Spam, privacy and profanity filters
With auto-publish on, these checks stop risky reviews from going live on their own. A review that trips one is held as pending for you to look at. The filters never change what a customer wrote.
All of them are in Settings › Publishing and moderation. They check reviews from your storefront and from request emails.
The filters
| Filter | What it catches | Default |
|---|---|---|
| Spam filter Filter out spam reviews and questions | Links (anything with http://, https:// or www.) and code such as script tags. | On |
| Personal information filter | Email addresses and phone numbers. Choose Hold for approval or Publish without filtering. | Hold for approval |
| Profanity filter Block reviews with profanity words and phrases | A built-in list of swear words. | On |
When auto-publish is off, every review is pending anyway, so the filters have nothing to hold.
A held review is still pending, and pending reviews publish themselves after 14 days. Check the Pending view regularly, or turn on moderation alerts in Admin notifications.
Reject instead of hold
Reject unsafe submissions before saving is off by default. When it is on, a review that trips any filter you have switched on is refused, and never saved. The shopper sees a general error and can try again.
With rejection on, Custom blocked terms also apply: up to 20 words or phrases of your own, comma separated, each up to 40 characters. They match whole words, in any case. Without rejection, blocked terms are not checked.
Rejecting loses the review for good, and the shopper is not told why. Holding is usually the better choice.
Questions
Questions get their own light spam check, whatever these settings say. You moderate them under Customer questions.