Using NotebookLM to Turn Competitive Research into a Private GTM Podcast
TL;DR — Competitive research that lives in a 40-page PDF gets read by almost no one. The same research, synthesised into a 12-minute private audio briefing through NotebookLM, gets listened to on commutes, between calls, and before big meetings. Format is not a cosmetic choice. It is a distribution strategy.
Forty Pages. Zero Listeners.
I uploaded 40 pages of competitive research into NotebookLM and got a 12-minute podcast I could listen to on my commute.
My sales team actually listened to it. That last sentence is the one that matters.
Here is some context: I had spent three weeks assembling that research. Competitor battle cards from the last two quarters. Two analyst reports covering the category. Win/loss interview summaries from eight closed deals. Feature comparison matrices I had updated manually. It was genuinely thorough work, and I was proud of it.
I shared it with the sales team as a Google Doc. Analytics showed me that four people opened it. Two of them scrolled past the first page. One person printed it and I assume it is now in a drawer somewhere.
This is not a criticism of sales teams. It is a recognition of how people actually consume information. A dense document full of competitive nuance is not a format that works for someone who has six calls on their calendar and is mentally preparing for all of them. An audio briefing they can absorb during the ten-minute walk from the parking garage — that is a format that fits how they actually live.
NotebookLM’s audio overview feature is one of the most practically useful things I have found in the AI tooling landscape for PMMs, and it is consistently underestimated because it does not look like a workflow tool. It looks like a research tool. The insight is that it is both.
What NotebookLM Actually Does
NotebookLM is a research assistant from Google that allows you to upload a set of source documents and then interact with them — asking questions, requesting summaries, generating study guides. The sources are treated as the authoritative context; the model works from what you upload, not from its broader training data.
The audio overview feature takes this one step further. It synthesises the source documents into a conversational audio summary — two AI hosts discussing the key themes, insights, and tensions in the material. The format is genuinely listenable: not a robotic read-through of bullet points, but a dialogue that surfaces the interesting parts and contextualises them for someone who has not read the underlying documents.
For PMMs, this is significant for a specific reason: it solves the source anchoring problem at the synthesis layer. The audio content is generated from your uploaded documents, not from the model’s general knowledge about your competitors. When the podcast discusses a competitor’s pricing model, it is discussing what your battle card says about that model — not what the model infers from training data about that company.
The GTM Briefing Podcast Workflow
Here is the exact workflow I use, without the Make.com trigger layer initially — this one can be run manually because the value is so high even without full automation.
Step 1: Assemble Your Sources
Gather into a single folder the competitive research that is most relevant to your current quarter’s deals:
- Current battle cards for your top two or three competitors
- Recent win/loss interview summaries (anonymised is fine)
- Relevant analyst report excerpts — even a few key pages, not the whole report
- Any recent competitive intelligence from deal notes or sales calls that you have documented in a structured format
The source set does not need to be enormous. NotebookLM works well with five to fifteen focused documents. More is not better — focused and current is better.
Step 2: Upload to NotebookLM and Generate the Audio Overview
Upload your source documents to a new NotebookLM notebook. Once the sources are processed, request an audio overview. NotebookLM will generate a conversational audio summary — typically eight to fifteen minutes depending on the volume of source material.
Listen to the first version yourself before you share it. You are checking two things: that the audio is covering the highest-priority competitive themes, and that there are no significant gaps or misrepresentations. Because the content is grounded in your uploaded sources, the accuracy should be high — but your own review is still the quality gate.
Step 3: Guide the Output With Source Curation
If the first version is not covering the right themes, the fix is usually in the sources rather than in trying to prompt the model differently. Competitive themes that are well-represented in your battle cards will surface prominently in the audio. Themes that are thin in your source documents will be thin in the output.
This is actually useful diagnostic information. If NotebookLM is producing a briefing that underemphasises your key differentiator, that is a signal that your battle cards do not make that differentiator explicit enough. The audio overview is testing your competitive documentation the same way a well-prepared sales rep would — and surfacing the gaps before they surface in a live deal.
Step 4: Share as a “GTM Briefing” Audio Asset
Export the audio and share it through whatever channel your sales team uses most reliably. A Slack message with the audio file attached and a one-paragraph summary of what the briefing covers tends to perform better than a shared folder link. The message itself is a signal that this is worth their time, not just another resource in the resource library.
I frame mine explicitly: “This is a 12-minute competitive briefing you can listen to before your next [competitor] displacement call. Covers their Q1 positioning changes, the three objections we are seeing most, and the win patterns from our last eight competitive deals.” Specificity in the framing drives listens.
Connecting This to Your Agentic Workflow
The manual version of this workflow is already high-value. The automated version connects it to the broader GTM Infrastructure.
A Make.com scenario can watch for new competitive intelligence inputs — a freshly updated battle card saved to a shared Drive folder, a new win/loss summary submitted through a form, a competitive alert from your monitoring flow — and trigger a NotebookLM source update. Set a recurring scenario to regenerate the audio overview weekly or when a source document has changed since the last generation.
The output is a living competitive briefing that stays current with your competitive intelligence cycle, not a static podcast you made once and gradually stopped maintaining.
For sales teams operating in a fast-moving competitive environment, this is genuine GTM infrastructure — not a content novelty but a reliable, consistently updated intelligence distribution channel that meets reps in the format they can actually use.
A Note on What This Is Not
NotebookLM audio overviews are not a replacement for deep competitive knowledge. They are a format conversion and an accessibility layer for research that already exists.
The research still has to be done. The battle cards still have to be maintained. The win/loss interviews still have to happen. NotebookLM synthesises what you give it and makes it listenable — but the quality of the output is entirely determined by the quality of the inputs.
This is actually the strongest argument for investing in proper competitive documentation. For years, teams have done the research and then watched it disappear into documents no one read. Making that research audible is the unlock that makes the investment in doing the research worthwhile.
Key Takeaway for PMM Leads
The problem with competitive research is rarely that it does not get done — it is that it does not get consumed. NotebookLM’s audio overview feature converts dense competitive documentation into a listenable briefing that fits how sales teams actually operate. The workflow is simple, the source anchoring is native to the tool, and the format shift from document to audio is the difference between research that sits in a folder and research that influences the next call. This is one of the highest-ROI things you can do with an afternoon and a set of documents you probably already have.
Frequently Asked Questions
Is NotebookLM audio output accurate enough to trust in a sales context?
Accuracy depends directly on the quality of the source documents. NotebookLM generates from what you upload, not from its general training data, which means the audio content reflects your battle cards and research rather than the model’s inferences. Review the audio yourself before sharing — specifically checking that competitive claims match your source documentation. In my experience, when the source documents are current and specific, the audio output is reliable enough for a sales briefing after a single review pass.
Can NotebookLM handle confidential customer information from win/loss interviews?
This is a data governance question worth taking seriously. NotebookLM is a Google product, and your organisation’s data policies regarding Google Workspace tools apply. For fully anonymised win/loss summaries — where customer names and identifying details have been removed — the risk profile is typically acceptable for most organisations. For sensitive deal information, check with your security or legal team before uploading. A best practice is to create a “sanitised for tools” version of win/loss documentation that strips identifying information while preserving the competitive and strategic insights.
How often should you regenerate the competitive briefing podcast?
The right cadence depends on how fast your competitive environment moves. For most B2B SaaS categories, a monthly regeneration tied to your regular competitive intelligence review is sufficient. In fast-moving markets or during periods of active competitive disruption — a competitor raising a significant round, launching a major feature, or changing their pricing — a more responsive update cycle is warranted. The Make.com automation approach, triggered by source document updates rather than a fixed schedule, is the most reliable way to ensure currency without creating a manual maintenance burden.
What is the best way to introduce this to a sales team that is sceptical about AI-generated content?
Do not lead with “AI.” Lead with “here is a 12-minute briefing on [Competitor X] that covers everything from our last eight competitive deals.” Let the format and the quality of the content make the case. Most sales reps become immediate converts the first time they use a competitive insight from the audio briefing in a call that same day. The format serves them in a way that a PDF never did, and that practical utility is far more persuasive than any explanation of how the audio was generated.