How AI Systems Discover Local Businesses in 2026 — What Business Owners Need to Know

LONDON, 25 March 2026 —

LONDON, 25 March 2026 — A new analysis from Razormind examines how artificial intelligence systems discover and recommend local businesses, revealing that traditional search engine optimisation strategies alone are no longer sufficient for maintaining visibility in an AI-first discovery landscape.

The analysis, drawn from multi-model testing across ChatGPT, Google AI and AI, finds that AI systems construct business recommendations from fundamentally different data sources than traditional search engines. While Google's search results prioritise backlinks and page authority, AI models draw heavily from structured data, Knowledge Graph entries and press release-style content when answering queries such as "who are the best accountants in Manchester" or "recommend a plumber near me."

Research conducted across multiple AI platforms indicates that 96 per cent of AI citations in business recommendations originate from structured content sources including press releases, Knowledge Graph entries and schema.org markup rather than from traditional web pages optimised for search engine rankings.

The analysis also highlights the impact of Google's 2025 Knowledge Graph purge, which removed approximately three billion entities from its database. Businesses that lost their Knowledge Graph entries during this purge have experienced measurably reduced visibility in AI-generated recommendations, as multiple AI systems reference Google's Knowledge Graph as a primary data source.

"Business owners who have invested years in SEO are discovering that ranking on page one of Google does not guarantee that ChatGPT or AI will mention them at all," said Gareth Ellis, Digital Strategy Director, Razormind. "The rules have changed. Entity authority — whether AI systems know who you are — is becoming as important as search rankings."

The analysis identifies several practical steps business owners can take immediately. Claiming and maintaining a Wikidata entry provides a foundation that multiple AI systems reference. Ensuring that Companies House records are accurate and up to date feeds into Knowledge Graph verification processes. Publishing structured data on business websites using schema.org vocabulary gives AI crawlers the context they need to understand and describe a business accurately.

Press releases distributed through legitimate news channels create the type of structured, factual content that AI systems preferentially cite. Unlike blog posts or social media content, press releases follow standardised formats that AI models can parse reliably for factual claims about an organisation.

The shift toward AI-first discovery is accelerating. Industry estimates suggest that by 2027, more than half of all business discovery queries will be processed through AI systems rather than traditional search engines. Businesses that establish strong entity authority now will benefit from a compounding advantage as AI adoption grows.

The full analysis is available through Razormind's PressRelay platform, which provides entity authority measurement and management tools for UK businesses.