Okay, so I've been deep in the weeds with Opus Clip for the last month, mostly using it to repurpose our longer Salesforce and RevOps tutorial videos into social clips. The algorithm is solid, but I kept wondering: does the actual *canvas* or aspect ratio of the clip make a difference in performance?
I decided to run a simple A/B test. I took the same core 45-second clip generated by Opus—one of me explaining a common data hygiene pitfall during migrations—and rendered it out in two formats Opus provides: the classic square (1:1) and the vertical 9:16 (full portrait/TikTok/Reels style).
My hypothesis was that 9:16 would crush it, given how native it feels to short-form platforms. The results were... not that simple.
I pushed the square clip to LinkedIn and Twitter/X, and the 9:16 to TikTok, Instagram Reels, and YouTube Shorts. Tracked everything for a week.
Here's what stood out:
- **Completion Rates:** The 9:16 version had a noticeably higher average watch time and completion rate on TikTok and Reels. No surprise there; it fills the screen perfectly.
- **Engagement (Likes/Comments):** This is where it got interesting. The square clip on LinkedIn generated significantly more comments and meaningful questions. My theory is that for a slightly more "professional" or detailed topic, the square format, which often feels more like a "video post" than a "full-screen experience," prompted more thoughtful engagement. People weren't just passively watching; they were pausing to read the text overlays and react.
- **Sharing:** Almost a dead heat. The 9:16 got more casual shares ("this is cool"), while the square got more "this is useful" shares with tags.
So, my takeaway for the community is this: the "best" format might depend less on the platform algorithm and more on your content's intent and your audience's mindset. For quick, eye-catching hooks and entertainment, 9:16 is king. But for community-driven platforms where you want to spark conversation on a technical topic, the humble square might still have a real edge.
Has anyone else done a similar style test? I'm curious if your findings match mine, especially for more niche, SaaS-oriented content.