How do answer engines work? Are they different to search engines?

What search engines do

• Traditional search engines (like classic Google or Bing) crawl web pages, index them, and rank them using signals such as keywords, links, and user behaviour.​
• The main output is a SERP: snippets, “10 blue links”, sometimes a featured snippet or “People Also Ask”, leaving you to click through and evaluate sources.​

What answer engines do

• Answer engines (e.g. Google AI Overviews, ChatGPT with browsing, Perplexity) are built around large language models that interpret natural-language questions, including long, multi-part prompts.​
• When you ask a question, they:
  • Parse intent and context using NLP.​
  • Fan out multiple related sub-queries to search and other data sources.​
  • Retrieve relevant passages, tables, and entities, then use the model to compose a single, coherent answer with citations.​