Skip to main content

📔 Lesson 1.3: Your First Notebook — Add a Source and Ask a Grounded Question

This is the lesson where NotebookLM stops being an idea and starts being yours. You'll create a real notebook, add your first source, watch NotebookLM read and process it, then ask a question and see a little numbered citation appear in the answer. Click it, and the app jumps you straight to the exact sentence in your source. That click — "the AI said it, and here's proof" — is the whole magic trick, and once you've felt it, everything else in the course is just variations on it.

📚 What You'll Learn

By the end of this lesson, you will be able to:

  • Create a new notebook and give it a clear name
  • Add your first source using at least one method — upload a PDF, paste text, or add a Google Doc
  • Recognize what "processing" is and why you wait a moment before chatting
  • Ask a grounded question and read the inline citation, then click it to jump to the source passage
  • Use NotebookLM's auto-generated summary and suggested questions as a running start
  • Describe the core loop — add source → ask → verify citation — and why answers stay inside your source

⏱️ Estimated Time: 40 minutes

🎯 Project: Build the real notebook from your Lesson 1.1 Notebook Brief with at least one source, ask it three questions, and click at least one citation to verify the answer.

In This Lesson

Create Your First Notebook

Remember the notebook home from last lesson — your empty bookshelf? Time to put a book on it. From the notebook home, click "Create new" (it may read "+ New notebook" or simply a big plus). NotebookLM opens a fresh, empty notebook and typically invites you to add your first source right away — often the very first thing it asks for.

Before anything else, give the notebook a name. A new notebook usually starts as "Untitled notebook"; click the title to rename it to something that matches your Brief — say "Thesis: Coral Reef Recovery" or "Home Solar Research." This tiny habit pays off fast: within a week you'll have several notebooks, and a shelf of clearly-named ones beats a shelf of "Untitled," "Untitled (2)," "Untitled (3)."

📖 Definition

Notebook (recap): a workspace built around a set of sources on one topic, holding those sources, your chat with them, and the notes and Studio outputs you create. One notebook per project or question is the sweet spot. Mixing unrelated topics into one notebook muddies its answers — focused sources give focused replies.

⚠️ Important Note: Notice that a brand-new notebook can't chat yet — the chat box is waiting. That's by design: with no sources, there's nothing for NotebookLM to be grounded in. The tool literally cannot answer until you give it something to read. That constraint is the source-grounding you learned about in Lesson 1.1, made concrete.

Add Your First Source

A fresh notebook shows an "+ Add" (or "Add source") control in the Sources panel — and usually a menu of the ways you can bring material in. In later lessons (all of Module 2) we go deep on every source type and how to choose them well. Right now we just need one source in the door, so let's walk through the two or three easiest methods and you can pick whichever fits what you have.

📄 Method A — Upload a PDF (or other file)

Choose the upload option and pick a file from your computer — a PDF is the classic choice, but NotebookLM also accepts other document types. This is perfect for a research paper, a report, a manual, or an ebook chapter you already have saved. The file uploads, and NotebookLM begins reading it.

📋 Method B — Paste text directly

Choose the paste text / copied text option, then paste in any text you've copied — an article, an email, your own notes, a passage from a book. This is the fastest way to get something in front of NotebookLM in ten seconds flat, and it's a great fallback when you don't have a tidy file. Give the pasted source a short title so you recognize it later.

📝 Method C — Add a Google Doc

Choose the Google Drive option and pick one of your Google Docs (or Slides, or a PDF stored in Drive). Because you're already signed in with your Google account, your Drive is right there. This is ideal if your material already lives in Docs — notes you've been keeping, a draft you're working on, a shared document from a class or team.

💡 Which method should I use right now?

Whatever gets a real, relevant source in fastest. Have a PDF handy? Upload it. Working from a web article? Copy the text and paste it (we'll cover adding web-page and YouTube links directly in Module 2). Everything already in Google Docs? Add it from Drive. There's no wrong choice for this first run — the point is to feel the flow, not to build the perfect notebook yet.

⚠️ Important Note: The exact wording and icons on the "Add source" menu shift over time and can differ a little by account type. If an option you expect isn't where these notes say, scan the menu for the nearest equivalent — upload, paste, Drive, link — the categories are stable even when the labels move.

Processing and the Auto-Summary

After you add a source, you'll see NotebookLM take a moment to process it — a spinner, a "loading," or a brief pause. It's not stalling; it's reading. NotebookLM ingests the text, breaks it into passages, and builds an internal index so that later, when you ask a question, it can find the exact spot an answer comes from and cite it. The bigger the source, the longer this takes — usually seconds, sometimes a bit more for a hefty document.

When processing finishes, something nice happens automatically: NotebookLM writes a short summary of the source (often shown in the center) and offers a few suggested questions you can click to get started. These aren't decoration — they're a running start. The summary tells you, at a glance, that NotebookLM understood roughly what you fed it, and the suggested questions are a gentle on-ramp if you're not sure what to ask first.

✅ Pro Tip

Read that auto-summary before you do anything else. If it describes your source accurately, great — grounding is working. If it seems off (say it summarizes the wrong document, or misses the main point), that's an early signal to check your source: maybe a PDF was mostly images with no readable text, or you pasted the wrong passage. Catching a bad source now saves you from confused answers later.

graph LR A["Create notebook
and name it"] --> B["Add a source
upload, paste, or Drive"] B --> C["NotebookLM processes it
reads and indexes"] C --> D["Auto-summary and
suggested questions appear"] D --> E["You are ready
to chat"]

Ask Your First Grounded Question

Now the fun part. Click into the chat box in the center panel and ask a real question about your source — in plain English, the way you'd ask a knowledgeable friend. You can click one of the suggested questions, or type your own, such as:

  • "What is the main argument of this document?"
  • "Summarize the key findings in five bullet points."
  • "What does this say about [a specific topic you care about]?"

Press enter, and after a short think, NotebookLM writes an answer drawn only from your source. This is the moment the abstract idea from Lesson 1.1 becomes real: it's not answering from the whole internet or a fuzzy memory — it's answering from the exact document you just added, and nothing else.

🧠 Mindset

Ask it something you already know the answer to from your source, on this first go. Why? Because it lets you check NotebookLM against your own understanding and feel — firsthand — how tightly grounded the answer is. Watching it correctly pull a fact you recognize is far more convincing than any explanation in a lesson. Trust that you build by verifying is the trust worth having.

And notice what happens if you ask about something your source doesn't cover: a well-behaved NotebookLM answer will tell you it can't find that in your sources, rather than inventing an answer. That honest "it's not in here" is a feature, not a failure — it's the boundary of grounding doing its job.

Click the Citation — the "Aha" Moment

Look closely at that answer. Sprinkled through it are little numbered markers — citations. Each one is a link back to the precise spot in your source that the sentence came from. This is the single most important feature in all of NotebookLM, so let's make a ritual of it right now.

Click a citation. NotebookLM opens the source and jumps you straight to the exact passage — often highlighting the very sentence — that it used to write that part of the answer. Read it. Does the source actually say what the answer claimed? Almost always: yes. And that is the aha — the AI didn't just assert something, it showed you receipts. You are never asked to take its word.

graph TD A["You ask a question
in the chat"] --> B["Grounded answer
appears with citations"] B --> C["Click a numbered
citation marker"] C --> D["Jump to the exact
passage in your source"] D --> E["Confirm the source
really says it"] E --> F["Trust the answer
with your own eyes"]

⚠️ Watch Out

Grounding dramatically reduces hallucinations, but it doesn't make NotebookLM perfect. It can still misread a table, over-summarize, or occasionally cite a passage that only loosely supports the claim. The citation is your safety net precisely because you use it — an answer you didn't verify is just a confident guess. Get in the habit now: read the answer, click the citation, confirm the source. Module 3 turns this into second nature.

Click the citation. Two words that separate people who use AI well from people who get burned by it. Make it a reflex today.

The Core Loop You'll Use Forever

Step back and notice what you just did, because you'll repeat this exact rhythm in every notebook you ever build. Strip away the specifics and it's three moves:

Step What you do Why it matters
1. Add a source Give NotebookLM material — a PDF, pasted text, a Doc, a link It can only ever answer from what you've added; sources are the foundation
2. Ask Ask a plain-English question in the chat You get a grounded answer drawn only from your sources
3. Verify the citation Click the numbered marker and read the passage You confirm the answer with your own eyes instead of trusting blindly

Add source → ask → verify. That loop is the beating heart of NotebookLM. Every fancy thing coming later — Audio Overviews, Mind Maps, study guides, timelines — is built on top of it. Master this simple cycle now and the rest of the course is you getting faster, sharper, and more creative with a loop you already own. Answers stay inside your sources by design, so the more thoughtfully you feed and interrogate them, the better everything downstream becomes.

🎯 Project: Build Your Real Notebook

No more sample notebooks — this is the real thing. You'll turn the Notebook Brief you wrote in Lesson 1.1 into an actual, working notebook and run the core loop end to end. This notebook is the one you'll grow across the entire course, so use your genuine topic and real sources.

🏋️ Add a source, ask three questions, verify a citation

Objective: Create your course notebook with at least one real source, ask it three questions, and click at least one citation to see the source passage for yourself.

Instructions (about 20 minutes):

  1. (2 min) From the notebook home, click "Create new" and rename the notebook to match your Brief (e.g. the driving question or topic).
  2. (4 min) Add at least one source from your Brief using any method — upload a PDF, paste text, or add a Google Doc. If it's quick, add a second so the notebook feels real.
  3. (2 min) Wait for processing to finish, then read the auto-summary. Does it describe your source accurately? Note your reaction.
  4. (6 min) Ask three questions in the chat: one you already know the answer to (to test grounding), one genuine question you actually want answered, and one about something you suspect isn't in the source (to see how it handles "not found").
  5. (4 min) On at least one answer, click a citation. Read the passage it jumps to. Confirm the source really says what the answer claimed — or note it if it doesn't.
  6. (2 min) Jot one sentence in your journal about how the "click the citation" moment felt.
💡 Hint — if something feels stuck

Chat box greyed out? Make sure at least one source has finished processing (the spinner is gone). Answer looks wrong or thin? Check that your source actually contains readable text — a scanned PDF that's really just images may have little for NotebookLM to read; paste the text directly instead. Can't find citations? They're the small numbered markers inside the answer text; hover or click them. And if the interface looks different from these notes, remember it's the same Sources / Chat / Studio map underneath.

✅ Project Completion Checklist

  • You created and clearly named your course notebook
  • You added at least one real source and it finished processing
  • You read the auto-generated summary of your source
  • You asked three questions in the chat
  • You clicked at least one citation and read the source passage it opened
  • You noted whether the source actually backed up the answer

🎯 Quick Quiz

Question 1: Why can't you chat with a brand-new notebook until you've added a source?

Question 2: What happens when you click a numbered citation in an answer?

Best Practices for Your First Notebook

✅ Do's

  • Name the notebook immediately. Future-you, staring at a shelf of notebooks, will be grateful.
  • Read the auto-summary as a sanity check. It's your fastest signal that a source loaded correctly.
  • Click at least one citation every session. Make verifying a reflex, not a chore you skip.

❌ Don'ts

  • Don't dump unrelated topics into one notebook. Focused sources give focused answers; keep one notebook to one topic.
  • Don't accept an unverified answer on anything that matters. The citation is right there — use it.
  • Don't panic if it says "not in your sources." That honesty is grounding working, not the tool failing.

💡 Pro Tips

  • Try the suggested questions first — they're a fast way to see what your source can answer well.
  • If a scanned PDF gives thin answers, paste the text directly instead; NotebookLM needs readable text, not just an image of a page.

📓 Learning Journal

Keep a learning journal as you work through this course — a document, a note, or even a page in NotebookLM once your notebook exists. After each lesson, take a few minutes to write down:

  • Key concepts you learned
  • Techniques that clicked for you
  • Questions or confusion points to revisit
  • Ideas you want to try
  • Your progress and feelings about learning this — especially how the first grounded answer landed

✍️ This lesson's prompt: Describe the moment you clicked your first citation and saw the source passage. Did the source actually back up the answer? How did it change — grow or shrink — your trust in the tool? And what's the first real question you're excited to point this notebook at? This is a fun one to re-read later, when the loop feels automatic.

📝 Lesson Summary

🎓 Key Takeaways

  • You create a notebook, name it, and add your first source by upload, paste, or Google Drive — one topic per notebook.
  • NotebookLM processes each source (reading and indexing it) and then auto-writes a summary and suggested questions.
  • You ask questions in plain English and get grounded answers drawn only from your sources — with a well-behaved "not in your sources" when something's missing.
  • Every answer carries inline citations; clicking one jumps to the exact passage so you can verify with your own eyes.
  • The core loop is add source → ask → verify citation — the foundation everything else in NotebookLM is built on.

🎉 What You've Accomplished

You built a real notebook, added a real source, and had NotebookLM answer a real question — and then you did the thing most people never learn to do: you checked, by clicking a citation and reading the source for yourself. That's not a beginner trick; it's the entire discipline of using AI well, and you've already got it. Everything from here is you getting more powerful with a loop you now own.

❓ Common Questions at This Stage

My answers seem thin or off — did I do something wrong?

Usually it's the source, not you. A scanned PDF that's really just images has little readable text for NotebookLM to work with — paste the text directly instead. A source that's off-topic or too short also gives weak answers. Remember: the chat can only ever be as good as what's in the Sources panel. Module 2 is all about choosing and preparing sources well.

Can I add more sources to the same notebook later?

Absolutely — that's the plan. You can keep adding (and removing) sources in the Sources panel, and the more relevant material you add, the richer your answers get. You can also select which sources are active to focus the chat. We build your notebook up throughout Module 2.

What if a citation doesn't quite match what the answer said?

Good catch — that's exactly why we verify. Grounding is strong but not perfect; occasionally an answer over-summarizes or cites loosely. When a citation doesn't support the claim, trust the source over the summary and rephrase your question. Spotting this is a skill, and clicking citations is how you build it.

🔭 Looking Ahead

That's Module 1 complete — you understand NotebookLM, you're set up, and you've run the core loop. Next comes Module 2, where we go deep on sources: every type you can add (Lesson 2.1 covers PDFs, Docs, Slides, and text; Lesson 2.2 covers web pages, YouTube videos, and audio), how to choose good ones, and how to get the most out of each. Since sources are the foundation of everything, this is where your notebooks start getting genuinely powerful.

✅ Before the Next Lesson

  • Have your named notebook with at least one working source ready to build on
  • Make sure you've clicked at least one citation and seen the source jump for yourself
  • Write your Learning Journal entry for this lesson

📚 Additional Resources

🌟 Encouragement for the Journey

You just felt the aha — the moment an AI stopped asking for your blind trust and started showing you its receipts. That feeling is why this tool is worth mastering, and you got there in three lessons. The loop you learned today, add → ask → verify, is one you'll use for years. Now let's go make your sources great. Onward to Module 2. 📔