Talking head videos are probably the most requested format I get asked to create and honestly, I get it. They are versatile, they feel personal, and when they are done right they are absolutely brilliant at converting viewers into customers. But there is a massive difference between a talking head video that lands and one that gets three seconds of watch time before someone scrolls on.
I am going to walk you through exactly how I approach these. No waffle, just what actually works.
It All Starts With the Hook
The first two seconds of any talking head video are everything. If your opening line does not immediately make someone think "I need to keep watching this," then it does not matter how good the rest of it is. Nobody is going to see it.
The hooks I find work best tend to fall into a few categories. There is the bold statement: "This product completely changed my skin routine and I was not expecting it." There is the question: "Why is nobody in Dubai talking about this?" There is the relatable frustration: "I have spent hundreds on gym gear that looked rubbish by month two." And there is the curiosity gap: "I tried this so you do not have to, and I have thoughts."
The key is that the hook has to feel genuine. Forced hooks sound forced and people can tell immediately.
My Setup for Dubai Shoots
People always assume you need loads of kit to make great talking head content. You really do not. Here is what I actually use most of the time:
- My phone on a decent tripod, shot at eye level or very slightly above
- Natural light from a window, positioned at roughly 45 degrees to my face
- A ring light as backup when we are shooting later in the day or in a darker location
- A clean, uncluttered background or one that is relevant to the brand
- A lapel mic for clean audio, because bad audio is a death sentence for watch time
In Dubai, the light is genuinely incredible for most of the year. Early morning shoots before around ten give you this gorgeous warm, soft light that looks beautiful on camera. Afternoons in summer are trickier but with the right indoor setup you can still get brilliant results.
The Script vs. Freestyle Debate
Some creators script everything word for word. Some completely freestyle. I sit somewhere in between, which I think produces the most natural results.
I will have a clear structure: the hook, the problem I am addressing, the product as the solution, a couple of specific benefits with genuine reactions, and a call to action. But within that structure I want to sound like I am actually talking, not reciting something. That means allowing for natural hesitations, genuine reactions, and language that does not sound like it came out of a marketing brief.
The biggest mistake brands make when briefing talking head content is being too prescriptive about the exact words. Give me the key messages you want to land and let me find a way to say them that actually sounds human.
Pacing and Editing
Talking head videos need to move. In 2026, audiences have incredibly short attention spans and your edit needs to reflect that. I typically remove any pause longer than half a second, tighten up any moments where the energy dips, and add jump cuts to create pace. Captions are non-negotiable because a significant chunk of people watch with sound off.
For length, I find 30 to 45 seconds is the sweet spot for most brand content. Long enough to land the message properly, short enough to hold attention all the way through. Some formats like tutorials or more detailed reviews can go longer, but the rule of thumb is always: cut anything that does not earn its place.
What Brands Should Brief Me On
To get the most out of a talking head video, I need a few things from you. The key message you want to land, the platform it is going to run on, who your target audience is, any specific claims you want to make (or avoid), and the tone you are going for. That is genuinely all I need to come back with something that works.
If you have a product or campaign and you are wondering whether a talking head format would be right for it, the answer is almost certainly yes. Drop me a message and let's figure it out together.