Why is the order-status queue the right first category?
In most stores the evening and peak-season queue is dominated by one question with a checkable answer in a system the store already runs: where is my order. The answer lives in the order record and the carrier's tracking status, the reply is the same shape every time, and a wrong answer is easy to detect. That makes it the category with the cleanest evidence and the fastest approval record, which is what a Customer Support Digital FTE on ftes.ai needs to be useful in the first month.
It is also the category whose overnight coverage matters most to shoppers and costs the team the most to staff.
How does a reply get its source?
The FTE reads the ticket, finds the order in the connected store through a dedicated, scoped account, reads the fulfilment and carrier status, and drafts a reply that cites the order record and the tracking status it read. How-to and returns-policy questions cite the knowledge base article they were answered from. A ticket whose answer is not in either source escalates with a note that the base has a gap rather than being guessed, and recurring gaps become proposed articles for the manager's approval.
Review of a cited reply is checking the citation and the tone, which takes seconds, which is why the approval record builds quickly.
What escalates every time?
Refunds and credits, returns disputes, chargebacks, address changes after dispatch, anything involving a damaged or missing item where a decision about money follows, complaints, legal or safety language, and any customer in distress. The always-escalate list is fixed in the remit at every autonomy level, grows after incidents, and never shrinks by a setting. Each escalation reaches the person who owns that category with a four-part summary: what was asked, what was found in the record, why it did not resolve, and a suggested next step.
An escalation that arrives as a reassigned ticket without the summary is sent back; the standard is a remit rule.
- Money: refunds, credits, disputes, chargebacks
- Fulfilment changes: address after dispatch, damaged or missing items
- People: complaints, legal or safety language, distress
How does the category earn autonomy?
At the first or second monthly review, if the order-status category's approval rate is clean and the send-back reasons have been about tone rather than facts for several weeks, the manager can promote that one category to auto-send within bounds: only tickets matched to an order, only statuses the carrier reports, and only replies whose template the manager approved. Everything else stays gated, and everything on the escalate list stays escalated. If send-backs rise after the promotion, the category steps back one level and the record says why.
Money never earns autonomy. That is a design decision, not a setting.
What should a store measure, and when?
First-line resolution for the order-status category against the baseline; median first response including overnight and weekends; escalation volume by reason and escalation quality, meaning summaries that met the standard; knowledge gaps closed as proposed articles approved; and approval rate with send-back reasons. Measure through a peak, because a first line that holds in a quiet month and folds in a busy one has not been measured. Satisfaction is measured where the store surveys it and marked unmeasured where it does not.
Never measure reply volume. Drafts nobody approved are tasks, not value.
What does the first month look like?
Week one: confirm the order-status and returns articles, write the always-escalate list with receivers and the summary standard, connect the helpdesk and the store with scoped credentials, and record the baseline. Weeks two to three: approve cited replies in seconds, send back with reasons, check every escalation against the standard, and approve the first proposed articles. Week four: read the FTE's own report and decide whether order-status has earned auto-send within bounds. Refunds and disputes were never on the table.
By day 30 the store's most repetitive question has an owner that answers from the record at any hour, and the team starts every hard case from a summary.