same build as our returns file; no published figure for chargebacks
Before: a specialist rebuilds chargeback evidence from scratch while the representment clock runs. After: the draft arrives with the evidence attached and the specialist decides whether to fight. Chargebacks have no published figure, so the hero names the mechanism instead.
Chargeback administration has no published benchmark. Narvar (2023) reports 50–65% less returns handling time — a returns figure that belongs to the parent file and is not applied to disputes. Not a win-rate.
Your shop platform, OMS and help desk stay — Shopify, NetSuite, Gorgias, Zendesk or the ones you run. The returns build is pointed at disputes:
The representment itself. Fraud judgement. The customer conversation. No model closes a dispute.
Not a win-rate.
A dispute is only as strong as the evidence in your systems; if proof of delivery is not captured or messages live in personal inboxes, the draft has gaps a person must fill. Acquirer portals differ, and some accept uploads only by hand. Narvar's 50–65% is about returns handling and lives on the parent file; there is no published chargeback figure, and no win-rate is implied.
The path: free 60-second estimate → free 20-minute review → paid audit of this one process (€1.5–3K, typically two weeks) → pilot with your people in the loop (€10–20K, weeks, not quarters). No transformation programme. Prices are public, on the services page →
Scoped in the audit — the playbook has no estimate for this exact desk.
That depends on your volumes and wage costs — this page will not invent the number. The free 60-second estimate runs that calculation from your answers, with every multiplier sourced.
Not a named Aperanda client. Process file · Ops.
Short process file. Same build as its parent file; the playbook has no separate volume or benchmark for this desk.