> For the complete documentation index, see [llms.txt](https://docs.smaq.io/documentation/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.smaq.io/documentation/features/knowledge-base.md). # Knowledge Base The Knowledge Base is the AI's memory of your business. When you chat or run an automation, the AI consults the KB to ground its answer in your specific context — your industry, your campaigns, your business model. A weak KB → generic answers. A rich KB → answers that sound like they came from someone who knows your business. > **\[Screenshot needed]** Knowledge Base page with the Business Profile section at the top and a Campaigns table below showing ingested Google Ads + Meta campaigns. ### What's in the KB #### Business profile The basic context about your business: * Business name and website * Industry / sector * Target audience description * Key competitors * Notes (free-text for anything else) Initially populated during [onboarding](/documentation/get-started/quick-start/onboarding-wizard.md). Edit anytime from the KB page. #### Campaigns Auto-ingested from your ad platform connections. Click **Fetch Google Ads campaigns** or **Fetch Facebook campaigns** to pull the current campaign list with names, statuses, spend. This becomes searchable context for chat — when you ask "How is our Brand campaign doing?", the AI checks the KB to know what "Brand campaign" refers to. #### Goals & strategy notes Free-text notes about what you're trying to achieve. Use this for: * Strategic context the AI can't infer ("We're shifting from acquisition to retention in Q3") * House style for content ("All copy targets a 8th-grade reading level") * Constraints ("Don't recommend pause on the Brand campaign — it's protected for legal reasons") ### Why this matters for chat The chat AI uses the KB at every turn. The richer it is, the better the answers: * **"How is campaign X performing?"** — without KB, the AI guesses what X is from your campaign names. With KB, it knows X is your D2C launch from April. * **"What should I try next?"** — without KB, generic suggestions. With KB, suggestions grounded in your business model and stated goals. * **"Write a recap for the client."** — without KB, generic phrasing. With KB, language tuned to your client's seniority and your existing relationship. ### Why this matters for automations [Daily Brief](/documentation/features/automations/daily-brief.md), [Dashboard Summary](/documentation/features/automations/dashboard-summary.md), and the agentic automations ([Media Plan Intelligence](/documentation/features/media-plans/intelligence.md), [Campaign Brainstorm](/documentation/features/automations/reference.md#campaign-brainstorm)) all consult the KB. The narrative they write is shaped by what's in there. ### Editing the KB Go to **Knowledge Base** in the side menu. Edit any section directly. > **\[Screen video needed]** Editing the business profile: changing the industry, adding two competitors, and updating the goal notes. \~45 seconds. ### Tips * **Spend 15 minutes on it once a quarter.** Bigger ROI than most things you can do in SMAQ. * **Add constraints, not just goals.** "Don't recommend pausing Brand" prevents the AI from making bad recommendations. * **Include competitor names**, not just URLs. The AI uses these when running [competitor intelligence](/documentation/features/creative-intelligence.md). * **Update after major changes** — new product launch, new campaign, change of strategy. ### What the KB is NOT * It's not a document store. Don't paste PDFs or long docs — they won't be processed. * It's not auto-updating beyond campaigns. Strategy notes are manual. * It's not a chat memory. The KB is reference material; chat history is in [Review historical user messages](/documentation/organization/review-historical-user-messages.md). ### Privacy * The KB is project-scoped. KB content doesn't leak across projects in your org. * Owners and Editors can edit. Other roles can view. * KB content is used by SMAQ's AI, which runs on Anthropic's Claude. See SMAQ's privacy policy for how AI processing handles your data. ### What's next * [Chat with SMAQ AI (tutorial)](/documentation/get-started/tutorials/chat-with-smaq-ai.md) — see what a rich KB enables in chat. * [Playbooks](/documentation/features/chat/playbooks.md) — formal analyses that complement the informal KB. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.smaq.io/documentation/features/knowledge-base.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.