How I Use Perplexity AI for Initial Research in Product Management

Struggling with Initial Research? You’re Not Alone.

If you’ve ever felt stuck doing initial research or creating documentation—especially for projects that are new or unfamiliar—you’re not alone. I experience this too.

This challenge becomes even more obvious when working in a startup agency or software house, where projects constantly change across industries, scopes, and expected outputs. While the variety can be exciting, it can also create stress and cognitive overload.

That’s where AI tools for research start to make a real difference.

Discovering Perplexity AI as a Research Tool

Some time ago, I came across Peter Yang explaining Perplexity AI, an AI-powered research tool designed to modernize traditional search and optimize the research workflow itself.

But what does that actually mean in practice?

At its core, Perplexity AI follows a workflow that closely mirrors how humans typically conduct research—just faster and more efficiently.

How Perplexity AI Works (Simplified Workflow)

When you submit a prompt to Perplexity AI, it generally follows these steps:

Searches multiple sources simultaneously
Extracts only the information relevant to your prompt
Synthesizes key insights into a structured explanation
Clearly shows the sources used to generate the output

If this sounds familiar, it’s because this is exactly how we already do research manually—just with far more effort.

The key difference is automation and efficiency.

A Simple Analogy

Google gives you the raw ingredients
Perplexity AI acts as the chef

And importantly, the chef isn’t hiding anything. You can still see the sources and understand how the conclusions were formed.

Why Perplexity AI Is Useful for Initial Research

From my experience, Perplexity AI significantly reduces the mental friction of early-stage research. It helps with:

Avoiding information overload
Reducing the need to jump between multiple browser tabs
Lowering the anxiety of not knowing where to start

(Some of that anxiety might just be me, but I doubt I’m alone.)

Key Benefits I’ve Experienced

Faster understanding of new topics
No need to manually check sources one by one
Research can run in the background while you focus on other work
- Depending on prompt complexity, results usually take 3–5 minutes
Important information is separated from noise and redundant content

When Perplexity AI Works Best

Based on my experience, Perplexity AI is especially effective in the following scenarios:

1. Learning a New Topic Quickly

When you don’t yet know the right keywords and need a high-level understanding before going deeper.

2. Initial Product or Market Research

I’ve used Perplexity AI for many projects during the discovery phase, particularly in product management contexts.

3. Helping Non-Technical Readers Understand Documentation

When someone is asked to “read the documentation first,” AI tools like Perplexity can help non-technical users understand technical documentation more easily.

That said, for document-heavy use cases, tools like Google NotebookLM may be even more suitable—but that’s a topic for another post.

Example: Using Perplexity AI for Market Research

Here’s one research question I recently explored using Perplexity AI:

Is q-commerce still relevant given current market conditions?

One of Generated Chart by Perplexity AI

You can view the full research output here:
👉 https://www.perplexity.ai/search/research-objective-conduct-com-CUdltk_kTtm6HTawxjjMvA#0

If you’re curious, don’t hesitate to try it yourself. The most important step is always the same: #YangPentingMulaiDulu or just start.

Key Takeaways: AI for Research and Product Work

AI tools like Perplexity AI can effectively help us “clone” ourselves—one instance doing research, another coding, while we focus on alignment and decision-making
AI improves speed and efficiency, but it doesn’t replace responsibility
Always verify and proofread AI-generated outputs

Every AI tool carries the same reminder for a reason:

“AI can make mistakes. Please recheck the information.”