Knowledge Base
Guide to training your team's assistant using the knowledge base
Overview
The Knowledge Base is where your AI assistant transforms from a helpful tool into an expert employee. Every snippet you add becomes instantly available to all your team's assistants when drafting emails, answering customer questions, and handling conversations.
What to Add to Your Knowledge Base
Your assistants will search this database whenever they need to answer questions or draft responses. The more comprehensive it is, the better they perform.
Product & Service Knowledge
Insurance policy types, coverage details, carrier guidelines
Plan comparisons and pricing structures
Coverage limits, deductibles, and exclusions
Product specifications and features
Service offerings and deliverables
Email Templates and Writing Patterns
Add email templates for the AI to reference in certain situations
Common scenarios ie. "When someone asks Z, write X in Y way"
Common Customer Questions
FAQ responses (pre-approved language)
Policy explanation scripts
Claims process walkthroughs
Coverage scenario examples
Eligibility requirements
Sales & Objection Handling
Competitive differentiators
Objection handling frameworks
Pricing justification talking points
Value propositions for different customer types
Case studies and success stories
Company Processes
Onboarding procedures
Claims submission steps
Policy change workflows
Escalation protocols
Service level agreements
Compliance & Regulatory
Required disclosures
Approved terminology
Legal language for contracts
State-specific requirements
Do-not-say guidelines
Internal Resources
Product sheets and one-pagers
Rate cards and pricing guides
Referral programs
Partner information
Contact information for specialists
Getting Started
Navigate to: Knowledge Base
You'll see your (likely empty) knowledge base ready for training data.
Adding Knowledge Snippets

To add a snippet:
Click "Add Content"
Choose Text Content
Give it a clear, descriptive name
Paste or write the information in the content field
Save
Writing Effective Titles
Your AI searches by title first, so clarity matters.
Good titles:
"Homeowners Policy - Water Damage Coverage Explanation"
"Objection Handling - Price Too High Response"
"Auto Insurance - Multi-Car Discount Details"
"Claims Process - First Notice of Loss Steps"
Weak titles:
"Policy Info"
"FAQ #7"
"Insurance Stuff"
"Important Document"
Pro tip: Include the topic category in the title so the AI knows when to reference it.
Content Best Practices
Be Specific and Complete
Don't summarize—give the full information. Your assistant will pull exact language from snippets.
Instead of: "We have good hurricane coverage" Write: "Our homeowners policies include comprehensive hurricane coverage with up to $500K in dwelling protection, additional living expenses if your home becomes uninhabitable, and debris removal up to policy limits. Unlike some carriers, we don't have a separate hurricane deductible—your standard deductible applies."
Use Approved Language
If there's specific wording required for compliance or consistency, capture it exactly in your knowledge base.
Include Context
Help the AI know when to use this information.
Example:
Title: "Objection Handling - Competitor Offers Lower Rate"
When to use: Customer says a competitor quoted them less
Response framework:
"I appreciate you doing your research. Price is definitely important, but let's make sure we're comparing apples to apples. A few questions:
- What coverage limits did they quote?
- What's their deductible?
- Did they mention their claims satisfaction rating?
Our rates reflect our A+ AM Best rating and 98% claims satisfaction. We've also got a price-match guarantee—if you find truly comparable coverage for less, we'll match it. Can I review their quote with you to see what we're working with?"
Key points to emphasize:
- Our financial strength rating
- Claims satisfaction
- Price-match guarantee
Structure for Readability
Use bullet points, headings, and clear formatting so your AI (and humans reviewing snippets) can quickly scan.
Finding What to Add
Not sure where to start? Use CustomerIQ to tell you what's needed.
Use CustomerIQ's Highlight Views
Navigate to Highlights and filter for:
Questions customers ask frequently
Topics where reps hesitate or give inconsistent answers
Objections that come up repeatedly
Product areas generating the most inquiries
These are your highest-impact training opportunities.
Ask Your Team
Your experienced reps know what questions they get asked daily. Have them spend 30 minutes brain-dumping common scenarios into snippets.
Review Recent Conversations
Look at emails and calls from the past week. What information did reps have to look up or wasn't readily available? Add it.
Mine Existing Resources
Do you have product sheets, training decks, or FAQ documents already? Don't recreate—copy the relevant sections into snippets.
The Evolution from Assistant to Expert Employee
Week 1: Basic Assistant (10-20 snippets)
Core product information
Top 5-10 FAQs
Basic objection responses Your assistants can handle routine questions with pre-approved language.
Week 2-3: Knowledgeable Assistant (30-50 snippets)
Expanded FAQs
Process walkthroughs
More objection scenarios
Product comparisons Your assistants rarely need to defer to humans for standard inquiries.
Month 1+: Expert Employee (100+ snippets)
Comprehensive product knowledge
Edge case scenarios
State-specific information
Competitive intelligence
Deep objection handling Your assistants can handle complex, nuanced conversations with confidence.
This is the transformation that makes CustomerIQ the most flexible revenue platform available. You're not just configuring software—you're training virtual employees who know your business as well as your best reps.
Best Practices
Start with High-Impact Information
Focus on the questions your team gets asked most often. These snippets will be referenced immediately and deliver instant value.
Add Continuously, Not All at Once
Don't try to document everything in one sitting. Add 5-10 snippets per day as you notice gaps.
Review and Update Regularly
Products change, policies update, competitors shift. Set a monthly reminder to refresh outdated snippets.
Get Team Input
Your reps are in the conversations daily. They know what information would be helpful. Make adding snippets a team effort.
Test Your Snippets
After adding knowledge, monitor how assistants use it. Are they pulling the right information at the right times? Refine as needed.
Common Mistakes to Avoid
Vague or incomplete information "We offer competitive rates" doesn't help your assistant. "Our homeowners rates average 15% below market in Tennessee, with discounts up to 25% for bundling" does.
Forgetting about discoverability Even great information is useless if the AI can't find it. Use clear titles and comprehensive tags.
Writing for humans only Your snippets need to work for both AI search and human reading. Include key terms the AI might search for.
Copying jargon-heavy documents verbatim Translate internal documentation into language your assistants can use in customer conversations.
Not leveraging existing conversations You have a goldmine of data in your CustomerIQ Views. Use it to identify what knowledge is actually needed.
Adding everything at once, then never updating The knowledge base should be living and growing. Schedule regular additions and updates.
What Happens Next
Immediate impact: As soon as you add a snippet, all your team's assistants can reference it when drafting emails or answering questions.
You'll notice:
More consistent messaging across your team
Faster response times (less searching for information)
Better handling of complex customer questions
Fewer escalations for routine inquiries
Continuous improvement: The first 1-2 weeks will involve the most additions as you build your foundation. After that, you'll add snippets as gaps emerge.
Measuring Success
Good indicators your knowledge base is working:
Assistants draft complete responses without needing human input
Reps report feeling confident in AI-suggested answers
Customer satisfaction improves (faster, more accurate responses)
Team messaging becomes more consistent
Signs you need more training data:
Assistants frequently say "I don't have enough information"
Reps heavily edit AI drafts for accuracy
Same questions get asked repeatedly but assistants can't answer them
Inconsistent responses across team members
Last updated