What is UGC (user-generated content)?

In social media marketing, "UGC" has come to mean the videos a brand commissions from creators so that they look like they came from customers. This is not that. Here UGC is what your customers actually write and upload on your site: the star-rated review, the photo of the backpack on the mountain, the question about sizing and its answer. It matters to you for two reasons. It is text about your products in the words of the people who buy them, which are not the words of your catalog and sometimes sound more like what the next customer will ask. And even though you did not write it, it is published on your site: what it says, what it links to and how you present it are your business.

What it includes

In an online store, UGC usually comes in four forms:

• Reviews and ratings. A customer's text and score about a product. These are product reviews, and when they are described for machines they use the Review type for each one and AggregateRating for the average.
• Customer photos and videos. The product in use, with real light and real size.
• Questions and answers. A customer asks on the product page and the store or another customer answers.
• Comments and forums. On the store's blog or in a community of its own, where users can leave links.

And there is UGC about your store that is not on your site: reviews on other platforms, forum threads, videos. You do not control it, but it gets read too.

How search engines and assistants treat it

Google accepts it as content on your site, with rules. In its general guidelines for structured data, it asks you to provide original content that you or your users have generated, and not to mark up misleading content, such as fake reviews.⁠^{[1]Google Search Central. General structured data guidelines} For links, it recommends marking links left by users, such as those in comments and forums, with rel="ugc", and suggests removing that attribute for contributors who have been adding quality content for a while, as a way of recognizing them.⁠^{[2]Google Search Central. Qualify your outbound links to Google}

Users' spam counts as a problem on your site. Google's spam policies have a section of their own for content that users add through a channel meant for them, and Google adds that the site owner often does not know it is there. Its examples are spam posts in forum threads and spam in blog comments.⁠^{[3]Google Search Central. Spam policies for Google web search} Its guide to preventing it suggests publishing an abuse policy and letting trusted users report content, adding the noindex tag to what new users with no reputation post until they earn one, using nofollow or ugc on the links in untrusted content, and manually moderating suspicious interactions, which is a workload but very effective.⁠^{[4]Google Search Central. Prevent user-generated spam on your site and platform}

Assistants read reviews, including those from other sites. OpenAI explains that, when choosing which products to show in its shopping results, ChatGPT may take into account the available options, price and reviews, among other things. It may also show review summaries, generated by the model from reviews on public websites, to highlight what people like most and least about a product, and OpenAI warns that it does not verify reviews or ratings.⁠^{[5]OpenAI, Help Center. Shopping with ChatGPT Search} So what your customers say about a product, on your site and off it, can end up summarized in an answer that neither you nor they wrote.

In a store

The product page of a backpack with four blocks of UGC. It is a fictional example: the store, the texts and the links are made up to show what each block asks for.

The reviews notice and the link in an answer, as they would look in the HTML:

<p class="review-notice">
  These reviews are written by customers who bought the product in our store. We check them
  before publishing to remove spam and personal data, and we publish the negative ones too.
</p>

<p>
  I use it with a case from this brand:
  <a rel="ugc" href="https://other-site.example/laptop-sleeve">case</a>
</p>

Each store decides the exact wording of the notice according to what it actually does; one that promises a verification that does not exist is worse than none. The link carries rel="ugc" because a customer put it there, not the store.⁠^{[2]Google Search Central. Qualify your outbound links to Google}

How to check it

1. List every place where your customers can post: reviews, photos, questions, blog comments, profiles.
2. Open the source code of a product page with questions and answers and look at which attribute the customers' links carry: ugc or nofollow.⁠^{[2]Google Search Central. Qualify your outbound links to Google}
3. Search your oldest questions and comments for links or text that have nothing to do with the product.⁠^{[3]Google Search Central. Spam policies for Google web search}
4. Check that your product pages say whether reviews are verified and how they are handled, and that what they say is what you do.
5. Ask ChatGPT about one of your products and read the review summary, if it shows one: it is what a buyer sees of your customers without ever visiting your site.⁠^{[5]OpenAI, Help Center. Shopping with ChatGPT Search}

Related concept: product reviews

Product reviews are one part of UGC, the most visible, but not all of it. UGC also includes photos, questions and answers, and comments, which carry no rating and do not feed an average, and which have rules of their own: the links they contain, who moderates them and whether they get indexed. Every customer review is UGC; not all UGC is a review. For what is specific to reviews, their markup and their average, see the entry on product reviews.

To keep going, the catalog and entities reading path orders the rest of the terms, and the entry on price and availability signals covers the other part of the product page that assistants read most carefully.