Image by Editor
# Introduction
Creating a product requirements document (PRD) is a common process in product management and a commonplace task in sectors like software development and the tech industry as a whole. Some of the typically found difficulties and hard requirements in creating a PRD include ensuring clarity, preventing scope creep, and preserving stakeholder alignment.
Thankfully, AI tools have risen to help navigate these challenges more effectively, without completely delegating the strategic decision-making underlying the PRD creation process — in other words, with the human still in the loop. One example is Google’s NotebookLM, which synthesizes grounded raw data or materials to answer questions, thereby turbocharging the workflow for creating grounded, useful PRDs.
This article will navigate you, based on a beginner-friendly use case, through the process of using NotebookLM’s features to turn raw, sometimes chaotic information into a grounded PRD in a matter of minutes. Spoiler: it won’t be just about chatting with an AI assistant.
# From Messy Notes to a Structured PRD Draft
Let’s consider the following scenario. You are the newly hired product manager for a startup that wants to develop a new mobile app called FloraFriend. The goal of the app is to help people stop accidentally killing their houseplants.
The team, including you, has collected a set of three “messy” documents that contain descriptions for what the potential app should be like:
- interview_transcript_matt.txt: a 30-minute interview with a user called Matt, who is the owner of over 50 plants. In these interview notes, Matt says existing apps are “overly complicated” and make it difficult to retain in mind aspects like “which fertilizer to use.”
- competitor_research_notes.txt: a rough list of bullet points made after analyzing competitor apps like “PictureThis” and “Planta”, highlighting their paywalls and interface drawbacks.
- brainstorming_whiteboard.jpg: random but somewhat “cool” ideas that have been mentioned by the team during lunch breaks and other casual conversations, e.g. “spotify playlists for plants”, “watering reminders”, and so on.
Imagine full documents containing all of the content described above. Manually turning these into a clean PRD that nicely brings it all together may sound like a pain, right? Enter NotebookLM!
Log in to NotebookLM with your Google Account and click “Create New Notebook“. Give your new notebook a name, something like “FloraFriend PRD.”
Once the new notebook has been created, you’ll be welcomed to the main NotebookLM interface, which looks like this:
NotebookLM Interface
A word of caution: this newly created notebook is not intelligent per se. It is not a regular large language model (LLM); it does not know plant care or any other specific topics. But we are about to teach it an “express” Master’s degree about it with our messy — yet enlightening for the tool — notes.
Suppose you have the three above mentioned files with some content related to the plant care app, or any other raw information files of your own. You can upload them to the NotebookLM canvas by using the upload button in the main, central section.
Once uploaded, you can think of your notebook as something similar to a tiny, toy-sized retrieval-augmented generation (RAG) system that can start thinking and behaving AI-like based on the information it has access to. In fact, without asking it, by clicking on either one of the uploaded files on the left-hand side, NotebookLM generates a concise, well-organized summary of the contents in that file: this is called a file’s Source guide.
Now comes the key part. We could simply ask in the chat box at the bottom something like “Write a PRD”, and that’s it. But we want to do this properly and provide clear, specific instructions, and that entails some prompt engineering, namely to force the newly born AI to prioritize what we want our PRD to reflect: prioritizing the user problems over the random ideas generated by the team (without totally neglecting them). Here is a well-crafted prompt that works:
I am the product manager for FloraFriend. Based only on these sources, draft a PRD.
Crucial constraints:
1. Prioritize features that solve the pain points mentioned in interview_transcript_matt.txt.
2. Exclude any ‘brainstorming’ ideas that don’t directly address a user problem.
3. Structure the output with these headers: Problem Statement, Core Features, Non-Functional Requirements (UI/UX), and Success Metrics.
Try adapting this prompt to your own business problem or use case. Once sent, chances are you will get a nice and clean PRD with key sections like Problem Statement, Core Features, Non-Functional (UI/UX) Requirements, Success Metrics, and so on.
Interestingly, the PRD contains something that looks like numerical citations you can hover on. If you do so, you will see the source (one of the source files) pop up:
Before accepting this first PRD as it is, remember that a first draft is rarely perfect. Keep engaging in conversation to gradually refine it, e.g. if you notice there is a missing monetizing section, ask: “Based on the competitor_research_notes.txt, what monetization models are our competitors using, and what should we avoid?“. After that, manually check the outputs, make sure they are consistent with the rest of the first PRD draft, and incorporate the main monetization insights into it, either manually or by asking NotebookLM’s AI to do so — if you opt for the latter, always check what you get before blindly approving it. Remember: AI can make mistakes!
The icing on the cake is the Audio Overview section on the right-hand panel (Studio). By just clicking on it, you will generate an audio overview of the information contained in the source files. This is an excellent way to absorb information when reading might be less appealing, e.g. while you are on your daily commute.
# Next Steps
This article introduces NotebookLM’s capabilities to generate grounded PRD specifications from raw, messy documents in a matter of minutes, taking very easy steps. From here, a worthwhile next step could be resorting to Google’s Antigravity to turn your PRD specification into a functional software prototype.
Iván Palomares Carrascosa is a leader, writer, speaker, and adviser in AI, machine learning, deep learning & LLMs. He trains and guides others in harnessing AI in the real world.
