Use Ask Effectively

Ask (/ask) answers questions from published stories. It also uses the internal glossary to interpret team jargon and acronyms without replacing story citations.

Use Ask for fast retrieval, then verify claims in source stories before external use.

Prompt pattern that works

Use this structure:

1. Context: segment, stage, or persona.
2. Need: what evidence you want.
3. Output format: bullets, quote-first, or short narrative.

Example:

"I am preparing a security-review call with an engineering leader in fintech. Give me 3 relevant proof points with one metric and one quote each."

Good Ask use cases

• Finding comparable outcomes quickly.
• Generating first-draft talking points.
• Spotting which stories to read deeply.
• Translating internal terms/acronyms into customer-facing proof.

Glossary-aware prompts

When you use team shorthand, Ask expands the query with glossary definitions in the background.

Example:

"Give me proof points for brownfield modernization and platform guardrails in enterprise rollouts."

If Ask detects glossary terms, it will show glossary chips in the response so you can trust how your prompt was interpreted.

What Ask is not for

• Final source-of-truth claims without checking original stories.
• Internal confidential analysis from unpublished content.
• Replacing judgment about deal fit.

Validation loop

After Ask returns results:

1. Open cited stories.
2. Verify metric wording and quote attribution.
3. Adjust for your prospect context.
4. Save top matches into a collection.

Troubleshooting weak results

• Add more context (industry, persona, stage, objection).
• Ask for fewer but higher-confidence examples.
• Use library filters directly if Ask seems too broad.