{"id":1215,"slug":"rankbrain-2015-bert-oct-2019-mum-2021-ml-ranking","title":"Machine-learned language understanding in ranking: RankBrain (confirmed October 2015), BERT (October 2019, ~10% of English queries), MUM (announced 2021)","kind":"reference","scope":"business","status":"current","audiences":["kevin","candid-team"],"topics":["seo-content","algorithm-updates"],"reference_body":"**Claim:**\n\n- **RankBrain** — machine-learning ranking component, publicly confirmed **October 2015** (live since ~early 2015). Google called it one of its top-three ranking signals at the time. Helps interpret never-before-seen queries.\n- **BERT** — transformer-based natural-language understanding, **October 2019**, initially affecting **~10% of English queries**.\n- **MUM** — announced **2021**.\n\n**Source:** Google + Bloomberg (RankBrain, October 2015); Google blog (BERT, October 2019); Google I/O 2021 (MUM).\n\n**Confidence:** Verified (dates + descriptions); **Industry-consensus** on relative importance.\n\n**Why this matters for Candid:** Reinforces the Hummingbird lesson — write naturally, answer real intent. You cannot directly \"optimise for BERT\"; you optimise by being clear and genuinely useful. Strengthens the case for [[rule-answer-customer-questions-in-first-100-150-words-for-ai-citation]].","rationale_body":null,"metadata":null,"links":{"outgoing":[{"slug":"hummingbird-sep-26-2013-semantic-search-90pct-queries","title":"Hummingbird (announced September 26, 2013) — near-total rewrite of the core algorithm; shifted from keyword matching to query-intent understanding; affected ~90% of searches","kind":"reference","scope":"business","link_type":"depends-on"}],"incoming":[{"slug":"research-brief-businesses-found-on-web-google-smb-june-2026","title":"Research brief: How businesses got found on the web — a then→now timeline (Google-focused, SMB lens, June 2026)","kind":"reference","scope":"business","link_type":"relates-to"}]},"created_at":"2026-06-18T00:10:55.835Z","updated_at":"2026-06-18T00:10:55.835Z"}