Examples
Example 1: Reusing a Proof Point Before a Pitch
An account manager is putting together a pitch for a prospective SaaS client and wants the agency's strongest evidence for paid social, without hunting through old decks.
Step 1: Ask. On the Knowledge page, the account manager types:
What are our strongest proof points for paid social campaigns?
Step 2: Answer with sources. Within a moment, Knowledge Base returns a short written answer naming three results: a campaign that cut a client's cost per lead, a follower-growth figure for another client, and a quote from a client about return on ad spend. Beneath the answer, under a "Sources" heading, it lists where each one came from: a case study, a campaign report, and an approved testimonial.
Step 3: Verify. The account manager opens the testimonial source to confirm the exact wording and that it is approved for outside use, then copies the figure and the quote into the pitch with confidence that both are real and attributable.
Step 4: Follow up. Still in the same conversation, the account manager asks:
Which of those are approved for public use?
Knowledge Base answers against the same thread and confirms which proof points are cleared for client-facing material.
Example 2: Checking Brand Voice Before Drafting
A content team member is about to write a LinkedIn post for a client and wants to match the brand's established voice rather than guess.
Step 1: Ask.
What does our content library say about brand voice for the SaaS client?
Step 2: Answer. Knowledge Base pulls from the client's brand brief and past approved content. It returns a plain summary: the brand favors short sentences, a confident but not boastful tone, no jargon, and a habit of opening with a specific customer moment. Underneath, it lists the brand brief and two recent approved posts as its sources.
Step 3: Draft with the rules in hand. The team member opens the brand brief from the source list to read the full guidance, then drafts the post knowing it lines up with how the brand actually sounds.
Example 3: An Honest "I Don't Know"
An account manager wants to know whether a particular client has a documented policy about discount language in email.
Step 1: Ask.
What is our policy on using discount language in emails for the retail client?
Step 2: A plain answer. Knowledge Base does not find anything on this in the connected knowledge. Instead of guessing, it replies plainly that it could not find anything on that and suggests rephrasing or broadening the terms. No sources are listed, because there are none.
Step 3: What it means. The account manager reads this correctly: there is no documented policy in the connected knowledge yet. Rather than inventing one, the account manager raises it with the client and asks the team to add the agreed policy to the agency's documents, so the next person who asks gets a real answer.
Example 4: A Count Across Your Library
An agency owner wants a quick read on how much vetted evidence the firm has, broken out by type.
Step 1: Ask.
How many proof points do we have by category?
Step 2: Exact figures. Knowledge Base recognizes this as a count question. The agency has its proof points available as a table the owner is allowed to see, so it returns the exact breakdown: a figure for client outcomes, one for quotes and testimonials, and one for statistics, each as a precise number rather than a rough estimate. The source of the figures is shown beneath.
Step 3: Act on it. Seeing the testimonial count is low compared with outcomes, the owner asks the team to prioritize gathering a few more client quotes, so the proof library is balanced before the next round of pitches.