Review Management & Bad-review Follow-up

#Use poor reviews to find unresolved buyer issues

Open Shops → target Shop → AI Customer Service → Review Management. Enable review optimization, then Bad-review follow-up, choose the rating, recent-day window, and contact limit, and save. The Shop also needs valid authorization, enabled Customer Service, and an online Desktop handling its CS work.

The goal is to investigate and resolve the buyer's issue. The Agent uses the review, order, and existing conversation. Refunds, replacements, or extra compensation outside Shop rules still require escalation and human review. Enabling follow-up does not promise that a buyer will change a review.

APP CAPTUREThe English light-theme Review Management settings with review optimization, bad-review follow-up, rating, recent days, and maximum contact count.

#1. Choose reviews to follow up

The system also checks order/buyer identity, current review state, and conversation eligibility. Current eligibility requires an order created within the last 60 days. A review inside the selected window therefore does not guarantee contact. Missing buyer identity, expired authorization, or a conversation with AI disabled can prevent sending.

Begin with a narrow rating or recent-day window. Check that messages address the issue without requesting information already supplied, then expand the scope if appropriate.

#2. Understand the three-contact schedule

The selected limit produces one of these plans:

• 1 contact: initial contact only.
• 2 contacts: a second follow-up at least four hours after the first successful send.
• 3 contacts: a final follow-up at least eight hours after the second successful send.

These are minimum waits rather than appointments. Actual delivery depends on the half-hour scheduler, Desktop availability, and eligibility. Failed sends do not count as successful contacts; a technical retry of the same round must not become another duplicate contact with the buyer.

Lowering the cap constrains existing plans. Raising it applies only to new plans and does not reactivate plans already ended.

#3. Connect buyer replies with human decisions

A buyer reply stops proactive follow-ups and the normal Customer Service conversation continues. A genuinely resolved issue should not receive another check-in. An open human Escalation pauses review reminders while a person decides.

For example, a buyer reports that a lamp will not turn on. Check earlier communication and product instructions first, without asking again for evidence already provided. If a replacement exceeds autonomous authority, escalate the order, completed checks, and recommendation. After the person decides, AI continues the agreed resolution.

A conversation marked as replied does not by itself prove that the complaint is resolved. Check whether the actual refund, replacement, or other commitment has completed.

#4. Query and export review details

In Chat & Sessions, ask:

Export Brightloom US's 1–3-star product reviews from the last 14 days, including Shop, Seller SKU, order ID, rating, original text, and review time. Summarize issue categories. Analyze only; do not contact buyers.

Review analysis is available independently of the follow-up switch. Dates refer to Shop-local review creation. Coverage includes acquired reviews rather than a guarantee of all platform history. Anonymous or unreliably linked buyer identity fields remain blank; do not infer a name.

#5. Check why no message was sent

1. Check Shop authorization, Customer Service, and the handling Desktop.
2. Check both review-management switches and the saved settings.
3. Verify rating, review date, order date, and buyer identity.
4. Check for a buyer reply, exhausted contact cap, or disabled conversation AI.
5. Inspect Escalations for a decision still waiting on a person.

Unpaid-order reminders have different eligibility and stopping conditions. See Unpaid-order outreach and Pickup reminders for the other workflows.