NotebookLM remains one of my favorite AI research tools thanks to its reliance on your own sources. But even though I've used the tool for plenty of things, I'm always intrigued by the ingenious ways people use it.

When I came across the Reports feature, I assumed it was just a way to generate summaries of your sources. However, when I finally explored what reports could do, I realized I had greatly underestimated how versatile they could be.

What are NotebookLM reports?

A useful feature that's more versatile than it seems

Reports in NotebookLM are one of the tool's most underrated features. They allow you to generate a variety of different content types and summaries based on your sources.

In this way, the title of "reports" is a bit misleading. When I first heard about the feature, I didn't realize how versatile it was. Standard formats include a briefing document, a study guide (with a quiz, suggested essay questions, and a glossary of key terms), and a blog post. So, in addition to audio and video overviews, reports are actually one of the best ways to turn complex notes and sources into digestible posts and documents.

I particularly like the feature because it isn't just aimed at students. For example, while study guides are useful for students and educators, the glossary generated in these reports is great for anyone trying to learn more about a topic.

Meanwhile, a briefing document is incredibly helpful for anyone who needs to catch others up on a topic. However, it's also useful for professionals who need to brief clients.

To access the feature in a notebook, simply use the right-hand pane and select the Reports option. This will bring up a pop-up with static report options, as well as suggested report options that NotebookLM generates based on your sources.

NotebookLM reports adapt to your sources

Each notebook yields different suggestions

NotebookLM doesn't only stick to the default formats it creates for reports — it also generates suggestions that depend on your sources. This means that you will get different types of report formats for different notebooks.

For example, for my notebook on fibromyalgia, NotebookLM had additional formats for a clinical practice paper, a therapeutic strategy report, an education primer, and an introductory wellness handbook. When I chose the wellness handbook format, it created a guide on managing chronic fatigue in fibromyalgia.

Meanwhile, for my Dungeons & Dragons notebook, I was able to generate a beginner's primer that introduces new players to the core concepts in D&D gameplay. For my garden care notebook, it suggested reports such as landscape design specifications, a grounds management strategy, and a botanical characteristic summary.

Custom reports feel like a NotebookLM superpower

The only limit is your imagination

If NotebookLM doesn't suggest a format you would like, you can create one from scratch using the Create your own option. This allows you to use your own prompt to generate a report.

For example, in my garden notebook, I created a custom garden care guide. This guide included a summary of the plants in my garden, with sections dedicated to their watering guidelines and potential issues I might encounter.

I've read so many websites about the plants in my garden, but often struggle to retain everything. By having my own custom report, I have a guide where I can quickly find the most important information. The report surfaced information that I had forgotten about from the original sources.

You can also use the custom reports feature to get NotebookLM to follow a very specific format. For example, I used it to generate a first-person account of what it feels like to live with fibromyalgia, using a template I had uploaded as a source.

This template was very simple, including formatted lines for "Heading for first topic" and specifications of how long each section should be. This template can be applied to any notebook thanks to its broad scope. Using the prompt, I specified which topics should be covered, as well as the tone the chatbot should use.

It was able to generate a report that matched my document format, down to the number of sentences and paragraphs for each section. Occasionally, it did expand the scope when I wasn't specific enough with my instructions, but this can be easily refined with more targeted prompts.

I was surprised by NotebookLM's adaptability in this respect. Since you can get the chatbot to use your own template, you can apply its generative abilities to an endless variety of tasks.

I underestimated what NotebookLM reports could do

While I have been concerned that NotebookLM may lose what made it special, the reports feature is a great addition to the tool. It's unfortunate that its name doesn't highlight just how much the feature can do.

If you have been ignoring it because you also assumed they were just document summaries, I suggest taking a look at what kinds of reports it can generate for your sources. You may find something incredibly handy in its suggestions.