The Executive's NotebookLM Guide
How to Use NotebookLM: A Business Guide to the AI That Only Knows Your Documents
How to use NotebookLM, in one sentence: create a notebook at notebooklm.google.com, upload the documents you want to work with—board packs, contracts, SOPs, meeting notes—then ask questions in plain English and get answers with citations back to your own pages. This guide walks through each step, with five real executive use cases.
1-on-1 with Pete Enestrom, on your actual work
Written by Pete Enestrom
Yale & Columbia, ex-Microsoft & Intel — 1-on-1 AI coaching for executives
Your coach’s background
NotebookLM is a free AI tool from Google that answers questions about the documents you upload—and only those documents. Add a board pack, a contract, a procedures manual, or a year of meeting notes, and NotebookLM responds with answers drawn from your pages, each claim cited to the exact passage it came from. When the answer isn't in the material, NotebookLM says so instead of guessing. For an executive whose real question about AI is “where does it get its answers,” that design is the whole story.
Among the tools I coach with, NotebookLM is the gentlest on-ramp for document-heavy executives, for one reason: your material does the heavy lifting. There's no prompt craft to learn at the start and nothing to install—NotebookLM runs in a browser, and the skill is uploading the right pile and asking plain questions. The founders and CEOs I work with don't need to be convinced AI is impressive. They need one place where AI is obviously, verifiably useful on their own paperwork. NotebookLM is that place.
This guide is the walkthrough I give clients in week one, written down: five steps from a blank notebook to a shared team resource, five real queries worth copying, and the honest limits to understand before anything confidential goes in. Everything here works on the free version.
Getting started: five steps from a blank notebook to a team resource
The sequence below is the same one I walk clients through—from sign-in to a notebook your whole team can use. An ordinary Google account is all you need.
Create a notebook at notebooklm.google.com
Sign in with a Google account and click “New notebook.” Name the notebook after a real project—“Q3 board prep” or “Supplier contracts”—not “test.” A notebook is a workspace scoped to one topic, and a good name is what keeps a growing library of notebooks findable three months from now.
Add your sources
Upload the material the notebook is about: PDFs, Google Docs and Slides, web pages by URL, YouTube videos, audio files, or pasted text. A single notebook holds dozens of sources, and each source can run to hundreds of pages. The board pack, the draft contract, the employee handbook—if the answer should come from a document, the document goes in here.
Ask questions and check the citations
Type questions in plain English—“what does the agreement say about renewal?”—and NotebookLM answers only from your sources, with numbered citations after each claim. Click a citation and the source opens to the exact passage. When the answer isn't in your documents, NotebookLM tells you. That honesty is the whole point of the tool.
Generate a briefing or an Audio Overview
Beyond the chat, NotebookLM assembles your sources into ready-made formats: a briefing document, an FAQ, a timeline, a study guide. The standout is the Audio Overview—a podcast-style conversation between two voices that walks through your material. Download the audio, and the commute or the treadmill becomes document-review time.
Share the notebook with your team
Share a notebook the way you'd share a Google Doc, and colleagues can ask their own questions of the same sources. Send the board-pack notebook to your directors before the meeting, or give the handbook notebook to every new manager. Each person needs a Google account, and on company Google Workspace accounts the administrator's settings apply.
What is NotebookLM used for in a real business?
NotebookLM is used for one job: answering questions about a specific pile of documents, with receipts. Executives upload board packs, contracts, standard operating procedures, meeting notes, industry reports, and interview transcripts—the material that piles up faster than anyone can re-read it. NotebookLM then acts like an analyst who has read only that pile and nothing else. Ask which clauses in a draft agreement create liability, and NotebookLM answers with the contract's own sentences. Ask what the board pack says about inventory, and the answer cites the page. Homework help made NotebookLM famous with students; the same grounding is what makes NotebookLM quietly useful inside a company.
The uses that stick with the founders I coach are unglamorous: preparing for a board meeting without re-reading ninety pages, letting new managers get answers from the employee handbook without interrupting HR, turning six industry reports into a short audio briefing for the drive home. Nobody in those scenarios is doing “AI.” Each person is getting answers out of documents the company already owns—which is exactly why NotebookLM is where I start executives who are skeptical of the whole category.
Five queries worth copying from real executive work
These are the queries that land with the founders and CEOs I coach. Copy them verbatim, replace the brackets with your material, and keep the instruction to cite—citations are what separate a grounded answer from a confident guess.
I've uploaded this quarter's board pack. Give me three things: the five numbers that changed most since last quarter and what the documents say about why; every decision the pack asks the board to make, with the page each appears on; and the three questions a sharp outside director is most likely to ask me. Cite the source passage for each point.
Click any citation and NotebookLM jumps to the exact paragraph in the source. That's the feature that makes board prep defensible.
I've uploaded a draft agreement from [counterparty]. Summarize the payment terms, the renewal and termination clauses, and every clause that creates liability for us. Quote the exact sentence behind each summary point, and flag anything unusual for a contract of this type. This is preparation for a call with our attorney, not a substitute for one.
NotebookLM reads contracts; it doesn't practice law. The quotes it pulls make the attorney call shorter and cheaper.
I've uploaded our operations manual and employee handbook. A plant manager asks: what is the exact process when an employee reports a safety incident, from first report to final filing? Answer step by step, cite the section each step comes from, and list anything the documents don't cover.
The last instruction matters. Gaps in your own documentation are often the most valuable thing NotebookLM surfaces.
I've uploaded notes from our last [number] leadership meetings. Which topics came up more than twice without a recorded decision? For each recurring topic, list the meetings where it appeared, who said they would own it, and what the latest note says. Skip anything we actually closed.
Run this before your next leadership offsite. The list of unresolved loops writes half the agenda.
I've uploaded [number] industry reports. Before I generate an Audio Overview, write a one-page briefing: the three findings that matter most for a [your industry] company of about [your size], and the two claims I should be most skeptical about. Cite the report and page for each finding.
Then generate the Audio Overview and download it. The written page is for the meeting; the audio is for the drive home.
Should you use NotebookLM or ChatGPT?
NotebookLM and ChatGPT answer different questions. ChatGPT is a general assistant: ChatGPT drafts, edits, brainstorms, and reasons from broad training on public text, which makes it flexible and occasionally wrong in confident ways. NotebookLM is deliberately narrow: NotebookLM knows only the sources you upload, and every answer arrives with citations into those sources. For interrogating your own material—the contract, the board pack, the handbook—NotebookLM is the safer instrument. For writing the email, the memo, or the speech that comes afterward, Claude or ChatGPT is the better one.
My own workflow uses both. NotebookLM turns a pile of documents into grounded answers and a short audio briefing; those answers then become the source material I paste into Claude when something needs to be written well. The tools don't compete so much as hand off to each other. If you're choosing where to start: start with NotebookLM when your work is reading-heavy, and start with Claude or ChatGPT when your work is writing-heavy. The Claude guide and the ChatGPT-for-work guide on this site cover the second tool when you're ready for it.
What are NotebookLM's honest limits?
NotebookLM is the gentlest entry point I've found for document-heavy executives, and NotebookLM is still a product with edges. Five limits are worth knowing before you upload anything.
- NotebookLM is Google's tool. Using NotebookLM requires a Google account, and notebooks live on Google's servers. In a company that runs on Microsoft, that's a policy question, not just a preference.
- The ecosystem locks in gently. Notebooks don't export cleanly to other AI tools, so the library you build stays with Google. Keep the original documents in your own drive and treat NotebookLM as a reading layer, not an archive.
- Your data policy comes first. Google states that NotebookLM doesn't use uploaded sources to train its models, but a vendor's statement isn't a company policy. Before anything confidential goes in, get your own written rules—the plain-English checklist in the ChatGPT privacy guide on this site applies to NotebookLM word for word.
- NotebookLM only knows what you upload. There's no live web inside a notebook, and a stale document produces a stale answer. Keep sources current, or the citations will be confidently out of date.
- Citations reduce hallucination; they don't replace judgment. NotebookLM can still misread a table or lean too heavily on one report. Verify the passage before you repeat a number in the boardroom.
When the tool stops being the hard part
NotebookLM you can set up tonight, alone, for free—and you should. What I've watched happen next, with client after client, is that the first notebook works, the questions get sharper, and then the real question surfaces: which of the company's document workflows—board prep, contract intake, the procedures nobody can find—should actually be rebuilt around this way of working, and who owns that change. The tool was never the bottleneck. The workflow is.
That's the point where a human teacher—not an AI coach bot, not another tutorial—is worth an hour of your time. In a 1-on-1 session we open your real document pile, build the notebooks and prompts around how your company actually runs, and set the privacy rules before anyone uploads something they shouldn't. This guide gets you to the first notebook. Coaching gets the practice into the business.
1-on-1 coaching
Want a teacher, not another tutorial?
Guides get you started. A 1-on-1 session gets you fluent—built on your actual work, at your pace.
Your coach
Coached by Pete Enestrom
Yale and Columbia grad, former Microsoft and Intel, and a venture-backed exited founder. Pete has spent the last four and a half years going deep on every major AI tool, and he teaches the way operators learn: on your real work, at your pace, with nothing assumed.
Common questions
Straight answers, the way I'd give them across a table.
Keep reading
Guide
How to Use Claude: A Practical Guide for Busy Executives
A plain-English guide to Claude for founders and executives: what it is, how to start in ten minutes, and five real workflows with copy-paste prompts.
ReadGuide
ChatGPT for Work: A Practical Guide for Non-Technical Executives
How non-technical executives use ChatGPT at work: drafting, summarizing, brainstorming, notes into action lists, and meeting prep—with six copy-paste prompts.
ReadResource
ChatGPT Privacy, in Plain English: What Really Happens to What You Paste
ChatGPT privacy in plain English for business owners: what happens to what you paste, the real risks, the settings to check today, and rules to hand your team.
ReadBeyond coaching
When the workflow needs to be encoded, not just learned — that's what our team at Zaigo builds.