What is a knowledge base?

Knowledge base, explained

A knowledge base usually has two faces. The public side is a help centre customers can browse or search. The internal side holds content for the team: edge cases, escalation rules and wording for sensitive situations. Many businesses start with a handful of FAQs and grow from there.

Good knowledge base articles answer one question each, lead with the answer, use the words customers use, and have an owner and a review date. Policies such as returns, shipping and warranties change, and a stale article does more harm than a missing one because it produces confident wrong answers.

AI has raised the stakes. When a support agent answers by retrieving from your knowledge base, every gap and every outdated sentence becomes an automated answer at scale. The upside is that fixing one article can fix many future conversations at once.

The fastest way to improve a knowledge base is to work backwards from real questions: look at what customers asked that the content did not cover, and write those articles first.

How to measure it

Track coverage (share of real customer questions answered by an existing article), article freshness (share reviewed within your chosen window), and the knowledge gaps your support tool reports.

Common mistake

How SuperBot relates

SuperBot trains on your knowledge sources (URLs, pasted text and documents) and published help-centre articles, and its analytics surface knowledge gaps so you know which articles to write next.

• Knowledge gaps: the AI support metric that matters
• How to train an AI agent without losing brand voice
• SuperBot features

Related terms

• Self-service supportSelf-service support lets customers find answers and complete tasks on their own, without contacting a support agent, using tools such as a help centre, FAQs, order tracking pages, account settings and AI chat. Done well, it is faster for customers and cheaper for businesses; done badly, it becomes a barrier between customers and help.
• Knowledge gapsKnowledge gaps are the customer questions that a support team's content, and therefore its AI agent, cannot answer. They show up as fallbacks, low-confidence answers, escalations and repeated rephrasing. Tracking them tells you exactly which help articles to write or fix next, making them one of the most actionable metrics in AI support.
• Retrieval-augmented generation (RAG)Retrieval-augmented generation (RAG) is a technique in which an AI system first searches a trusted knowledge source for passages relevant to a question, then gives those passages to a language model to write the answer. In customer support, RAG keeps replies tied to the business's current policies and help content instead of the model's general training data.
• Grounding and citationsGrounding is the practice of making an AI system base its answer on specific, approved source material rather than its general training data. Citations make that grounding visible by showing which source supports the answer. Together they let customers and reviewers verify a reply and make wrong or unsupported answers easier to spot.
• Deflection rateDeflection rate is the percentage of customer questions answered by self-service, such as a chatbot, help centre or FAQ, without creating a ticket or reaching a human agent. It estimates how much work automation keeps away from the support team, but it does not by itself prove the customer's problem was solved.