Merit ACLabs
Sample report · Review Responder · fictional sample data

Sample data. This is the unedited report Review Responder produced on its bundled sample files, which describe a fictional business, run the same way a subscriber's upload is run (as of 2026-10-14). No AI model and no person touched it. Your own results depend on the exports you upload.

The full download also contains: needs_person.csv, replies.csv, replies.md, report.html, topics.csv.

Review replies

Review Responder report, as of 2026-10-14

Summary

  • 38 reply drafts ready in replies.md and replies.csv, starting with the 10 flagged below.
  • 10 reviews need a person to answer: health or safety (3), legal threat (2), discrimination or harassment (1), abusive language (1), names an employee (1), refund or chargeback demand (2), check if genuine (2). Their drafts are deliberately cautious.
  • Average rating 3.1 out of 5 across 40 rated reviews.
  • Top complaint: Damaged or defective (5 reviews). Top praise: Food & drink (9 reviews).
  • 4 reviews already answered and skipped.
  • 1 draft is for a review more than a year old; answer the recent ones first.

Answer these yourself

Each of these has a careful draft that invites the reviewer to contact you privately and does not admit fault or repeat the accusation. Read the full review before posting. For a legal threat, talk to your lawyer or insurer before you reply at all.

# Platform Date Reviewer Stars Why Review
1 Google 2026-10-08 Priya Shah 2 Health or safety issue My oat latte was made with regular milk and I had an allergic reaction. I ended up in urgent care.
2 Google 2026-10-07 Victor Hale 1 Health or safety issue; Legal threat or complaint to an authority Got food poisoning after the chicken wrap. I'll be reporting you to the health department.
3 Yelp 2026-09-25 Greg H. 1 Health or safety issue; Legal threat or complaint to an authority I'm calling my lawyer. The chair broke and I hurt my back.
4 Trustpilot 2026-10-13 - 1 Discrimination or harassment claim Treated like a criminal - Racist staff at the Harbor Street location. I was followed around the shop.
5 Yelp 2026-10-10 Brianna S. 1 Threats or abusive language This place is a f***ing joke. Worst coffee I've ever had.
6 Google 2026-10-11 Daniel Kim 1 Names an employee in a complaint (Jessica) Server named Jessica was rude to my mother and rolled her eyes at us.
7 Amazon 2026-10-07 Karen J. 1 Asks for a refund or chargeback Wrong item - Ordered the dark roast and got decaf. I want a full refund.
8 Amazon 2026-10-02 Pat Wilson 1 Asks for a refund or chargeback Chipped - Mug arrived chipped. Filing a chargeback with my card.
9 Amazon 2026-10-12 J. Smith 1 Check if genuine: 1 star with no details Buyer beware - Do not buy. Total scam.
10 Google 2026-10-09 Kevin O'Brien 1 Check if genuine: 1 star with no details Terrible.

By platform

Platform Reviews Avg stars Drafts Need a person Already answered
Amazon 8 2.8 8 3 0
Etsy 4 4.0 4 0 0
Facebook 2 - 2 0 0
Google 14 3.3 12 4 2
Trustpilot 6 3.0 5 1 1
Yelp 8 2.8 7 2 1

What reviewers talk about

Counted across all reviews, including ones you've already answered. A review counts as a complaint about a topic when the sentence mentioning it is negative (or, with no opinion words, when the review has 1-2 stars).

Topic Reviews Praise Complaints Avg stars Share negative
Food & drink 18 9 4 3.2 22%
Staff & service 10 6 4 3.3 40%
Shipping & delivery 6 2 4 3.5 67%
Damaged or defective 5 0 5 1.4 100%
Price & value 5 2 3 3.8 60%
Wait time & speed 4 1 3 3.3 75%
Cleanliness 3 2 1 3.7 33%
Communication 3 1 2 3.0 67%
Refunds & returns 3 0 3 1.0 100%
Product quality 2 2 0 5.0 0%
Booking & appointments 1 1 0 5.0 0%
Wrong or missing item 1 0 1 1.0 100%

Star ratings

Stars Reviews Share
5 stars 13 31%
4 stars 7 17%
3 stars 3 7%
2 stars 5 12%
1 star 12 29%
No rating 2 5%

Before you post

  • Replace [Your name], [Business name] and [phone or email] in each draft.
  • Drafts only refer to what the reviewer actually wrote. They never offer refunds, discounts or freebies and never ask anyone to change or remove a review. If you want to make an offer, do it privately.
  • replies.md and replies.csv list the flagged reviews first, then everything else newest first.
  • Amazon does not let sellers post public replies to reviews. Use the Amazon drafts as notes for your own follow-up, for example through Brand Registry's Customer Reviews tool.
  • On Yelp you can answer a flagged review with a private message instead of a public comment.

What we read

File Platform Rows Reviews Skipped Columns used
EtsySoldOrders_reviews.csv Etsy 4 4 0 date = 'Date Reviewed', order id = 'Order ID', rating = 'Star Rating', reviewer = 'Reviewer', text = 'Message'
amazon_reviews_q3.csv Amazon 8 8 0 date = 'Review Date', product = 'Product Title', rating = 'Star Rating', review id = 'Review ID', reviewer = 'Author', text = 'Body', title = 'Title'
google_reviews.csv Google 13 12 1 date = 'Create Time', product = 'Location', rating = 'Star Rating', reply = 'Reply', review id = 'Review ID', reviewer = 'Reviewer', text = 'Comment'
my_reviews_sheet.csv from its Platform column 5 5 0 date = 'Date', platform = 'Platform', rating = 'Stars', reply = 'Response', reviewer = 'Name', text = 'Review'
trustpilot.csv Trustpilot 6 6 0 date = 'Review Created (UTC)', product = 'Location Name', rating = 'Review Stars', reply date = 'Company Reply Date (UTC)', review id = 'Review ID', text = 'Review Content', title = 'Review Title'
yelp_export.csv Yelp 8 8 0 date = 'Date', rating = 'Rating', reply = 'Owner Response', reviewer = 'Reviewer Name', text = 'Review Text'
  • 42 reviews after removing duplicates.
  • 1 row skipped because it had no rating and no review text.
  • 1 exact duplicate row removed.
  • 1 date could not be read and is left blank.

Limits: topics and flags come from keyword matching, not a person reading every review, so sarcasm or unusual wording can be misread. Ratings on 10- or 100-point scales are converted to 5 stars. Skim every draft before posting.

← Back to Review Responder