The Speakable Schema Markup is a powerful tool designed to enhance content by enabling Text-to-Speech (TTS) on Google Assistant-enabled devices. This markup identifies specific sections within an article or webpage that are ideal for audio playback. Through this, search engines and other applications can easily identify sections of the content that should be read aloud to users on Google Assistant devices.
By utilizing Speakable schema, web pages can distribute content through new channels, reaching a wider user base. Specifically, Google Assistant uses this feature to respond to news queries on smart speakers, offering users an audio playback of articles.
When a user asks Google Assistant about a specific topic, the assistant selects three articles from the web and reads aloud sections of these articles that are marked with Speakable structured data using TTS.
PRO TIP: When Google Assistant reads a Speakable section aloud, it credits the source and sends the full URL of the article to the user's mobile device through the Google Assistant app.
Currently, Speakable schema is only available in the US, and only for users with Google Home Devices set to English. Publishers must also publish content in English to use the feature. Google plans to expand this feature to more countries and languages as more publishers implement Speakable schema.
Follow these steps to get started with the Speakable Schema:
Ensure compliance with guidelines: Make sure to follow all necessary guidelines provided by Google.
Add Speakable structured data: Implement Speakable structured data to your webpage content.
Submit for onboarding: Submit your content for review. Google will assess your eligibility for the feature, and this process may take up to 5 business days. You will be notified about the status of your eligibility.
In addition to Structured Data and Webmaster Guidelines, here are the key guidelines to consider when using Speakable Schema:
Avoid adding Speakable structured data to content that may sound confusing when read aloud in voice-only scenarios, like photo captions, source attributions, etc.
Focus on clear, concise content that offers key points to ensure users can follow the story easily. Avoid cluttering the content with unnecessary details that might confuse the listener.
Content must be structured with precise headlines or summaries that provide clear, useful information to users.
For top stories, break up the content into single, clear sentences to make it easier for TTS systems to read.
Aim to keep content sections between 20-30 seconds or around 2-3 sentences for optimal audio user experience.
The Speakable Structured Data uses an Article or Webpage object. You can define the following properties for the content:
cssSelector: This defines the specific section of content that should be marked as "speakable" by referencing it using CSS selectors.
Xpath: This can also address content by using XPath, assuming the content follows an XML structure. You may use either cssSelector or Xpath, but do not use both at the same time.
If your content is not triggering on Google Assistant using TTS audio, try these commands:
"What is the latest news about [topic]?"
"What is the latest news on [topic]?"
"Play news about [topic]."
If your content is still not triggering, it could be due to the algorithmic ranking system that determines which articles Google Assistant will choose. Typically, Google Assistant will select up to three articles from different sources for audio playback.
Visit the Google Webmaster Central Help Forum: If you're encountering issues not covered in this documentation, this is a helpful resource.
Contact Google representative: Reach out to your Google representative or audio news support for further assistance.
You can also explore additional audio solutions provided by Google Assistant for your news content, including more advanced integrations. For custom applications, consider checking out the Actions on Google Developer Guide.