If you’ve been around Local Service Spotlight for any length of time, you’ve heard us preach that you should talk to the AI instead of just typing at it. The reason is simple — when you speak, you give way more context than you ever would by typing. Making videos works the exact same way, just at a bigger scale. You’re taking a thought, a task, or a lesson out of your head and putting it somewhere the agents (and other humans) can actually learn from it.
That’s what this post is about. I’m Dylan Haugen, and I want to walk you through why I think everyone working in this space — clients, teammates, contractors, even people just getting started — should be putting their thoughts on video.
Speaking Gives More Context Than Typing — Same Reason We Talk to the AI
The same principle that makes a voice prompt better than a typed prompt is what makes a video better than a written note. When I sit down to type something out, I usually trim it down to what feels “important.” When I talk, I cover the why, the side-tangents, the gotchas, and the little human details that actually make the explanation click for someone else. All of that extra context becomes fuel — for the agents, for the team, and for me when I come back to it later.
Throw that video on YouTube and it transcribes itself. Take the transcript and turn it into an article (this one, for example), and now you’ve taught the agents too. That’s exactly the same loop we run with meta articles — the agent finishes a task, then writes up what it did so the next agent (or the next human) can learn from it.
A Real Example: Paul Ryazanov’s Knowledge Panel Push
I just made a video about one of our clients, Paul Ryazanov, who we’re helping earn his Google Knowledge Panel. We pointed agents at his existing content — YouTube videos, LinkedIn posts — to inventory it, and then repurposed those pieces into articles on his site. That activity raises his confidence score with Google and pushes him closer to earning a panel.
When the agent finished its part, I had it write a meta article that walked through the process — what it did, what it pulled, where it got stuck. Then I went back personally and shot a video walking people through the same workflow from my perspective. Why both? Because some people are visual learners, some like to hear it, some just want to read it. We try to hit every type of learner with the same body of work, and the video usually catches nuance the agent missed during QA, which makes the final asset stronger than either piece alone.
This Is LDT — Learn, Do, Teach
One of Dennis’s frameworks — and one of the nine triangles — is LDT: Learn, Do, Teach. Making videos is the “Teach” step, and it’s the one most people skip.
- Learn — I’m learning the task as the agent learns it alongside me (inventory the videos, repurpose them, push them to the site).
- Do — I actually do the work. The articles get written, the links get added, the panel signals start stacking.
- Teach — I take what I just did and put it into words and video so the next person (and the next agent) doesn’t have to figure it out from scratch.
The Teach step is the one we only want actual practitioners doing. You can’t teach a workflow you haven’t run. But once you have, putting it on video is the highest-leverage thing you can do with that knowledge.
Why Video Is the Raw Material for Content → Checklist → Software
This is the part that ties everything together. Another of the nine triangles is CCS — Content, Checklist, Software. You can’t jump straight to “software.” You have to walk it up the ladder.
- Content — me making the video about Paul’s repurposing process, plus the meta article the agent wrote when it finished the task. That’s the raw ingredient.
- Checklist — we take those raw assets and turn them into a checklist for the task itself. “Bold the column headers. Drop the published URL back into the sheet after each article. Tag the post with the right category.” Every step the video covered and every step the agent documented becomes a checkbox.
- Software — once you have the content and the checklist, you can automate. Claude skills are a great example — once a workflow is captured cleanly, the agent can run it on repeat.
If you skip the content step, your checklist is incomplete. If your checklist is incomplete, your automation breaks. The video is the cheapest, fastest way to capture the content layer with full context, including all the little decisions you make that you’d never think to write down.
The Camera Is the Hard Part — But It’s Worth It
I get it. A lot of people don’t want to get in front of a camera. They don’t want to hear their own voice played back. I’ve been there, and honestly I’m still guilty of not doing this enough myself — I’m trying to get back into it more consistently. But personally, I convey my thoughts a lot better when I’m speaking them out loud than when I’m staring at a cursor. A lot of people are the same way once they actually try it.
And in this space specifically — where the goal is to keep climbing toward the software layer of CCS — you have to have those raw ingredients. You yourself completing the task. The agents documenting alongside you. The two combined into a checklist, then into automation. The video is what kicks the whole loop off.
Your Turn
So — are you making videos? Are you teaching what you do, or are you just doing it and keeping the lessons in your head? I’m genuinely curious. Drop a comment, tag me, or better yet, make a video about it and send me the link. Catch you in the next one.