Section 10.3.5 : Perplexity and AI-powered research

🎯 Learning objective

Mastering Perplexity AI as a conversational search engine to obtain reliable, sourced, and up-to-date information in a professional context. Understanding why Perplexity has become essential for monitoring, fact-checking, and data research.

Perplexity: search reinvented

Perplexity AI is an AI-based conversational search engine. Unlike ChatGPT which primarily responds from its training data (with a risk of hallucination), Perplexity searches the web in real time and provides synthesized answers with cited, verifiable sources — numbered and clickable.

It's the answer to every professional's question: "How can I trust what the AI tells me?" With Perplexity, every claim links back to its source. You can verify.

Perplexity vs Google: a paradigm shift

Criterion	Google Search	Perplexity AI
Response format	List of 10 blue links	Structured synthesized response + sources
Time to get info	~5 min (browse 3-5 pages, filter ads)	~10 sec (direct answer, ready to use)
Cited sources	You must evaluate reliability yourself	Automatically cited, numbered, clickable
Conversational follow-up	No — new search = lost context	Yes — ask follow-up questions building on context
Academic research	Via Google Scholar (separate)	Integrated "Academic" Focus mode
Advertising	Omnipresent, sometimes misleading	No advertising in results
Personalization	Based on your history (filter bubble)	Objective, no filter bubble
Export	Manual copy-paste	Structured export, page sharing

Focus search modes — The complete guide

Perplexity's Focus mode lets you target your search to a specific source type:

Focus mode	Sources queried	Professional use case
All	Full web (default)	General research, daily monitoring
Academic	PubMed, arXiv, Semantic Scholar, CORE	Scientific research, literature review, medical data
Writing	Source-augmented writing assistance	Writing articles with references, sourced copywriting
Math	Calculation engine + mathematical reasoning	Problem-solving, formula verification, statistics
Video	Primarily YouTube	Finding tutorials, conferences, training on a topic
Social	Reddit, Quora, forums, discussions	Real user reviews, experience feedback, community trends

Advanced tip: combine Focus modes. Start with "Academic" for factual data, then "Social" for real-world feedback — this gives you a 360° view on a topic.

Detailed professional use cases

1. Competitive intelligence

What are the latest products launched by [Competitor X] 
in 2025-2026? Include:
- Pricing and positioning
- Media reactions (trade press)
- Customer reviews (G2, Trustpilot, Reddit)
- Their communication strategy

Complete workflow: run this query every Monday morning for each key competitor → export results → compile into a weekly monitoring dashboard.

2. Market research

What is the size of the online professional training market 
in Europe in 2024-2025? 
Specifically search for data from:
- Gartner, Statista, McKinsey, Deloitte
- European Commission reports
- Trade press articles (TechCrunch, Les Echos)
Give me exact figures with sources.

3. Systematic fact-checking

After receiving a response from ChatGPT or Claude, verify the key figures:

Verify this claim: "The global AI market 
will reach $850 billion by 2030 according to McKinsey."
Is this accurate? What is the primary source?

4. Regulatory monitoring

What regulatory changes affect the 
[fintech/healthcare/education] sector in France in 2025-2026?
Focus on:
- New laws and published decrees
- European directives being transposed
- Compliance deadlines

5. Meeting/pitch preparation

I'm meeting the CEO of [Company X] tomorrow. 
Give me a complete briefing:
- Recent financial results
- News from the last 3 months
- Challenges identified in the press
- CEO's background (LinkedIn, recent interviews)
- Relevant conversation topics

Collections and Spaces — Professional organization

Collections (renamed "Spaces" on Pro) let you organize your research by professional project:

→Create a Space per project: "Product Launch Q2 2026", "Legal Monitoring", "Competitor Benchmark"
→Custom instructions: each Space can have specific instructions ("Always respond in English, cite figures in euros, focus on the European market")
→Organized history: track how your research evolves over time
→Sharing: share a Space with your team for collaborative monitoring

Perplexity Pro ($20/month) — The best value in AI

Perplexity Pro offers: access to the best models (GPT-5, Claude Opus 4.6, Gemini 3.1 Pro) via a single interface, file upload for sourced analysis, 300+ Pro queries/day, and unlimited Spaces. It's the only subscription that gives access to all leading models — ideal for comparing answers. For professionals whose work requires daily research (consulting, journalism, analysis, monitoring), it's the most cost-effective AI subscription.

Advanced technique: the Perplexity + LLM workflow

Perplexity excels at finding information. ChatGPT and Claude excel at processing information. Combine both:

→Perplexity → Search for factual data with sources
→Copy the sourced results into ChatGPT or Claude
→Process: request an analysis, rephrasing, or integration into a document

This workflow eliminates hallucinations (Perplexity provides sourced facts) while benefiting from ChatGPT/Claude's analytical power.

Practical exercise with Perplexity

Duration: 25 minutes

→Create an account on perplexity.ai (free with limits, or Pro at $20/month)
→Factual research: "What is the size of the [your industry] market in France in 2025?" — verify the cited sources
→Test Academic mode: "What are recent studies on AI's impact on workplace productivity?"
→Test Social mode: "What are the user experience reports for [tool X] on Reddit?"
→Fact-check: take a claim from ChatGPT and verify it on Perplexity
→Create a Collection/Space "AI Monitoring" and add 3 thematic searches

Quick Check

What is Perplexity's main advantage over ChatGPT for research?