Definition
Social signals are the engagement metrics content collects on social platforms: likes, shares, comments, reposts, upvotes, views and follower counts, along with how often a brand is mentioned in posts. The term comes from SEO, where people long debated whether search engines use these numbers to rank pages.
The short answer for Google is no, not directly. Google has said for years that likes and share counts are not ranking factors. Much social engagement sits behind logins where crawlers cannot see it, and raw counts are easy to inflate. Correlations between social engagement and rankings exist, but they mostly run through indirect paths. Content that spreads gets seen by journalists and bloggers who link to it, it raises branded searches, and it brings direct visitors. Those links, mentions and searches are what search systems actually respond to.
AI search follows a similar pattern with a twist. Answer engines rarely appear to filter sources by engagement. Promptwatch's February 2026 study of YouTube videos cited by ChatGPT, Perplexity and Google AI Overviews found that 78% of cited videos had fewer than 1,000 likes and that videos under 100,000 views earned nearly 80% of citations. Channels under 100,000 subscribers accounted for about 80% of citations as well. The engines matched video titles, descriptions and transcripts to the question; social proof was not the gatekeeper.
What does reach AI answers is the content of social posts, not their counts. Reddit threads, LinkedIn articles, YouTube videos and Facebook pages are cited as sources (see social media citations), and how often depends heavily on the engine. In early October 2026, social platforms made up 12.2% of Google AI Overviews citations and under 0.5% of ChatGPT's. Brand mentions in social posts also feed the broader picture models build of a brand, through training data and through the third-party pages they retrieve. In that sense the useful 'signal' is what people say and how consistently they say it, which overlaps with third-party mentions and entity consensus.
So where does social activity help visibility in practice?
It helps when it creates citable material: a detailed LinkedIn article, a YouTube tutorial with a clear title and chapters, a Reddit answer that solves a real problem. It helps when it earns coverage and links that search systems do weigh. It helps when it keeps public profiles accurate, since AI Overviews cite Facebook and Instagram pages for basics such as opening hours and menus. It does little when it chases vanity metrics. Buying likes or followers adds nothing that search systems use and can get accounts penalized by the platforms themselves.
For measurement, separate the two things the term mixes together. Track engagement as a channel metric for your social team. Track social citations and social mentions inside AI answers as a GEO metric, per engine, since that is the number connected to AI visibility. Related: content authority, Reddit citations, UGC citations and AI brand mentions.
Examples of Social Signals
- An original salary survey shared on LinkedIn reaches trade journalists, who link to it from 25 articles. Rankings improve because of those links, not because of the 2,000 reactions on the post.
- A niche YouTube channel with 18,000 subscribers publishes a 12-minute tutorial on configuring a specific firewall model. Google AI Overviews and Perplexity both cite it for that query, while a viral general video on the same brand is never cited.
- A brand buys 50,000 Instagram followers to look established. Its AI visibility does not change, and Instagram later removes most of the fake accounts, cutting its follower count sharply.
- A B2B team stops reporting total social engagement to leadership as an SEO metric and instead reports how often LinkedIn and YouTube URLs about the brand are cited in AI answers per engine.
