MontikThe advice below is how we set up Montik’s own variant workflow. Treat it as a starting method, because your account’s results should overrule any rule of thumb, ours included.
Finish the master first
Every flaw in the master gets copied into each variant, so finish it before you branch. Get it approved, then save a version. Find its parts too: the hook in the first few seconds, the body, the proof and the call to action.
Make hooks that differ in kind
Three rewordings of one idea only test wording. Try different kinds of hook instead:
- The result first: show the outcome, the number or the after shot.
- A bold claim: say something your audience might disagree with.
- The viewer’s question: “Why does my coffee taste burnt?”
- A direct call-out: “If you run a small shop…”
- A pattern break: an unexpected shot, movement or number.
You often already have the footage. The strongest line in the middle of a testimonial can move to the front.
Build each one
Keep every other part of the ad the same. Change one thing per variant, and name it after that change, such as “Hook B: question” or “Hook C: result first”. For each hook:
- Bring the line or shot to the front, then close the gap it leaves.
- Put text on screen almost immediately. Three to seven words, readable with the sound off.
- Skip the fade-in. The first frame also works as a thumbnail, so show the face or the product.
- Cut within two to three seconds. A long first shot gives people time to leave.
- Regenerate subtitles if the timing changed.
- Check text placement. On 9:16, keep it clear of TikTok’s buttons, roughly the top 14% and bottom 35% of the frame.
In Montik
Variants are separate projects made from the approved cut, each with its own hook and cover. You can make them by hand, or ask Claude:
Make three hook variants of this ad: one that opens on the result, one with the viewer’s question, and one with a bold claim. Keep everything after the hook the same, then check each one and tell me what changed.
Read every new line before it ships. You’re responsible for what an ad claims, whoever wrote it.
Montik’s edit check flags a slow first shot and text under the platform’s buttons. You can then export all the variants together, each to its own file.
Run a fair test
- Test the hook alone. If the length or offer also changes, you won’t know what made the difference.
- Run them together, with the same audience and budget split, so timing doesn’t skew the result.
- Judge the opening by the opening. Early retention tells you how the hook did. Conversions depend on much more, so look at both.
- Give it enough spend before calling a winner. A few hundred views can mislead you.
Then go shorter
When a hook wins, cut it down. A common ladder is 30, 15 and 6 seconds. Keep the hook and call to action, and take from the middle. Then try other formats: 1:1 and 4:5 for feeds, 16:9 for YouTube. Montik refits the edit to a new frame, but check each cut, because reframing can crop the subject.
