According to Stanford University's AI Index Report 2024, there are now hundreds of thousands of AI tools for a wide variety of tasks. It's understandable if you can't see the forest for the trees and don't know which tools are worth using and which are better left out. To help you separate the wheat from the chaff, we will discuss the various tools that we recommend in our AI workshop in a series of articles.

In this first edition, I take a look at Google's NotebookLM. Chances are you've already heard about it. But what exactly can you do with it? Is it a valuable addition to your current AI toolkit or is it mostly new hype? What makes NotebookLM different from the tools you already use, and is it worth investing time in?

As a content manager at Buzzlytics, I have used and tested NotebookLM extensively since its launch. I have experience with NotebookLM as well as with other AI tools such as ChatGPT and Claude, allowing me to compare well where this new tool from Google really excels. I have saved a lot of time in several tasks by using NotebookLM alongside other AI tools.

In this article, I'm going to show you exactly what NotebookLM is and how it differs from tools like ChatGPT. You'll learn what the main functionalities are, how to get started with it, and for which business applications - especially in marketing and sales - the tool is best suited. I also openly discuss its limitations, so you get a realistic picture.

After reading this article, you will know exactly when NotebookLM is the right choice for your specific needs and how to use it effectively in your daily work.

NotebookLM focuses primarily on analyzing, organizing and making sense of existing information.

NotebookLM is a free AI tool from Google designed specifically to help you understand and process large amounts of information from a variety of sources. Unlike most other AI tools, which often focus on generating content, NotebookLM focuses primarily on analyzing, organizing and making sense of existing information.

Notebook, as mentioned above, is free for individual users, but there is also a paid version(NotebookLM Plus) in which, among other things, you can create more notebooks and add more resources per notebook.

NotebookLM's strength lies in its ability to interpret and analyze up to 50 different sources simultaneously. It supports a wide range of formats:

• PDF files
• Website URLs
• YouTube videos (including transcription).
• Copied text
• Google Docs and Slides
• Audio files (automatically transcribed).

A special feature of NotebookLM is the so-called source grounding functionality. When answering questions, the tool explicitly cites the sources from which the information comes, including small numbers that point to the exact location in your documents. This makes it easy to verify the information and reduces the chance of so-called hallucinations.

If you ask NotebookLM a question for which the answer cannot be found in any of the resources, it will indicate it, instead of making up an answer:

NotebookLM's context window is impressively large: up to 25 million words per notebook in the free version. This capacity allows you to analyze complex documents and large amounts of information simultaneously.

When answering questions, NotebookLM explicitly cites the sources from which the information comes. This makes it easy to verify the information and reduces the risk of hallucinations.

NotebookLM offers several features that allow you to process information from your sources in new and useful ways:

One of NotebookLM's most innovative features is that you can create personalized podcasts based on your uploaded resources. You can direct the content of these podcasts by adding context and details before you generate a podcast. For example, you can specify that the focus should be on a specific topic or that the podcast should be aimed at a specific audience. This functionality is currently only available in English.

With one click of the "Study Material" button, you can have documents generated that contain questions and answers based on your resources. This is useful when studying or to prepare your training sessions.

For those who need quick insight into a project or topic, NotebookLM can create a summary overview document. This is ideal for onboarding new team members or informing stakeholders.

NotebookLM can create an organized timeline of events described in your sources. This is valuable for project planning, historical research or mapping developments.

For example, I myself used NotebookLM to get a quick understanding of a book's content. I uploaded a pdf of the book into Notebook, added a YouTube video on the same topic, and then asked a few questions about the sources. This got me up-to-speed much faster than if I had had to read the book from cover to cover first.

My colleague Lucas used NotebookLM to prepare a training course. He added a 30-minute YouTube video to NotebookLM, got a nice summary in the summary document, asked 3 in-depth questions, and was able to use the content of the video in a training that same afternoon.

This might also be the place to comment on how safe your files are with Google. Google itself states that your personal data, including your source uploads, queries and model responses, are not used for training. However, the paid version, NotebookLM Plus, does seem to offer a higher degree of confidentiality, and it is recommended that you use the paid version for sensitive information.

If you want to know more about this, watch the video below:

The AI market is crowded with different tools, but NotebookLM differs from popular alternatives such as ChatGPT in important ways.

• NotebookLM is primarily a tool for "understanding. It helps you gain insight from your own sources by organizing, summarizing and asking targeted questions about them.
• ChatGPT is first and foremost a generative AI model, trained on Internet data, that is strong at creating new content, answering general questions and conducting conversations.

• NotebookLM is specifically designed to interpret multiple sources simultaneously. The chat interface takes into account all selected sources.
• ChatGPT can also work with uploaded documents, but is less focused on interpreting many different sources at once.

• A crucial feature of NotebookLM is its emphasis on source grounding. It provides inline citations that lead directly to the specific location in the source, which helps verify the reliability of answers.
• ChatGPT is known to hallucinate and generate factually incorrect information, especially without a direct source. ChatGPT's output is therefore more difficult to verify.

• NotebookLM has a significantly larger context window than ChatGPT. This makes it possible to process huge amounts of information simultaneously.
• ChatGPT has a smaller context window, which limits the amount of information it can effectively process.

• NotebookLM offers specific features aimed at processing resources, such as AI podcasts, study materials, timelines and summary documents.
• ChatGPT is more versatile and can be used for a wide range of tasks, including content creation and programming.

NotebookLM is the better choice when:

• You work with large amounts of your own documents and information
• you need high accuracy and source identification
• You need to combine and analyze information from different formats (text, video, audio).
• restructure your information and present it in new formats (podcasts, timelines).
• minimize your hallucinations from AI

ChatGPT remains the better choice for:

• Generating creative content
• Answer general questions
• Coding and Programming
• Brainstorming and idea generation

While NotebookLM is powerful, it also has limitations.

Google has been known to discontinue products (known as Google Graveyard). If you use NotebookLM heavily, it is wise to store important information externally as a backup.

NotebookLM is geared toward accuracy and avoiding hallucinations, but at the expense of creativity. For out-of-the-box ideas, other tools may be better suited.

The quality of the output is highly dependent on the quality of the input sources. Reliable articles produce better results than low-quality "clickbaity" content. So make sure you always use quality and reliable sources.

It is important to note that conversations and generated output in the chat interface are not saved automatically. So you have to manually save important output as notes. This is rather cumbersome.

The user interface, especially for managing notes, is not optimal and could be improved. The platform undergoes regular changes, indicating ongoing development.

You can add Google Drive documents, PDF files, website URLs, YouTube videos, copied text, Google Slides or audio files to a notebook.

Getting started with NotebookLM is easy. Just follow these steps:

1. Go to notebooklm.google.com. In your browser, navigate to the NotebookLM website and click the "Try NotebookLM" button.
2. Log in with your Google account. You need a Google account to use NotebookLM.
3. Create a new notebook. Click the '+ Create New' button after logging in.
4. Give your notebook a title. Choose a relevant name, such as one based on the project or topic.
5. Upload your resources. Click '+ Add Source' to add sources. These can be Google Drive documents, PDF files, website URLs, YouTube videos, copied text, Google Slides or audio files.
6. Start by asking questions. Once your resources are uploaded and processed, you can ask questions in the text field at the bottom of the interface. NotebookLM will provide answers based on your sources, referencing the specific resources.

Good to know: not all YouTube videos are accessible in NotebookLM. Only public YouTube videos with subtitles (self uploaded or automatically generated) are supported. Also recent videos (uploaded 72 hours or less ago) may not be imported, probably because the automatically generated subtitles are not yet available then. In these cases, you can try again later.

In addition, NotebookLM cannot import content from websites that have a pay wall or that have explicitly disabled web scraping. Of course, you can get around this by copypasting the text of the article.

NotebookLM offers marketers countless ways to get their work done faster and more efficiently. Below I provide some examples:

• Upload market reports, competitive analysis and transcripts of relevant videos
• Ask targeted questions about trends, competitor strategies and market developments.
• Analyze multiple sources simultaneously to discover deeper connections
• Generate audio overviews (podcasts) of your research material to keep up with when you are in the car, for example

• Analyze existing content to generate new ideas and identify gaps
• Structure content and summarize complex topics
• Easily add website content from competitors or sources of inspiration as a resource
• Create personalized podcasts based on your research.

I myself also used NotebookLM when writing this article.

Here's how I approached it:

1. I searched for reliable sources and added them to a notebook in NotebookLM. So I created a notebook specifically for this article.
2. I asked a number of questions within this notebook based on the content compass I filled out in preparation for this article.
3. I saved the answers to the questions as notes and copied them into a Google Doc. A Word file also works, but Claude (the AI tool I use to write my content) has a link to Google Drive, so I can easily attach Google Docs without having to download them first and then upload them again.
4. I asked Claude to use the contents of this Google Doc as input for the article.
5. I have added my own anecdotes and experiences.

The advantage of working this way is that I am less likely to reach my user limit in Claude this way, because the input only contains things that are 100% relevant to the article. Thus, no tokens are wasted.

• Upload campaign reports and presentations and ask targeted questions about performance and ROI
• Analyze Google Slides including images, charts and tables
• Discover long-term trends by analyzing multiple reports simultaneously

• Use internal guidelines and training materials to generate study materials and summary documents
• Create audio summaries of important documents for new team members

• Analyze transcripts of customer conversations and feedback to identify recurring themes
• Research social media data to discover sentiments and pain points

Sales teams can use NotebookLM to work more efficiently, make more informed decisions and improve their sales performance.

NotebookLM also offers valuable applications for sales professionals:

• Upload recordings and transcripts of successful and less successful sales calls
• Analyze language patterns, message consistency and overall effectiveness
• Identify trends and success factors that lead to better results.

• Bundle all relevant product documentation, price lists and competitive information
• Create a central learning centre that the whole team can access.
• Ask specific questions about products and get answers with sources.

• Load background information on potential customers (CRM exports, LinkedIn profiles)
• Prepare personalized and relevant sales calls.
• Analyze previous offers and contracts for insight into successful deals

• Generate briefing documents and FAQs based on existing sales material.
• Create audio summaries of important training materials

• Use NotebookLM as a "project context engine" for long-term sales cycles
• Centralize all relevant meeting notes, project plans and client information
• Maintain an overview and find answers quickly throughout the sakes cycle.

By leveraging these features, sales teams can work more efficiently, make more informed decisions and improve their sales performance.

If you regularly need to work through large amounts of text - whether for market research, product documentation, customer feedback or training materials - you can save a lot of time with NotebookLM. Imagine not having to manually search through countless PDFs, presentations and web pages, but instead being able to ask targeted questions and get immediate answers with clear source citations.

The real power of NotebookLM lies not in replacing other AI tools, but in strategically deploying the right tool for the right task. For writing, I personally continue to use Claude, but for analyzing my own library of sources and generating reliable insights from them, NotebookLM has become my go-to.

Want to experience for yourself how NotebookLM can improve your information processing processes? Go to notebooklm.google.com, create an account and start uploading some of your most frequently used documents. Experiment with different source types and features such as the AI podcasts or study materials. Discover how this tool can help you get more value out of information you already own, but can now use more effectively.

Want to learn how to use AI to improve your marketing and sales processes and results? Then our 1-day AI workshop might be for you.