Video Tutorial Production

I grew Jubilee Media to 14M subs on YouTube with $2k video budgets. 90% of brands can’t do the same with $200k budgets. Here’s how to build TV quality production on a tiny budget: Most brands overspend to make these videos because they're used to working with large-scale TV productions. But times have changed. Creating Netflix-quality content has a much lower barrier to entry. You really only need 3 basics: 1. Good audio 2. Good lighting 3. Good composition There’s a caveat to the budget of course: talent costs more than $2K. BUT… All you need to start making REALLY beautiful stuff is 1 producer (~$9k / mo) and 1 editor ~$8k / mo). That's it. They can get you 1 high-quality long-form episode every 2-3 weeks. We helped Proton scale to 100K subscribers in just 7 videos - from zero - with only a producer and editor. Cost of production is $3-5K per month… - Peerspace studio (7 hrs @ $150/hr) = $1050 - Talent (expert) = $500 - Table/Chairs (FB Marketplace) = $250 - 2x light rental (Aputure 300d) = $250 - Softbox + diffusion = $100 - C-stands rental (4) = $100 - Wireless lav system (4-channel rental) = $200 - Lunch (4 people @ $15/head) = $60 - Travel = $500 TOTAL = $3,010. You can shoot a couple episodes in one day and have a nice backlog to start. I promise you don’t need to make it complicated. I remember one time at Jubilee, we wanted to test the antithesis to this… We kept asking “what if we had a bigger budget” and eventually we tried it out. Turns out, the more we spent, the worse the content performed. You see there’s actually a tipping point with content - when things look too polished it actually hurts you because the YouTube platform and audience is built on authenticity. So beyond a certain budget, every dollar you spend on production actually makes it LESS likely to work. Now obviously, there are exceptions... Mr. Beast running trains into holes isn't cheap. But, for what most brands are trying to do... You don't need much. At the end of the day, winning with brand content isn’t about “expensive”. - You win subscribers by understanding the platform. - You gain traction by showing up consistently. - You retain viewers with quality content. This is what separates brands that grow REAL audiences, from brands that burn money on unenjoyable content. So if you have budget fears right now, reach out. I’d love to help you execute organic brand content the lean and effective way.

It's never a "quick" edit. Let's break down my process for creating a 2-minute video as an amateur: 1. Planning This involves meeting with everyone involved in the project (stakeholders, product teams, designers, etc.) to flush out the goals and purpose of the content. 2. Concepting Ideating and researching ways to bring the video to life. Depending on the platform, there are various nuances to consider beyond just the format. Sometimes, I spend hours just scrolling through different channels for inspiration. 3. Scripting This one is pretty straightforward but needs to be emphasized. Writing a script helps flush out the message so you're hitting all the points. Start with freewriting before making edits. Let the words fall out. 4. Storyboarding Before recording, I like to visualize the video's flow and structure. I've found that when I storyboard, especially when working with others, it's a lot easier to see if it makes sense for the overall project. You can do this using tools like Miro or even just a piece of paper to map out when things happen. 5. Recording Setting up the camera, lighting, and mics is just the beginning. Multiple takes are definitely required to capture the best footage. 6. Editing This is the most time consuming part. It involves more than just cutting and trimming footage. You need to spend time selecting the best takes, adding transitions, applying color correction (especially since iPhone cameras can be ugly af), and integrating audio (music, voiceovers, etc.). Every frame is carefully scrutinized to make sure the video flows smoothly, fits within 'safe zones,' and aligns with the overall goal and messaging. Then there are captions and graphics to consider. 7. Review This is the second most time consuming part because the more people involved in the approval stage, the tougher it becomes. Subjectivity plays a big role here—what one person likes might not align with someone else's preferences. This feedback loop can be tricky because there can be several rounds of feedback and adjustments, requiring you to jump back and forth between steps 6 and 7. 8. Export Finally, after all is approved, you need to go back and format and do a quality check, making sure everything fits correctly. Post-production tasks include exporting SRT files, creating thumbnails, and determining where the content will live after it's finalized. 📌 Just because a piece of content is lofi doesn't mean it's low effort. My tip for keeping the process smooth—and something I learned a little too late in the game—is that having a really clear brief at the beginning is the single most important aspect of any video project.

I used to spend an hour building a single motion graphic in After Effects. Now I build them in 3-5 minutes — and I haven't opened Premiere or After Effects in months. A quick bit of context: I'm a former LA filmmaker and still a member of the Producers Guild of America. I've worked on feature films, commercials, and music videos. So when I say I've replaced my editing stack, it's not because I didn't know the old one. My entire video workflow now runs on three tools: → Claude Code (the orchestrator — I describe the scene, it builds it) → HyperFrames (open-source video framework that renders programmatically) → Descript (transcript-based editing + final assembly) Here's the pattern: 1. I load a brand style guide into Claude Code once — colors, fonts, typography, reusable components. Every future scene inherits it automatically. 2. I paste a line from my Descript transcript into Claude Code and describe what I want the motion graphic to do. 3. Claude Code generates the HyperFrames scene, renders it to MP4, and drops it into my project folder. 4. I drag the MP4 into Descript on the matching line of the transcript. Done. The unlock isn't the tools individually — it's the brand style guide. Once Claude has persistent context on how your content should look, you stop prompting from scratch every time. The style guide is to video what a brand book is to design: one-time setup, compounding leverage. A few things this has changed for me: • Output velocity is up 10x on motion graphics • I can ideate visually at the speed I think • Non-developers on my team can now produce branded video scenes without learning a single editing tool If you make videos for your business, your brand, or your content — this is worth 20 minutes of your time to watch. Full walkthrough is in the comments below. I also published the full GitHub repo with skills, templates, and a music-from-transcript plugin pre-loaded — also linked in comments below. If you enjoy, please share with a friend 😃

AI doesn't edit your VSL. Let me be clear about that upfront. What AI does is eliminate the 6-8 hours your video editor spends hunting for stock footage and writing briefs for what they need. Here's how the production process actually works: Step 1: Your script is approved. You or the founder records A-roll — talking head on iPhone 14+ in a clean, well-lit room. Dark or simple backgrounds outperform busy ones for direct response. Use a teleprompter app. Record at 130-150 words per minute — slower than feels natural because the viewer is reading captions at the same time. Step 2: Feed the approved script into Claude with a B-roll extraction prompt. For every sentence or logical unit in the script, Claude outputs: scene type (problem visual, solution visual, proof visual, authority visual, lifestyle visual, product visual), a specific visual description in plain English, a generation prompt formatted for Midjourney or Runway, and whether the asset should be static or motion. Step 3: Output is a CSV — your production manifest. Script line, scene type, description, generation prompt, asset type. This goes directly to your video editor as their brief. Step 4: Generate all static assets in Midjourney. Generate all motion assets in Runway. Your editor now has every b-roll clip labeled, timestamped, and ready to drop in. Step 5: Your editor assembles the final cut — pacing, transitions, caption styling, color, the feel of the ad. This is the part you don't automate. The editor is what makes it feel like a real ad, not a slideshow. What this process eliminates: your editor spending hours searching stock footage sites, you going back and forth on "what kind of image should go here," and the miscommunication between creative direction and what actually gets produced. The AI generates assets. The human assembles them. That split is intentional. We've tested fully automated edits and they're not there yet. The editor's judgment on pacing and emotional timing is still the difference between a 2x and a 5x ROAS VSL.

Check out my AI workflow for creating video tutorials! I create a lot of software tutorials in my role doing technical training. Each video highlights a single use case on a platform, such as “Create a Customer” or “Generate an Invoice.” Incorporating AI into my process speeds everything up. So let’s say you met with a SME who walked you through the use case, maybe you have a recording or some supporting documentation. First, I generate a transcript of that SME presentation using Microsoft Stream or Clipchamp. These come with my Enterprise Microsoft license and are super easy to use. Then, I feed that transcript into Copilot and have it generate a video demo script. I then use another Copilot prompt from my prompt library to edit the script and apply my company’s style guide to ensure I have the right tone and voice. Next, I generate text-to-speech audio using the ElevenLabs voices in Articulate Storyline. I export this mp3, and produce the video in Camtasia using the AI audio and SME video. This part I still like to do manually. Finally, I export the mp4 from Camtasia back into Storyline and use their AI Captions Generator to perfectly sync captions to my video for increased accessibility. Voila! Increased efficiency, AI speed, and a human touch make technical training magic. Hope you found this helpful. #eLearningDevelopment #AI #InstructionalDesign #TechnicalTraining #LearningAndDevelopment #CustomerEducation Image Description: Employees in an office with headsets use AI on their computers.

You don’t need 7 different ideas to create 7 high-performing videos. You just need 1 topic and 7 angles. Want to create content that actually earns attention? Start by choosing a topic that solves a real problem your audience faces or sparks a moment of “I’ve thought that too.” The right topic creates relevance. Keep each video to 60 seconds or less so it delivers value quickly and fits the way people consume content. Context (C): I want to create engaging video content that effectively captures a single topic from 7 angles to maximize audience reach and impact. Role (R): Act as a skilled content creator with experience in video production and audience engagement strategies. Inspiration (I): Focus on a single topic and develop seven distinct angles for high-performing videos, utilizing various approaches to capture viewer interest. Scope (S): The angles should include: 1. Shock: Present surprising facts that engage viewers. 2. Comprehensive: Offer thorough insights on mastering the topic. 3. Common Mistake: Identify one key error that hinders success. 4. Comparison: Analyze two methods to highlight effectiveness. 5. Question: Pose a thought-provoking query related to achieving the topic. 6. Negative: Warn against a specific action that leads to failure. 7. Tutorial: Provide actionable steps to achieve the topic in a structured format. Prohibitions (P): Refrain from using overly complex language or jargon; avoid vague statements that do not directly relate to the topic. You (Y): Ask me all the questions you need to complete this task, one at a time. If you want more prompts like this for LinkedIn, video, email, and beyond, grab your copy of Prompt Writing Made Easy at promptwritingmadeeasy.com #LinkedInTips #PromptWriting #VideoContent #askSSL #AIforSales #sslinsights #SocialSelling .

One of the things we often talk about in class at UW Professional & Continuing Education is how uncomfortable it can be to use Gen AI and achieve in seconds, things that might normally take hours. Video production is one of those things and with a combination of #copilot #synthesia and #clipchamp, I can create training videos (four of them recently in fact) in less than two hours. Here’s the workflow: 1. Having recorded an SME on a 30-minute product demo call, I asked copilot for the transcript and, importantly the main points from the demo. This last part is an important early indication of whether the demo made sense to something with no context). 2. All being well, I then ask copilot to create a transcript for a short demonstration script for use in a video. The “trick” here is to ask for chapters within the demo/script so that you can more easily break the content and processing time down into much smaller chunks. 3. With a few tweaks of the script both in terms of spelling and a little bit of emphasis and clarification, and from one 30-minute call, I have four 90s second product demo scripts. 4. Using synthesia, I paste my scripts into the tool, having selected the voice and tone I want, and in under 10 minutes (often much less), I have a human-realistic , AI-generated audio file of my demo. note: synthesia’s huge benefit is that is has some incredibly realistic avatars, some of whose expressions alter based on the emphasis in the script - and many folks have had to be told that these are not real people. Well worth a look. 5. Finally, I open the audio clip in clipchamp and simply take my product screens and load them into the timeline, add some simple zoom animation, then export. Or I may have some product demonstration videos video from my prior Teams call I can quickly drop in. All that to say, and to go back to my prior point, what would have typically taken days barely took a couple of hours, and with only one person needed - all because Generative AI did such important grunt work. Upskilling and leveling the playing field for people who want to easily add value to their role or clients should be experimenting daily - think of what you can create, how much time you can save, or new recommendations you could make. #ai #copilot #genai #generativeAI

I've been learning how to make demo videos for Relay.app and a bunch of people have asked me about my tools and workflow so I thought I'd share. Disclaimer: I am a complete noob and I had never recorded or edited a video before in my life! So this will be really basic and ideal for beginners. Here's my full workflow: 1. I record the front-facing camera and screen capture using Screen Studio (essentially a more polished Loom). A few tips: - Use the "area" recording to get just the right part of the screen. - Do some editing on the auto-zooms. I find that they tend to zoom in too much (1.4x typically looks best to me). - Make sure to hit the "hide cursor when not moving" setting. - I like to hide the camera preview when recording because I find it distracting. - Export both the full video and the track of just the front facing camera; you'll need both later. 2. Once I have the export from Screen Studio, I open it in Davinci Resolve to do a few things: - Select which parts of the video should just have the camera on me and and which should focus on the screen recording. - Cut any long pauses or misspoken words. - Make sure the audio is smooth with no moments that are too loud or too soft. - When done, export the full video in Youtube mode on highest quality. Also output the audio separately for the next part. 3. From there, I run the audio through a Relay.app workflow that transcribes it and passes it through a few prompts to automatically generate a title, description, and summary for Youtube. We're working on expanding this to also generate a full accompanying blog post with screen shots pulled out, but haven't cracked that yet :). 4. We have a thumbnail template in Figma that I use to create a thumbnail with the right title and a few choices of imagery, along with any product logos of integrations we're highlighting. 5. Upload to YouTube and if appropriate send out an email to our newsletter or post it in our Slack community. For a 3-5 minute video, the process currently takes about an hour total, but I think we'll be able to get it down to ~30 minutes with a combination of getting better at the tools and doing a bit more with AI. Here's an example from today if you have any feedback on the final product: https://lnkd.in/gfNNN3T9