Transcription
Both NotebookLM and Perplexity just got some amazing updates, and I found that they actually make a powerful combo when used together for research or learning.
You have Perplexity's real-time search capability combined with NotebookLM's deep analysis. In this video, I'll show you six incredible ways you can use Perplexity and NotebookLM to level up your research or learning process to a whole new level. Let's go.
So, why do we need to use both and not just one of them? Perplexity can do web searches and find sources, but the limitation is that every time you ask a question, it will trigger a new search and use different sets of sources. Even if you upload your own knowledge base to the Perplexity Space, it may still use some pre-trained data to answer your question.
This is why NotebookLM is useful. It will only give you answers based on your sources, and the source reference will not change unless you import a new source. It's better to organize your notes and insights with its closed database.
First, use Perplexity to gather specific sources based on your prompt or create a Perplexity space for collecting relevant sources like reports, studies, videos, and articles. Then, select the most valuable sources and import them into NotebookLM. Quality over quantity here.
In NotebookLM, ask questions to extract insights or patterns. Use its audio features to discuss your research and organize different research projects. Finally, which is optional, you can go back to Perplexity to validate your new insights against open web data and do cross-checking.
There are six amazing ways you can use this workflow. The first one is doing market trend analysis. Oftentimes, as marketers, you will need to understand market trends so you can better plan content and thought leadership pieces. This is what Perplexity space is so useful for because you can use it to organize your mini-project and set up custom instructions to semi-automate the process.
Let's say I'm doing market research about Responsible AI. Then on Perplexity, go to space, create a new space, and let's name it "Responsible AI Market Research" and pick the AI model. For the custom instructions, I would set up two task codes. The first one is used to search for reports about responsible AI from reputable consulting firms like Deloitte, McKinsey, and BCG. The second task code will focus on finding research papers about AI risk assessment.
The magic happens whenever you type the task code or the number. Perplexity will automatically perform the prompts for you. Double-check each source and get the source link that you would like to import to NotebookLM. To get the links easier, you can use the copy functions on Perplexity and paste them on a notepad. This way, you can have a list of source links ready.
Then, repeat the task for research papers from tech companies. Now, on NotebookLM, create a new notebook and import all the sources or any of your own sources if you have them, and then start asking questions. For example, "What are the major shifts happening in responsible AI?" It will pull the trends that it is seeing, like increased adoption and awareness, Human-in-the-Loop Oversight, and it will only retrieve from the source, unlike Perplexity.
If it can't find anything related from the sources, it will say no. This is something that I like because it can always minimize the level of hallucination. You can ask other questions like, "What is driving companies to invest in this topic?" so you will know what they care about. Then, ask it to propose some content ideas for content creation or any specific topics you want to know more about.
The next one is product enhancement research. If you're a product startup founder, this use case will be useful. One of the powerful ways of using Perplexity is to source real user feedback and do competitive analysis. You can find real user pain points and identify potential product improvements or new product ideas to develop.
It doesn't have to be a direct competitor. You can research some top well-known products in your niche to get ideas. Let's say you are planning to enhance your email marketing automation tool. First, we can go to Perplexity to identify the top software providers in email marketing automation. It will show a list of the top 10 email automation tools like ActiveCampaign, Klaviyo, and MailChimp.
Now, let's say we pick ActiveCampaign as the research target. Then, ask it to find all the review links about ActiveCampaign from G2, Product Hunt, TrustRadius, or any other B2B software review sites that you know. You can see it will pull a summary table for you, which is great. You have a list of review links to import to NotebookLM, but we are not done yet.
We can actually use the focus search to narrow down just Reddit or YouTube that have more in-depth real user reviews to get more source links. Again, click show all and review every source to select just the relevant ones to import to NotebookLM. I also recommend searching for related industry reports in this niche so you understand what the trends look like to shape your product development.
Now, on NotebookLM, import all the sources. For demo purposes, I just imported one competitor, but I recommend you import at least two to three competitors' review data so the analysis will be more accurate, and you are utilizing the power of NotebookLM.
First, ask, "What are the common pain points mentioned across these sources?" You will see what some weaknesses that competitors fail to deliver might be your opportunities to fill the gap in your product, like technical issues, email design limitations, and customer support. Then, ask for specific features that users consistently mention as missing.
In this case, that is from ActiveCampaign. This is super useful because you will know the exact features that users are looking for, so you can add them to your product or get some ideas. Then we can ask, "What are some technologies that we should be aware of when developing this product?" like AI for personalization, lead scoring, and customer data management. All these are great ideas to shape your product development.
If you want, you can even upload your website product page and ask it to give you an improvement plan on the features or messaging. The next is audience research. Audience research is a very important part of digital marketing. You can understand your audience better for your campaign and craft your messaging.
By combining Perplexity and NotebookLM, you can do a much more in-depth audience research analysis. For example, let's say I have a childcare service business, and I want to understand more about my target audience, who are modern parents. First, on Perplexity, we can ask for recent survey studies about parents' needs for childcare. This will focus more on some survey analysis from our ideal customers.
Again, use search operators to narrow down the timeframe. Browse through the sources and save the links. Then, also ask for research reports about parent childcare needs. This will focus more on the research findings in the industry. Also, search for recent studies about how parents choose childcare services. It’s not necessarily in PDF report format; it can also be a webpage.
If you know any reputable media sources, also use the search operators to narrow the scope. We're not done yet. Now, turn on focus search mode to videos and search for expert interviews about the childcare industry. This will give you multiple perspectives from industry practitioners.
Then, on NotebookLM, import all the sources. First, we can ask about the biggest challenges target customers mention about childcare. It gives you more ideas about the pain points, like availability, quality childcare, and learning. Then we can be more direct and ask, "What are the considerations when they choose a childcare service?"
You will see it will give you specific details like cost, location, convenience, and the branding of the childcare. A powerful way is to import your own business website and ask for an improvement action plan. Let's say this is my childcare business website. Import it to NotebookLM and ask for the top three changes to the homepage messaging based on the research.
It will give you some good suggestions like promoting flexible and extended hours. We can also ask about the needs that are not mentioned on the website so we can highlight them a bit, like specific learning philosophy and staff qualifications. We can even take a step further to ask for specific services to highlight more.
You will get some really good ideas like different schedule options, Full Day/Full Week Schedule, and I like how it bolds some of the responses to emphasize the importance. The next is podcast research. For sure, podcasts will continue to be one of the trending content formats in the coming years. If you're planning to start a podcast, you can actually use Perplexity and NotebookLM to do more in-depth podcast research and understand what your audience is looking for, get some content ideas, and how you should position your podcast before you start.
Let's say we're planning a new podcast in leadership. First, we're going to ask Perplexity for the top 10 leadership podcasts launched in the past year with the highest listener ratings on Apple Podcasts and Spotify. Most importantly, we want the podcast show links. Note that I have used the search operators to focus on searching from ListenNotes.com, which is a very popular podcast search engine.
In case there is no podcast show page, you can directly search the podcast name on ListenNotes and get the link. For each podcast show, download one of the most recent podcasts for analysis purposes later. Like this podcast, "No BS Leadership," you can find the recent episodes and then click download audio. I suggest downloading at least five sample episodes for analysis purposes.
Now, also ask Perplexity to search for the user review link for these top podcast shows. Note this time, I'll ask it to focus on Pod Paradise's website as it has podcast reviews. Sometimes Perplexity may not be able to find you the review page link, but if the podcast is popular enough, you can just search on Pod Paradise and access the reviews on the show page.
Now, on NotebookLM, import all sources, reviews, and sample episodes. Then ask how these podcast shows describe their ideal listeners, like specific roles, experience levels, or company types mentioned. You can see it would give lots of details about the target audience for inspiration.
Then you can further ask it to summarize the target audience into five categories. Now it’s much clearer, and it gives you the top five target audiences you should be targeting, like aspiring leaders and executive coaches. You can further ask to compare the podcast show pages for unique angles and positioning that they emphasize.
This is helpful if you want to get more ideas when you brainstorm your own positioning for podcasts. Now you can ask to summarize the top five positive aspects and five negative aspects from the podcasts' reviews. This way, you can double down on the positive things that you're seeing and avoid the negatives, like lack of episode numbering.
You can also ask for top content ideas that can resonate with these audiences if you were to make a new podcast. These are all good ideas like “Building a Culture of Feedback” and “Time Management for Leaders.” You can also dig deeper into the podcast style and summarize the best practices.
I would say this is really useful if you're planning your new podcast. You can get inspiration from these successful podcasts about what best practices you should follow, like giving actionable advice, focusing on a specific theme, and sharing self-assessment tools. They're all useful guidance.
Next, learning a new subject. I can't stress enough how AI research tools are helping me to speed up my learning process. With NotebookLM, you can even learn anything faster. Now, with its interactive audio mode, you can join the conversations to elevate the whole learning experience and make it more engaging.
Let's say I'm doing self-learning about the topic of Marketing Psychology. First, on Perplexity, search for detailed guides or learning materials about marketing psychology. We can also use the search operator "INTITLE" to narrow the search results to only marketing psychology in the title.
Again, browse through sources and pick the ones that are the most relevant and trustworthy. Then search for the top five podcast episodes on the topic. Episodes like "The Psychology Behind Effective Pricing" and other popular episodes are great sources for more in-depth knowledge.
Then search for these episodes on ListenNotes or Pod Paradise and download them. We can also search for academic studies about the topic, so save all the good source links. I also recommend finding expert interviews and lectures about the topic using the focus search mode on YouTube. This way, you can ensure your sources are more well-rounded before you import them to NotebookLM.
Now, on NotebookLM, import all the sources. The easiest way is to use the pre-built prompt to generate a study guide and FAQ to quickly get an idea about the topic. What are some commonly asked questions? Now we can ask it to give a beginner-friendly overview of the subject using simple language and real examples.
It will give you a brief overview of what it is and why it matters. Then we can ask it to summarize the top marketing psychology principles and how to apply them in marketing. This is useful because within a few seconds, you can get a quick understanding of the core marketing psychology principles like scarcity and anchoring bias.
Then we can ask it to show examples of how companies use them. Now, import the childcare business website we just demoed and ask how these psychology principles can be applied. You can see the suggestions are great, like highlighting limited enrollment and asking parents to create profiles to personalize their experience.
The powerful thing here is you can always learn something new and directly apply it to your own projects, business, or campaigns. Ask NotebookLM to link all the sources together and find the linkages for you.
To elevate the learning experience more, you can generate the audio overview so we can customize it to focus on the marketing psychology principles targeted to someone who is a complete beginner. You can then download and upload it to Google Drive, so you can listen and learn on the go.
Another way is to use the interactive mode. Once you click, you can stop playing the conversation and then interrupt the conversation and ask questions by clicking Join. For example, let's ask, "What makes a good anchor price when using the anchoring pricing bias principle?" based on the audio overview.
If the generated audio overview didn't mention anything related to the question, it will not be able to answer, even if it appears in the imported sources.
Next, speaking and presenting. As marketers or working professionals, I believe we would come across scenarios where we need to present to stakeholders or deliver pitches for different audiences. Speaking and presenting is an important skill for most professionals, and we can actually use Perplexity to find some successful speeches and presentations for us, and NotebookLM to be our speaking guide.
First, on Perplexity, search for the most viewed TED talks of all time using the focus search mode with YouTube. If you have specific speaking topics or scenarios, like delivering a product launch keynote, startup pitch, or graduation speeches, you can also specify in the prompt to tailor your search results.
Now you will get a list of TED talks, and they're all amazing talks. Like the first one, "Do Schools Kill Creativity?" One thing is that some of them are from a long time ago, so you might not be able to import them to NotebookLM, like in this case. You can use a free transcript generator to grab the transcript and import it to NotebookLM as pasted text.
Let's say we have imported all the talks, and then I can ask for common elements in the presentations' openings because openings are usually the most important. It will give you ideas like asking questions, direct address, and audience engagement.
Then we can also ask it to analyze deeper on the storytelling structure for each of the talks, like how they usually begin with a personal anecdote or question, and then research or case study with details. This will give you ideas on how to deliver a successful presentation.
If you want, you can always dig deeper into a specific speech. We can also ask it to summarize the top five speaking techniques across these talks. I would say this is like best practices that you should follow.
Now, copy this response to somewhere else or save it as a note, as we're going to use it again. Let's say I have created this Perplexity page about the "Future of Work Trends," and I need to deliver a speech. Now, publish it and get the link.
On NotebookLM, create a new notebook and re-import the link. The reason for this is I found that even if I deselect the source, NotebookLM will still use it to generate the audio generation. Now, import the summarized speaking techniques you just copied to its knowledge base and ask it to summarize the top five speaking techniques again, but within 200 characters.
We will use some of the keywords to customize the audio overview. Now, click customize and ask it to apply the speaking techniques it analyzed and also host one to focus on presenting and host two to focus on highlighting the speaking techniques, so they demonstrate the techniques to you.
I won't play it all here, but you can hear how it is following my instructions using the speaking techniques, and host two highlights the techniques used. Of course, this is not perfect, but by customizing, you can tailor the audio overview generation for your specific learning needs.
There are more ways you can customize the audio overview generation, and I share this in my community. You can find the link in the description to join.
Alright, make sure you try these use cases using both tools. For more inspiration, join my community. You can also ask me questions there. Before you go, I also recommend you check out my other video about smart prompting techniques you can use on Perplexity or SearchGPT. I'll see you next time.