Voice Search

Voice search is searching by speaking a question to a device, such as a smartphone, smart speaker, car system or smart TV, instead of typing it. The assistant converts speech to text, finds an answer and often reads a single result aloud, which makes being the chosen answer especially valuable.

Written by Paul Hoda, SEO consultant since 2005. Last updated 8 November 2026.

How voice search works

A voice search has three main steps:

  1. Speech recognition: the device turns spoken words into text.
  2. Understanding: the assistant works out the meaning and intent of the question, often using context such as location.
  3. Answering: it finds a response from a search engine, a knowledge base, a business listing or an app, then speaks it or shows it on screen.

Common voice assistants include Apple’s Siri, Amazon’s Alexa and Google’s assistants. Google has been replacing Google Assistant with Gemini on many devices, which brings more AI-generated, conversational answers to voice.

How voice queries differ from typed ones

  • Longer and more natural: people say "What time does the nearest chemist close tonight?" rather than typing "chemist opening hours".
  • More questions: voice searches often start with who, what, where, when, why or how. See question keywords.
  • Strongly local: many voice searches are about nearby businesses, directions and opening times, similar to near me searches.
  • One answer, not ten: on a smart speaker, the user usually hears a single response.

Where voice answers come from

Sources vary by assistant and question. Search-based answers have often been drawn from content shown in featured snippets or the Knowledge Graph. Local answers draw on business listings such as Google Business Profile, Apple Business Connect and Bing Places, depending on the device. Newer AI assistants may write their own summary from several web sources.

How to optimise for voice search

  1. Answer questions concisely. Give a clear answer in one or two sentences near the top of a section, then expand. Short, direct answers suit being read aloud.
  2. Use natural language. Write headings as real questions people ask, and answer in plain English.
  3. Get local listings right. Complete and accurate business profiles, with correct hours, address, phone number and categories, are essential for local voice queries.
  4. Keep facts consistent. Make sure your name, address and phone number match everywhere, an idea known as NAP consistency.
  5. Add structured data. Schema markup such as LocalBusiness helps assistants understand your details.
  6. Make pages fast and mobile-friendly. Most voice searches happen on phones.

Myths and limits

  • There is no separate voice ranking report. Search Console does not split out voice queries, so measure indirectly through question queries, local actions and calls.
  • Voice search is not a separate discipline. It builds on good SEO, local SEO and clear content.
  • Speakable schema is limited. It remains a beta feature aimed at news publishers and is not a general voice ranking tool.

Key Takeaways

  • Voice search means speaking a query to a phone, speaker or other device.
  • Voice queries are longer, more conversational and often local.
  • Assistants usually give one answer, so concise, direct answers matter.
  • Accurate business listings are essential for local voice searches.
  • Voice search builds on good SEO rather than replacing it.

Related Terms

Wondering what to fix first on your site? Get a free check of your site’s SEO, or keep learning in the SEO Encyclopedia.