Video Scaling Strategies · Issue 3

How Gong Won With the “Worse” Hook

What happens to content strategy when making the clip stops being the hard part.

In short

Gong published three cuts of one CEO interview on the same afternoon. The one with the messiest opening line did 984 views; the polished, contrarian one did 133. The winner opened on a full-frame Economist cover rather than on a face, so it was answering an accusation instead of making a claim. When AI makes producing the clip free, the remaining work is choosing which clip and what the first second is pointed at.

Hi, I'm Lera. Every issue of Video Scaling Strategies takes one piece of content that worked and one that didn't, and uses what the market is doing to explain the gap.

Opening frame one: a full-screen Economist cover headlined Fear of the SaaSpocalypse is tormenting techland. No face on screen. Opening frame two: a wide shot of the CEO, arms crossed, captioned Focus on value, not just features. Opening frame three: a close-up of the CEO mid-sentence, no text on screen.

The first second of each. Same CEO, same interview, same afternoon, all under a minute.

I've been picking at these all week. Three cuts of one conversation, posted together — and one of them ran away from the other two.

Bar chart of views on each short. A magazine cover, 984 views. A face plus caption, 193. A face, 133. The magazine-cover clip is the only one that opened on someone else's headline, and the only one that travelled.
984 views against 133. Same speaker, same interview, same afternoon.

The Worst Hook Won

Here's what bugs me. Rank them by the quality of the talking and the order flips completely.

The one that finished last opens with a properly contrarian line — “You can't start a category” — clean, complete, the kind of thing I'd screenshot. It got zero likes.

The one that won opens like this:

There's some truth in that, but it's like a way like overreaction of the market right now.

The winning clip · first words

Mid-thought, answering a question you can't hear, filler everywhere. I would have cut it.

It isn't captions either, before you ask — the winner has none and neither does the loser. What the winner did was hand its opening to somebody else: a full-frame Economist cover, “Fear of the SaaSpocalypse is tormenting techland.” Only then does the CEO turn up to answer it. It isn't making a claim. It's answering an accusation.

Why I Think This Is the Whole Game Now

For two years the instruction to every B2B team was make more. Volume was a moat because volume was hard: an editor, a week, a budget.

That's finished. A model will find the fifteen moments in your footage, caption them, reframe them and schedule them before lunch — and it will do the same for everyone you compete with, on the same afternoon. When production collapses to nothing, the constraint moves. It moved to which one, and what do you open it with.

The feed stopped covering for you, too. Gong has 6,930 subscribers and 308 videos on that channel, and it can still post something that barely clears a hundred views. Shorts and Reels are recommendation surfaces, not subscription ones. Nothing reaches anyone because they once followed you.

Your follower count is not a floor. It's a rounding error.

Which is the same story as everything else in AI right now, honestly. The capability gets cheap. The judgment about where to point it doesn't. Content just got there first, because the scoreboard is public and it settles in a day.

What I'd Change on Monday

Comparison. The cut most teams ship: opens on a face, nothing to read on mute, leads with your claim, starts at the answer, judged with the sound on. The cut that travels: opens on something readable, leads with their fight, starts at the tension, judged on mute.

The entire right-hand column is one move: stop opening with here is our view and start opening with here is your problem.

Three takeaways. Making the clip is free now. So the whole game is the first three seconds. Spend them on their fight, not your face.

That's the memo. Go watch your last five clips on mute and stop them at three seconds — if you can't tell what they're about, neither could anyone else.

Workshop · held Tue Aug 11, 2026

Cut the Tool Clutter: The 1M+ Creator Podcast Blueprint

Tuesday, August 11, 2026 · 12:00–1:00 PM EDT · Virtual, on Luma

If making the clip is free, the only work left is choosing it and opening it right. That's the hour — a live panel and Q&A on collapsing a four-tool podcast workflow into one engine that records once and publishes everywhere.

  • Consolidate recording, editing, transcription and scheduling into one stack
  • Turn one 30-minute interview into short-form clips, LinkedIn posts, show notes and a newsletter
  • Keep publishing consistent with AI agents and templates instead of constant manual oversight
  • Live demo of an end-to-end B2B podcast workflow, raw recording to multi-channel distribution

For marketing leaders, founders and content managers. Spots are capped to keep it interactive.

Book office hours

Written by Lera Leonteva. Sources: Gong (@Gongio) on YouTube, three Shorts published July 1, 2026; view, like and comment counts, subscriber and video totals read July 13, 2026. The Economist, “Fear of the SaaSpocalypse is tormenting techland” (June 10, 2026). Opening lines transcribed from the clips.

Frequently Asked Questions

What people ask after seeing the three clips side by side.

Why did the clip with the worse opening line get more views?

Because the opening line was not what the viewer met first. The winning clip handed its opening to a full-frame Economist cover — “Fear of the SaaSpocalypse is tormenting techland” — and only then brought the CEO in to respond. It was answering an accusation rather than making a claim. The clip that finished last opened on a face with a clean contrarian line and got zero likes.

Was it the captions?

No. The winner has no captions and neither does the loser. The only structural difference is what fills the first second.

Does a big subscriber count protect your reach on Shorts?

No. Gong has 6,930 subscribers and 308 videos on that channel and can still post something that barely clears a hundred views. Shorts and Reels are recommendation surfaces, not subscription ones — nothing reaches anyone because they once followed you. Your follower count is not a floor.

If AI can produce unlimited clips, what is the actual constraint?

Choice. A model will find the fifteen moments in your footage, caption them, reframe them and schedule them before lunch, and it will do that for your competitors on the same afternoon. When production collapses to nothing, the constraint moves to which clip you ship and what you open it with.

How do you test whether your own hook works?

Watch your last five clips on mute and stop each one at three seconds. If you cannot tell what it is about, neither could anyone else. The move is to stop opening with your view and start opening with the viewer's problem.

Lera Leonteva
Lera Leonteva
CEO & Founder, Leo AI

Lera Leonteva is the founder and CEO of Leo AI, where she works on turning long-form footage into short-form video that earns attention on its own merits. She writes Video Scaling Strategies, taking one piece of content that worked and one that did not, and using what the market is doing to explain the gap.

Have a video?

Leo AI edits it into scheduled social posts, guided by what your audience actually watches.

See this working on your own channels

Leo turns your existing footage into posts and shows what each one earned. Book a demo or see the channels Leo posts to.