Cookbook/Reviews

=Publishing Review Data=

Overview
Relevance for Google and Yahoo Available vocabularies - Google and rev

Note: The Google documentation for RDFa currently uses the wrong property v:rating instead of v:average for the average rating, which can lead to incorrect star ratings in the preview. We have already asked them to fix this.

Important: The min/max values refer to the scale used for ratings in your shop system, NOT for the lowest/highest actual rating. So if you have zero reviews for an item and you are using a scale from 0 to 5, it would be


 * min (review:minRating) = 0
 * max (review:maxRating) = 5
 * mean (review:rating and v:average) = 0
 * number of reviews (review:totalRatings and v:votes = 0

Note: Google has clarified that they accept the v:count' property only if the respective reviews actually exist on the page, i.e. that there is the text of the review available. Thus, use v:votes instead in order to publish the number of people who at least rated the product (i.e., assigned a star rating without necessarily writing a textual explanation). If you want to use v:count, you must also mark-up and make available the individual review text on the same page.

RDFa


Product Reviews:

Average: 4.5, avg.: 0, max: 5 (count: 45 )

Recommendation 1: In your HTML template, you should include the review markup only if there is review data for this article, because Google will otherwise display zero stars, which may convey a worse reputation than no star rating.

The best approach is a conditional pattern, which could look like

{% if {review_count}>0 %} ... markup for the review data ... {% endif %}

Recommendation 2:' You may feel tempted to suppress bad ratings, i.e. by a condition in the template to add the data markup only for ratings above 3, like {% if {average_rating}>3 %} ... markup for the review data ... {% endif %}

Don't do this! Google may use the distribution of ratings across your site as an important parameter to judge the content quality of your site, and this may make Google think that your rating data is faked.