Skip to main content

    Speakable Schema

    Speakable schema is a Schema. org structured data type (currently in beta with Google) that identifies sections of a webpage particularly suited to being read aloud by voice assistants and other voice-enabled surfaces such as Google Assistant, Alexa, and Siri. Pages mark specific elements as Speakable using CSS selectors or XPath inside JSON-LD. Why it matters for AEO and voice search: Voice-first answer surfaces — Google Assistant, in-car voice search, smart speakers, and increasingly the voice modes of ChatGPT and Gemini — preferentially read content marked as Speakable. For brands building authority in voice and AI assistant surfaces, Speakable schema on key takeaway sections, FAQ answers, and direct definitional content is one of the cheapest investments with the longest tail of payoff as voice queries grow. Smart Money Media's pillar guides mark Key Takeaways and direct answers as Speakable for this reason.

    Why Speakable Schema matters

    This technical layer ensures that voice assistants grab the most relevant, natural-sounding snippets instead of stumbling over sidebars or image captions. Without it, a smart speaker might recite navigational breadcrumbs or technical metadata, destroying the user experience and decreasing the chances of a brand becoming a primary voice source.

    In practice

    The BBC uses JSON-LD Speakable properties to highlight lead summaries, allowing Google Assistant users to hear a 30-second audio digest of the morning headlines.

    Common mistake

    Restricting markup to the entire body text rather than granular, punchy sentences that avoid the repetitive verbal clutter of headers and navigation menus.

    How it connects

    This practice bridges the gap between traditional Answer Engine Optimization and the evolution of multimodal Large Language Models.

    Frequently Asked Questions

    What is Speakable Schema?

    In short: Speakable Schema is speakable schema is a Schema. See the full definition above for context.

    Can any website use this markup effectively?

    Google specifically requires that Speakable markup only be applied to news-oriented sites registered in the Publisher Center. While non-news sites can technically add the code, the actual text-to-speech functionality in localized search results is currently a priority for verified journalistic entities.

    What is the ideal character count for marked-up sections?

    Smart speakers generally prioritize the first 650 characters of marked content to ensure the user receives a concise answer. Content exceeding this length may be truncated or ignored by the voice assistant to prevent long-winded, non-conversational responses.

    How does this differ from standard Article schema?

    While standard schema helps search engines understand data, Speakable specifically targets the audio output layer for hands-free devices. It essentially acts as a direct instruction to the TTS engine on which exact paragraphs provide the most auditory value.

    If You're Invisible in AI, You're Losing Clients Right Now.

    See exactly how your company appears across AI, search, and investor research — and uncover the hidden gaps costing you trust and deals.