A title company showed me something cool. That was the problem.

Here's what they built. You record your voice once. The system clones it. Then it pulls the photos and the description for any listing straight from the MLS and, in about a minute, spits out a finished video: your listing photos panning across the screen while your cloned voice reads the property description aloud. No filming. No editing. A whole narrated listing video, generated while you get coffee.

And I'll be honest — the first time I saw it, I thought it was genuinely impressive. Your own voice, on a video you never recorded, describing a house you never filmed. That's a real piece of technology.

Then I watched it back. And what I was actually looking at was a slideshow of photos with a voice reading an MLS description over the top. Which is to say: I was watching 1980s cable real estate. The narrated-listing infomercial that used to run at 3 a.m., rebuilt with AI and posted to Reels.

It's cool. It also won't work. And the gap between those two things is the most expensive mistake in our industry right now.

The tool solved a problem. Just not the one that matters.

It answered "how do we make a listing video without filming one?" But production was never the bottleneck. Attention is. Making it free and instant to produce a video nobody watches is not a win — it just lets you fail faster and at higher volume.

And here's the part that turns a bad idea into an expensive one: the platforms are now working against exactly this kind of content. Through 2025 and into 2026, both Instagram and TikTok have been openly down-ranking mass-produced, unoriginal, and obviously AI-generated video. Instagram's own leadership has said they're prioritizing original content and reducing the reach of the reposted, auto-generated stuff. So the auto-narrated slideshow doesn't just bore the viewer — it gets quietly buried by the algorithm before most people ever see it. The tool fights you twice.

None of this means AI has no place in your video. It means this particular use aimed it at the wrong target — it automated away the two things that actually make video work. To see what those are, I asked someone who does this for a living.

My daughter Madeline runs MediaProsAI, where she makes short-form video for agents and trains them to make their own. Ask her what matters most in a video and she doesn't hesitate: the hook. The first few seconds. Everything else is downstream of whether you survive them.

The data backs her up bluntly. A viewer decides whether to keep watching or keep scrolling in the first one to three seconds. The steepest drop-off on any short video is right at the start. If you don't earn the next second, there is no next second — and nothing you put later in the video matters, because nobody got there.

So Madeline trains agents on what not to do, and her number-one example is the exact way most agents open: "Hi, my name is Josh Marquez, and today I'm going to show you three things about…" By the time you've finished saying your own name, they're gone. You spent your most valuable three seconds on an introduction nobody asked for.

Now look back at the title company's tool. How does its video open? With a cloned voice reading "Welcome to 123 Main Street, a stunning four-bedroom…" It is built to open with the weakest possible hook. It can't lead with a face, a question, a surprise, or a reason to care — because it doesn't know your listing has a story. It only knows the MLS description. The format has no hook by design, which means it fails at the one thing Madeline says decides everything.

Here's the harder truth, and it's the one I hammer with my own agents.

Say you fix the hook. Say you learn to stop the scroll. You still haven't won, because engagement and income are two different things. I personally know people with big followings on TikTok and Instagram — real reach, thousands of views — whose production is almost zero. They're great at getting likes. They are not great at getting paid.

Likes don't pay the bills. Leads do. And leads don't come from a good hook alone — they come from a good hook attached to something genuinely useful to the person watching. Value is what converts a viewer into a phone call. A house tour narrated by a robot has no hook and no value. A viral clip with a great hook and nothing underneath gets you applause and an empty pipeline. The thing that actually works is the rarest of all: a real person, stopping the scroll, and then saying something worth hearing.

What's the best hook you've ever used — the opening line or first shot that actually stopped people? Hit reply and tell me. I read every response, and the good ones might show up in a future issue.

And if you want to get serious about the video side of this — hooks, short-form, actually making the stuff — that's exactly what my daughter Madeline does at MediaProsAI. Separate company from the school, same standard: make the thing that actually works, not just the thing that looks cool.