How a Baltimore roofing-data startup ships 50 clips a month for $15.
Roof Radar is my other product — a small B2B that tracks storm and permit data for the roofing industry. I needed weekly marketing video to grow it, and the SaaS quotes were $300+/mo. Hybrig replaced that. 50 clips a month run on my home rigs. The marketing budget went from a line item to a rounding error.
Who, what, how much
- Who
- Small-business owner / founder running Roof Radar (me — James).
- What
- Marketing content for Roof Radar — storm-damage explainer videos, before/after roof shots, FAQ shorts, paid-ad creatives, weekly social drops.
- How much
- ~50 clips/mo. SaaS equivalent: ~$300/mo (Synthesia Creator at cap + overage, or HeyGen mid-tier). Hybrig: ~$15/mo electricity across two rigs. Net: ~$285/mo saved = ~$3,420/yr.
The workflow in plain words
Step 01
I generate the source images on Flux locally — storm shots, roof close-ups, product mockups.
Step 02
I batch-queue 10–30 clips at once via /tools/batch-image-to-video. They render across both my GPUs while I’m on sales calls.
Step 03
I download a folder of MP4s, hand them to my editor (or post them direct to social if it’s a quick drop).
What it looks like
Six recent clips. 16:9 widescreen — the format we ship paid ads and explainer content in.
Clip 01 · storm-damage explainer
Clip 02 · before/after roof
Clip 03 · FAQ short
Clip 04 · paid-ad creative
Clip 05 · weekly social drop
Clip 06 · product mockup
“My old plan was paying $300/mo for Synthesia and rationing renders. Now I crank out a clip every time we get a new storm event. The marketing budget went from a line item to a rounding error.”
— James, Roof Radar
By the numbers
Numbers are real-world cadence, not a projection. Verify against your own usage before you quote them.
The thing that isn’t about money
Roof Radar’s competitors are all paying SaaS bills. We’re not. Every marketing dollar we save is a dollar we put into product development or paid ads. The compounding gap is the moat.
Get started
Just one face on camera? Read the solo-creator case study →