Review moderation
Last updated: April 17, 2026
Every review a customer submits lands in your moderation queue. Auto-approve rules can publish some of them automatically — everything else waits for you to decide. This page walks through the Reviews tab and the workflow around it.
The Reviews tab
Screenshot of the Reviews tab in the BetterReviews admin — manual capture, drop in here.
You’ll find it in the main navigation of your BetterReviews admin. The top of the page shows:
- Your total review count (across all statuses)
- Your average rating across approved reviews only — so your displayed rating never moves because of pending or rejected submissions
Below that, four tabs filter the list:
- Pending — reviews waiting for your decision (direct and public submissions only; imported platform reviews skip this tab)
- Approved — reviews currently live on your widget
- Rejected — reviews you’ve rejected (kept for record-keeping, never shown publicly)
- All — every review regardless of status
Each tab shows reviews in a table you can sort and filter.
Sort and filter
At the top-right of the review list:
- Sort — Most Relevant (default), Most Helpful, Newest, Highest Rated, Lowest Rated, Most Detailed, Featured, Most Informative, Critical First, Best Media, Marketing Potential
- Rating filter — Any rating (default), or a specific star count (5, 4, 3, 2, 1)
These work within the current tab. Pending tab filtered to 1-star reviews shows only 1-star reviews currently awaiting your approval.
Screenshot of the Reviews tab with sort and rating filter dropdowns open — manual capture, drop in here.
Opening a review
Clicking any row opens a detail modal with the star rating, status badge, title, body, attached photos, reviewer name, redacted email, verification badge, platform, product, and submit date. If the review came from an AI conversation, an “AI Conversation” badge is shown.
Below the content, a Signals row surfaces our AI analysis of the review:
- Sentiment —
positive,neutral,negative, ormixed. Tone-coloured so you can scan at a glance (green = positive, red = negative, yellow = mixed). - Quality score —
Quality: N/10, from 1 (low-signal) to 10 (marketing-grade). Green when ≥7, red when under 4, plain otherwise.
Reviews imported from a platform (Okendo, Judge.me, Yotpo, Loox) or CSV don’t always have signals yet — the row hides when both are absent. See review signals for how the scores are produced.
If the review came from an AI conversation, a Show conversation transcript button appears below the content. Click to expand the full turn-by-turn chat that produced the review. See viewing the conversation transcript for what’s in the transcript and how PII is handled.
Screenshot of the review detail modal open over the list — manual capture, drop in here.
From the detail modal you can approve, reject, or delete the review — buttons shown match the review’s current status (see approving, rejecting, and deleting reviews).
Bulk actions
Select multiple reviews using the checkboxes on the left. Selecting at least one row reveals bulk action buttons in the list header: Approve, Reject, Delete.
Bulk actions are great for:
- Clearing out a large pending queue after you enable auto-approve for the first time
- Approving a batch of 5-star reviews that all look solid
- Rejecting or deleting a batch of obvious spam
Bulk delete asks for confirmation since it’s irreversible. Bulk approve and reject don’t — you can always flip status back later.
Auto-approval
Auto-approve is off by default. Turn it on in Collect when you’re ready. See what happens after your customer submits for the three rules you can enable.
Reviews in the Pending tab are the ones that didn’t auto-approve.
How many reviews should be pending?
If you’ve enabled auto-approve verified buyer and auto-approve high quality, the pending queue should stay small — usually a handful at a time, mostly from widget submissions (public, unverified buyers). Check it daily or every few days.
If your pending queue is growing faster than you can work through it, turn on more auto-approve rules or lower the quality threshold.