Study: 58 Percent of Voice Search Results Come from Featured Snippets

A new study by SEO platform Semrush analyzing 100000 voice search queries across Google Assistant Siri and Alexa has found that 58 percent of voice search results are sourced from featured snippets in traditional search results.

Key Findings

The study reveals the primary source mechanisms for voice answers. 58 percent of voice results come from featured snippets. 21 percent come from knowledge graph data. 14 percent come from local business listings. 7 percent come from other sources including video and app content. Google Assistant uses featured snippets most frequently at 67 percent. Siri uses featured snippets 52 percent of the time. Alexa relies on featured snippets 45 percent of the time as it also uses Bing results and Yelp data.

Content Type and Length

The study also analyzed the content characteristics of voice answers. The average voice answer is 41 words long. Paragraph-style answers average 43 words. List-style answers are read in their entirety 34 percent of the time with only key items read otherwise. The average reading time for voice answers is 15 seconds. Answers that begin with a direct statement or definition are preferred over those that begin with context or background.

Implications for Voice Search Strategy

Featured snippet optimization is the single most important factor for voice search visibility. Content should provide direct answers of approximately 40-50 words after question headings. List-based answers should present the most important items first as not all items may be read aloud. Local businesses should focus equally on Google Business Profile optimization as 14 percent of voice results come from local data. Building knowledge graph presence through structured data and entity optimization can contribute to the 21 percent of results from knowledge graph sources.