‹ Marketing recipesMarketing & Positioning · Agent Blueprint
Case study production
A customer story composed from what they said, not what we wished they'd said.
Side by side: token usage, with and without Paperbase
~125k tokens saved · 78% less work for the agentWithout Paperbase
~160k tokens
- Ask the account team for the story.
- Skim the customer's emails and calls for what actually happened.
- Reconstruct the timeline from sales and CS memory.
- Write the case study from the highlight reel, not the record.
- Run it past the customer twice for inaccuracies.
With Paperbase
~35k tokens
- Pull the account's Interactions in chronological order.
- Pull the outcomes achieved, with evidence.
- Extract the customer's own language about the result.
- Draft the case study from the actual record.
Agent prompt
You are drafting a case study for {{account_name}}.
Using Paperbase:
1. Pull the account's journey: initial problem, interactions, milestones, and outcomes.
2. Pull the numbers or results with their evidence.
3. Extract the customer's own words about the result.
4. Note anything in the record that would not support a heroic narrative.
Output:
- A case study draft built on the actual record.
- Verifiable outcomes, each cited.
- Customer language quoted, not paraphrased.
- A flagged list of claims the customer might dispute.Placeholders in {double_braces} are inputs the agent will ask for at runtime. Give the prompt to any agent connected to Paperbase, and the rest grounds in your own memory graph.