Introduction to media search and recommendations

This page introduces and describes the capabilities of Vertex AI Searchfor media. The page also provides links to more information,tutorials and checklists, to get you started withVertex AI Search for media.

Vertex AI Search includes two capabilities specific for the mediaindustry:

  • Media recommendations. Get recommendations for media content such asvideos, news, and music. With media recommendations, audiences can discovermore personalized content, like what to watch or read next, withGoogle-quality results customized using optimization objectives.

  • Media search. Get Google-quality search results with advanced query anddocument understanding designed for media content.

Note: Don't be misled by the product name. With Vertex AI Search,you can create recommendations apps as well as search apps.

Key features of media apps

There are many similarities between media apps and custom appsin Vertex AI Search. Here are some key features of mediaapps:

  • Media apps require user events. You upload user events to personalizerecommendations and rank search results for your audience.

  • Media apps require media metadata to conform to a predefined schema orto use a custom schema that contains a minimum set of key properties.

    • Predefined schema. This lets recommendations and search rankinguse Google-defined, media-specific fields such as content ratings,aggregated ratings, persons, and production year to help generate resultsbased on media engagement.

    • Custom schema. The custom schema gives you moreflexibility than the predefined schema. However, your schema fields mustmap to the followingrequired key properties:title,category,uri,media_available_time, andmedia_duration. Thecategory propertymust be an array of strings, and the other four properties are strings.

      In addition to the required key properties, Google recommends that youmap as many other schema fields as possible to thesuggested keyproperties. The suggested key properties represent similar media metadatato that in the predefined schema—for example, content ratings, aggregatedratings, persons, and production year.

  • Media recommendations apps offer you a choice of recommendation type.Media recommendations apps let you choose what kind of recommendation togenerate, such as recommending other content that users might like, similaritems, or the most popular items.

  • Media recommendations apps offer you a choice of optimization objectives.For example, you can decide whether to optimize recommendations forclick-through-rate to increase the number of interactions with content orfor conversion rate to increase the consumption of content.

The following table outlines some functional differences between media andcustom data stores.

Media apps and data storesCustom apps and data stores
Data stores are always structured.Data stores can be of any type(website, unstructured, structured).
Require structured data with apredefined schema or a custom schemawhere you map your data fields tosome required key properties.No key properties are required forstructured data.
For media apps, userevents are required.For custom recommendations, userevents are highly recommended but notrequired.
Imported historical user events arejoined synchronously.Imported historical user events arejoined asynchronously.

For more information, seeAbout media data stores anddocuments andAbout apps and datastores.

Getting started tutorials

If you are new to Vertex AI Search, try out the gettingstarted tutorials. These tutorials guide you step-by-step through the creationof an app. Data (documents and user events) are provided for the tutorials soall you need is a Google Cloud project and a billing account to create yourfirst app:

Checklists

There is a lot of commonality between working with media apps and working withcustom apps, but some features apply only to custom apps andother features only to media apps.

Use the following checklists to guide you through typical workflows specificto media:

Except as otherwise noted, the content of this page is licensed under theCreative Commons Attribution 4.0 License, and code samples are licensed under theApache 2.0 License. For details, see theGoogle Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.

Last updated 2026-02-18 UTC.