daBongo LMS AI Training Courses

Effective Prompting for Perplexity AI

Lesson 2: Context, Constraints, and Structured Output

Log in and enroll to track lesson completion.

Lesson Objectives

By the end of this lesson, students should be able to:

  • Build layered context that dramatically improves result relevance
  • Apply time, geography, source, and scope constraints to focus research
  • Request structured output formats appropriate for different research purposes
  • Use the "negative constraint" to exclude irrelevant content

Lesson Content

Layered context building.

Context has layers – each layer narrows the search to what is actually relevant:

Layer 1 – Role: "I am a mid-level HR professional" Layer 2 – Organization: "at a 50-person technology company" Layer 3 – Specific situation: "developing our first formal performance review process" Layer 4 – Constraints: "with a limited budget and no HR software currently in place" Layer 5 – Goal: "to implement something that takes managers no more than two hours to complete"

Combined:

"Research best practices for designing a simple, effective employee performance review process. Context: I am an HR professional at a 50-person technology company building our first formal review process. We have no HR software and need a process that takes managers no more than two hours to complete. Prioritize practical frameworks and templates used successfully by small to mid-size technology companies."

Time constraints.

Time constraints are especially important for Perplexity because it searches current web content:

  • "Published in the last 12 months" for rapidly evolving topics
  • "Guidance current as of 2026" for regulatory or compliance research
  • "Historical context since [year]" when you need trend perspective
  • "Most recent available data" when currency matters more than comprehensiveness

Geographic and regulatory constraints.

Many research topics are geography-specific – especially law, compensation, regulation, and labor markets. Always specify:

  • Country (especially for legal and regulatory research)
  • Region or state when state-level regulation matters
  • Industry sector when sector-specific regulations apply

Negative constraints – what to exclude.

"…and exclude academic theoretical frameworks – I need practical implementation guidance, not research literature."
"…excluding marketing and promotional content from vendors"
"…not including guidance designed for enterprise organizations – I need small-business appropriate approaches"

Negative constraints are efficient – they exclude entire categories of irrelevant content rather than filtering individual results.

Structured output that is immediately usable.

Match output format to how the research will be used:

For briefings: "Organize as: Key Finding (one sentence), Supporting Evidence (three bullets with citations), What This Means for Us (one paragraph)"

For comparison: "Comparison table: rows are criteria I care about, columns are each option. Brief assessment per cell. Citation number included."

For action planning: "Organize into: Immediate actions (this week), Short-term steps (next 30 days), Longer-term initiatives. Each with specific example and source."

Lesson Quiz

Log in and enroll to take this lesson quiz.

Scroll to Top