How I built my video content kit, and how you can build your own

Why I built it: what my clients were struggling with

I have photographed Auckland business owners for years, and they kept telling me the same thing. The hardest part of marketing is making the content. They knew they needed to show up regularly, with videos and posts that keep their business visible and share what their clients need to know. But it took hours they did not have, so it kept sliding to the bottom of the list.

• Time. Making a video the usual way means writing, filming and editing. With the kit, you talk into your phone for a few minutes, and the kit does the editing and design.
• Being on camera. Many people put off video because they do not want to be filmed. A voice memo is enough, and nothing has to be filmed.
• Editing and design. Most business owners are not video editors or designers. The kit edits what you said and puts it in your brand colours, fonts and graphics for you.
• Technical know-how. It was made for people who do not have a high level of technical knowledge. Claude, the AI tool made by Anthropic that the kit runs on, guides you one step at a time in plain language, and you can talk to it instead of typing.

People often ask whether they could build something like it themselves. You can. If you have some time and enjoy trying new things, the tools are open to anyone, and most of them are free. Below is what is in my kit and the stages I went through, so you have a map to follow. If you want every step in detail, or would rather skip the building, there are options at the end.

What is in the kit

Before the stages, here is what the finished kit can do. Everything below is built to work from a single voice memo on your phone:

• Videos in portrait for phones, landscape for YouTube and websites, or both.
• Sound clean-up that takes out low rumble, evens out a boxy tone and sets a steady volume.
• Captions timed to your words.
• Moving graphics in your brand, made fresh for each video in your colours and fonts, so there are no templates to buy or design fees to pay.
• Your own photos, footage and music, if you have some, worked in around your words.
• Sound effects made from scratch by the kit itself, with no downloads, so you own them outright.
• A cover picture and the post text for each video.
• Carousels, single graphic posts and Google Business Profile updates, from the same memo.

A voice memo is all you need. If you feel like it, you can also use a short video clip or selfies from your phone.

The editing, captions and graphics all run on free tools on your own computer. The only ongoing cost is your Claude Pro plan.

1. Decide exactly what goes in and what comes out

Before anything else, I wrote down what the kit had to do. In: a voice memo or a quick phone video, in whatever format a phone produces. Out: a finished video in portrait for phones, landscape for YouTube and websites, or both. The same memo can also become a carousel, a single graphic post or a Google Business Profile update. Every later decision came back to this list.

2. Choose the tools

Claude follows the instructions and talks the person through each step. The video and sound work is done by free, open tools: FFmpeg for video and sound, and Whisper for timing the captions to your words. Choosing the tools is quick. Getting them to work together smoothly is where most of the time goes, so be patient with yourself here.

3. Set up each brand

Every client's colours, fonts and logo go into the kit once, so every video comes out in their brand without them choosing anything. This includes finding and installing the right fonts, and making sure a logo looks right on both light and dark backgrounds.

4. Clean up the sound

Phone recordings vary a lot. The kit measures each memo, takes out low rumble, evens out a boxy or hollow tone and sets a steady volume. Getting this right took many rounds of testing on real memos. In one early version, the clean-up ran too many times and left the voice thin and muffled. It now makes one measured pass and checks it has not made things worse. It also listens for a phone notification ding, so a chosen line never runs across one. Expect your own first version to need a few rounds too. That is normal, and it is how you learn.

5. Edit the words

The memo is turned into text, then the kit picks the lines to keep, in an order a stranger can follow. It cuts the ums, restarts and repeats, and ends each line where the speaker really paused, so the edit sounds natural and every word is still theirs.

6. Design the layouts

Word cards, the client's photos, their own footage, captions and music all have to work together. Each social platform covers part of the screen with its own buttons, so text has to stay inside a safe area for each one. And words must never sit across someone's face.

7. Build in automatic checks

Before anyone sees a draft, the kit checks its own work: the sound level, whether text lands on a face, whether anything sits outside the safe area. A person should never be the first one to spot a problem the kit could have caught.

8. Write the instructions Claude follows

This is the biggest job, and the one that makes a kit easy to use. The instructions are written in plain language, for someone who has never used a tool like this. They cover every step, every question a beginner might ask, and what to say when something goes wrong. They now run to more than 1,000 lines, and nearly every line is there because a real test showed it was needed.

9. Make it set up on someone else's computer

Something that works on my computer is not a kit. It has to install on a client's Mac or Windows computer, with whatever they already have on it, guided one step at a time.

10. Test, break, fix, repeat

Every stage above was redone more than once. The kit now runs a self-test of nearly 90 checks, and it is tested with real voice memos and real people, not just my own recordings. When something improves, I update it, which is why clients can choose to have their kit updated later.

Ready to build your own?

If you have the time, I encourage you to give it a go. This technology is open to anyone, and building something for your own business is one of the best ways to learn what it can do. Start small, with one voice memo and one video, and build from there. These are the main ingredients you will need:

• An AI assistant that can follow written instructions and run tools on your computer. Mine runs on Claude.
• A speech-to-text tool to turn your recording into words and time the captions. I use Whisper, which is free.
• A video and sound tool to cut, clean up and put everything together. I use FFmpeg, also free.
• A way to draw graphics with code. In my kit, small programs draw each card in your brand colours and fonts, so nothing has to be designed by hand.
• A brand file holding your colours, fonts and logo, so every piece comes out looking like you.
• Written instructions for the AI, in plain language, covering each step and what to do when something goes wrong.
• Automatic checks that catch problems before you see a draft.

Three ways to buy the kit

• Build your own with the step-by-step guide, NZD $150. A PDF guide for people comfortable with technical set-up. Roughly 10 to 20 hours. See the guide.
• The ready-made kit, NZD $250. The finished kit as a download, which walks you through set-up on your own laptop. About an hour to set up and make your first video. Designed for beginner to medium technical skills. See the kit.
• The kit with a lesson, NZD $795. I set it up for your brand, and we spend two hours together in person in Auckland while you make your first real video. You also get a how-to video to keep and a 30-minute check-in afterwards. See what is included.

All three run on your own Claude Pro plan, about NZD $40 a month. Once you are comfortable with the kit, it also works with a Voice Twin or an AI video digital twin later on, if you want them. I have also written about why you do not need a voiceover for your business videos.