Skip to main content
1.7

The Research Sprint

Estimated time: 30 minTool: Perplexity + Claude Pro
After this drill, you can:

After this drill, you can conduct a 30-minute research sprint using Perplexity and Claude together — source-grounded discovery plus deep synthesis.

Why this matters

Single-tool research is a crutch. The power move is tool-chaining: use Perplexity to find what exists and verify sources, then feed the best material to Claude for deep synthesis, pattern recognition, and implications. This two-step workflow produces research output that would take 3–4 hours of traditional methods in under 30 minutes.

How to do it

  1. 1

    Choose a research topic you genuinely want to understand better

    It should be substantive enough to require synthesis — not just a factual lookup. Good examples: understanding a policy area, investigating a market trend, exploring a technical domain.

  2. 2

    Run Phase 1: Discovery in Perplexity (15 minutes)

    Use the Perplexity prompts below to gather sourced overviews of your topic. Note the sources Perplexity cites — check 2–3 of them.

  3. 3

    Run Phase 2: Synthesis in Claude (15 minutes)

    Paste the Perplexity output (or the key sections) into Claude. Use the synthesis prompt below. Claude doesn't have Perplexity's real-time sources but excels at pattern recognition and implications.

  4. 4

    Document the two-stage workflow

    Note what each tool contributed. This two-tool pattern is reusable for any research task.

The prompt

PROMPT — Phase 1: Discovery (Perplexity)Model: Perplexity
Research [YOUR TOPIC] comprehensively.

Give me:
1. A 3-paragraph overview of the current state
2. The 5 most credible sources I should read
3. The 3 most important debates or disagreements in this area
4. What has changed in the last 12 months

Cite all sources with URLs I can click.
PROMPT — Phase 2: Synthesis (Claude)Model: Claude Pro
I've done initial research on [YOUR TOPIC]. Here's what I found:

[PASTE PERPLEXITY RESPONSE OR KEY SECTIONS]

Now help me go deeper:

1. What patterns do you see across these findings?
2. What are the implications for [YOUR SPECIFIC CONTEXT — why you care about this topic]?
3. What important questions are NOT answered by this research?
4. If I wanted to understand this topic 3 months from now, what would be most likely to have changed?

Success criteria

  • ✓You completed both phases (Perplexity discovery + Claude synthesis)
  • ✓You verified at least 2 Perplexity sources manually
  • ✓You have a two-stage research output you could actually use
  • ✓You can describe what each tool contributed that the other could not

Common mistakes

Using Claude for Phase 1 instead of Perplexity

→ Claude doesn't have real-time internet access by default. Perplexity is specifically designed for sourced research. Use the right tool for the right phase.

Pasting all of Perplexity's output into Claude without filtering

→ Claude's synthesis is better when you give it the key sections, not everything. Spend 2 minutes identifying the most important parts of the Perplexity output before feeding them in.

Not giving Claude your specific context in Phase 2

→ "What are the implications?" is too generic. "What are the implications for a nonprofit education organization in Iceland trying to serve rural communities?" produces actionable insights.