I’ve seen thousands of pitches. As an investor in over 200 companies, including AI leaders like Anthropic, OpenAI, and Scale AI, I’ve learned to spot the difference between a gimmick and a genuine breakthrough. And right now, the world of AI video is full of gimmicks.
Most of what I see online looks the same. Float-y, generic, slightly weird videos that scream “I just typed a prompt into an AI.” It’s the same trap people fell into with early image generation tools. They become prompt-jockeys, endlessly tweaking words without a real strategy, and the results are predictably mediocre.
This isn’t another basic tutorial. This is my strategic guide to using RunwayML to get massive results. I’m not just talking about making a cool-looking video; I’m talking about creating work that drives real-world outcomes. I’m sharing the exact workflow I use to create visuals that stand out, get attention, and actually achieve a goal.
Stop Being a Prompt-Jockey
The biggest mistake I see people make with tools like RunwayML or even Midjourney is focusing 99% of their effort on the prompt. They’ll spend hours trying to find the perfect combination of words, thinking that’s the key to unlocking some hidden magic. It’s not.
When we were building RemoteTeam (which was later acquired by Gusto), we didn’t just describe a product. We had to build a vision for a new way of working. It required a story, a clear sequence, and a deep understanding of the problem we were solving. The same is true for creating powerful AI-generated content. You need to be a director, not just a writer typing commands.
Your job is to have a strong opinion and a clear vision. The AI is your crew. It’s your cinematographer, your actor, and your editor. But you’re the one in charge. You need to tell it what to do.
My “Director’s Workflow” for RunwayML
This workflow moves beyond simple text-to-video and treats the process like a real film production. It’s how you get from a vague idea to a polished, impactful final product.
1. The Concept & Storyboard (The Director’s Vision)
Before I even think about opening an AI tool, I know what I want to create. I have a story in mind. It doesn’t have to be complicated. For a marketing video, it might be:
- Shot 1: Show a frustrated user dealing with a common problem.
- Shot 2: Introduce the solution with a clear, heroic visual.
- Shot 3: Show the user happy and successful, using the new solution.
I’ll even sketch this out on a piece of paper. A few crude boxes are all you need. This simple step is what separates professional results from amateur ones. It forces you to think about composition, pacing, and emotion before you get lost in the tech.
2. Asset Generation (Midjourney as my Cinematographer)
I don’t use RunwayML to generate my initial scenes from scratch. The quality just isn’t there yet compared to the top image models. My go-to is Midjourney. I use it to create the key frames for my storyboard. These are my “master shots.”
My prompts for this stage are very specific and focused on cinematic quality. I’m not just saying “a person using a computer.” I’m thinking like a cinematographer:
- “cinematic shot, a young woman in a dimly lit office, frustrated, head in her hands, glowing monitor screen illuminates her face, shallow depth of field, 35mm film grain”
I’m giving the AI instructions on lighting, lens, and mood. I’ll generate dozens of options until I get the perfect image for each beat of my storyboard. This is the foundation. High-quality inputs are the first step to high-quality outputs.
3. The RunwayML “Editing Room”
Once I have my master shots from Midjourney, I bring them into RunwayML. This is where the magic happens. This is where we bring the still world to life.
Gen-2 is Your Actor: I use the Image-to-Video feature (Gen-2) to animate my still images. The key here is to be subtle. You’re not trying to create a blockbuster action scene. You’re trying to add a spark of life. For the frustrated user, I might just want her to sigh or slowly look up at the screen. A simple prompt like “a woman sighing in frustration” applied to the image is often enough.
Motion Brush is for Nuance: This is one of the most powerful and underutilized tools. Instead of letting the AI guess what should move, Motion Brush lets you paint the motion directly onto the image. For a shot of a landscape, I can paint the sky to make the clouds move, or paint the trees to make them sway in the wind. This control is what creates a believable, non-generic result. You can direct the viewer’s eye and add a layer of polish that most AI videos lack.
Camera Controls are Your Dolly Grip: A static shot is boring. I almost always add some subtle camera movement. A slow push-in can build tension. A gentle pan can reveal new information. Again, subtlety is key. A 1-2% zoom over 4 seconds is often all you need to transform a static image into a living scene.
4. Sound and Post-Production (The Final Polish)
No video is complete without sound. I can’t stress this enough. A silent video feels like a demo. A video with sound feels like a story.
I’ll export my clips from RunwayML and bring them into a simple video editor like CapCut or Adobe Premiere. I’ll add:
- Music: To set the mood.
- Sound Effects: A keyboard typing, a gentle whoosh for a transition, a sigh.
- Color Grading: A final pass to make sure all the clips feel like they belong in the same world.
This final 10% of the work makes 90% of the difference.
Why This Matters
Is this more work than just typing a prompt into a single tool? Yes. But the results are in a completely different league. When you’re pitching your startup, trying to build a brand, or launching a new product, you can’t afford to look generic. You can’t afford to look like everyone else.
I’ve had two successful exits and invested in over 200 companies. I can tell you that the people who succeed are the ones who have a clear vision and execute it with relentless attention to detail. They don’t just follow the instructions; they build a framework that works.
This workflow is about taking control. It’s about moving from being a passive user of AI to being a director. Stop just asking the AI what it can do for you. Start telling it what you want to create. That’s how you’ll get results you won’t find in any course.
Frequently Asked Questions
What are the most common mistakes when using runwayml to create results you won't find in any course?
The biggest mistake I see is overcomplicating things early on. Start with the simplest version that works, get real feedback, and iterate from there. Another common trap is copying what worked for someone else without understanding the context behind their decisions.
Do I need technical skills to use runwayml to create results you won't find in any course?
Not necessarily. While technical understanding helps, the most important skills are clear thinking and the ability to break problems into smaller pieces. Many successful founders I've invested in started with zero technical background and either learned enough to be dangerous or found the right technical partner.
How do I measure success with this approach?
Pick one or two metrics that directly tie to your goal and track them weekly. Vanity metrics like page views or follower counts rarely matter. Focus on metrics that reflect real engagement or revenue impact.
How long does it take to use runwayml to create results you won't find in any course?
The timeline varies depending on your starting point and resources. For most founders, expect 2-4 weeks for initial setup and 2-3 months to see meaningful results. I've seen teams move faster when they focus on one thing at a time rather than trying to do everything at once.