How to Use Perplexity AI for Content Research 📹

Read time: 5 mins

The Automated Research Engine ⚙️

I’ve been automating my content creation process using this method.

I call it my “Automated Research Engine”.

And it’s powered mainly by Perplexity, an AI tool I use and recommend.

This Automated Research Engine helps me find topics to write about for my newsletter every week.

Here’s a 5-step process for setting it up:

1. Create a Free Perplexity Account

Start by heading to Perplexity AI to create a free account.

You won’t need to pay a dime with this method.

2. Craft Relevant Prompts (or Copy Mine)

Next up, you’ll need to craft some prompts relevant to the information you want to aggregate from Perplexity.

This step is crucial as the prompt can make or break the output of the response Perplexity gives.

My tip?

Include templates in your prompt like:

Here’s what result it came up with.

Disclaimer: AI can make mistakes.

You MUST fact-check the information by clicking on the sources provided before using it for work.

Look how neat the formatting is here.

You can then copy this and paste it into your:

• Blog post

• Twitter Thread

• LinkedIn post

And feel free to write your templates to optimize for each platform.

3. Use Focus Modes to Find Unique Content

Focus modes are fantastic.

Creators are always sharing the same content over and over again.

Focus modes can help you narrow down research to a particular medium.

I use them to gather information from sources that people don’t normally check, like:

• Reddit

• Quora

• YouTube

• Academic papers

Let me show you an example.

To start, let’s click on “Home”.

Then, under the search bar, click on “Focus”.

We’re going to do some video research. Click on “Video”.

Now you’re in the Video focus mode.

To get the best result, we want to use templates in our prompt to tailor our response from Perplexity.

Here’s an example.

Let’s paste in this prompt:

Here’s the result:

Notice how they provide video sources from YouTube at the top?

The content you research now is specific to videos.

In this case, YouTube.

From here, feel free to continue the thread with follow-up questions for further research expansion.

4. Create a Collection of Threads (Prompts)

Who likes a messy table filled with unorganized papers, notes, and research?

That’s like your prompts right now. 👀

Creating a Collection helps keep all your research items in one place.

To create a collection, click on “Library” in the left sidebar.

Click on “Get Started” in the “Collections” section on the right side.

Give your Collection a title, description, AI prompt, and privacy settings.

Click on Create.

You now have your collection.

5. Add Threads to a Collection

Now you have your Collection.

It’s time to add some of the previous threads to it.

Your new Collection will then build upon these previous threads when outputting new responses.

You are essentially training it to give better responses.

To add a thread to your collection, click on the “…” three dots at the bottom of it and click on “Add to Collection”.

Then click on the Collection you just created.

Once added, you’ll see that they are now within the Collection in the Library.

Now you have your completed Collection!

The AI model now understands your intent well and will continue to understand more as you add more threads to it.

Your content research is now automated! ✅

To re-run all the prompts, just open up this Collection.

Start a new thread, or access old threads you’ve created!

If you want to share a collection, you can do so. See my collection.

Here’s an example:

If you want to re-run a search for all the newest AI research journals in the past week, head to the collection and re-run your thread.

As Perplexity uses live web search, it will be updated to show results from the past week.

To understand Perplexity more, here are some helpful YouTube videos I found:

That’s all for this week.

See you next Wednesday!

Justin Chia

Find me on LinkedIn or shoot me an email at [email protected]

If you have any questions or feedback, reply to this email. (I'll personally reply! :)