How I Turn Zoom Meetings Into High-Value Content Using AI

Many people struggle to come up with ideas. Or, they have ideas but don't have time or patience for the writing process. If you're more of a talker and less of a keyboard warrior but want to publish your ideas to the world, this is the content-creating strategy you've been looking for.

This strategy also works well for customer stories and gathering insights from busy CEOs—use cases where it's often a struggle to get someone to commit to writing a few hundred words. 

I often use Zoom and AI to solve these content creation problems. With minimal effort, a one-hour call can become a 1,500-2,000-word article. Best of all, you retain the author/speaker's voice, tone, and cadence, infusing their personality in the final piece.

Getting Started — The Tool Stack

I typically use Zoom and ChatGPT for this process. You must have a paid Zoom account to enable recording with transcripts. Google Meet is a great free alternative, with transcripts published in a Google Doc. 

Admittedly, I have done limited testing with Google Meet, so if you're after a free alternative, you'll have to test the transcript quality yourself.

Otter.ai is an excellent tool for existing videos. For example, if you're a YouTube creator and want to accurately convert your videos to text, I highly recommend Otter.ai. I've used it successfully a few times.

Next, you'll need a paid ChatGPT account. You might get away with the free version, but there's a lot more you can do with ChatGPT Plus, especially with the new OpenAI o1-preview and OpenAI o1-mini models.

Lastly—and this one is optional—I highly recommend Grammarly for cleaning up grammar and spelling and tightening up the language overall. Again, I use the paid version for its premium features, transforming my funky writing into something readable and digestible.

Tool stack recap:

• Zoom (paid) or Google Meet (free)
• Otter.ai (paid) for existing videos—also connects to Zoom
• ChatGPT (paid) for editing transcripts fast
• Grammarly (paid) for polishing and editing

Prepping Your Video Call

Whether interviewing someone or recording myself, I always do a quick prep with questions I want to answer. ChatGPT is an excellent tool for prepping the structure.

Usually, I have an idea for the article and a rough outline of the topics I want to cover, but this should be as comprehensive and valuable as possible, so I ask ChatGPT.

I prompt ChatGPT with a brief on the article, its goals, audience, etc., and ask for a comprehensive outline. I combine this outline with my ideas and feed them back into ChatGPT, asking for a list of questions I can use to guide the video call.

These questions will usually become the headings and subheadings for the article, and now you're ready for the video call.

The Recording Process

You only need Zoom and your subject for recording. You can also use Zoom to record yourself talking and follow the same process to create text and video content. 

Make sure you are the host of the Zoom meeting. If someone else is the host, they must have a paid account.

In the Zoom navigation bar, hit Record and then "Record to the cloud." Do not use "Record to this computer." You lose Zoom's automated transcript creation feature when you record and save to your computer and must then use a tool like Otter.ai to extract the transcript.

I run these video calls like a podcast/interview/meeting hybrid. I follow the question structure but often go off track or probe topics if something interesting arises.

Often, you cover the same topic with different insights at different parts of the call. Don't worry; this is normal and keeps the conversation flowing naturally. ChatGPT will identify and combine these topics into a single subheading later—a massive time saver.

Prepping ChatGPT

If you have a paid ChatGPT account, you can create GPTs for specific tasks. I won't go into how to customize a GPT in this article, so check out OpenAI's docs.

Inside the GPT's "Instructions," I provide an overview of what I want the model to do and how I want it to process what I send via the chat. I also include the questions I asked during the interview so the model knows what to look for and how to organize content.

Example Prompt:

"You are an AI assistant that helps edit interview transcripts into coherent articles. Please organize the content under relevant headings and maintain the original speaker's voice and tone."

I find the more instructions I can add to the configuration, the better the outcome and the shorter my prompts need to be to achieve that outcome.

Your GPT is now ready for the transcript.

Converting The Transcript Into An Article

If you're using Zoom, go to Recordings & Transcripts in your account and find the meeting you recorded. Open the meeting by clicking on it and hover over "Audio transcript" to view the download icon.

The transcript will download as a basic text file. Don't worry about editing this; ChatGPT does an unbelievable job in seconds. I usually add a short prompt at the top of the transcript informing the GPT who is who—for example, if I were interviewing Jane, "Here is an interview with Jane; her answers appear next to "Jane:". Please convert this into a comprehensive blog post only using Jane's answers."

The GPT will use the instructions from the backend configuration you set up in the previous step to convert the transcript into an article with headings and subheadings in seconds. 

The results still blow me away every time I do this. An article that would typically take me a few hours to create is done in less than 90 minutes—including the 60-minute call and prep time. And it's impossible to tell if AI was involved; it's as if the interviewee wrote the article.

Cleaning Up and Publishing

The final step is a Grammarly clean-up. I usually dump the entire article from ChatGPT into Grammarly and accept or ignore changes. The goal is to maintain the voice and tone but ensure it's clean and easily digestible.

Lastly, as laid out in my Distributed Content Strategy white paper, I like to publish on multiple channels, including LinkedIn, Medium, X, and the associated blog and email newsletter. This type of content performs well across social channels because it's unique and authentic. 

I'm interested to hear how you create original content in an AI world. Leave a comment 👇