The first 3 seconds of a TikTok ad

Three seconds is roughly seventy-five frames, one short sentence, and about two visual beats. It is not much, and most product ads spend it on throat-clearing.

This is a frame-by-frame account of what to put in that window, and the four habits that waste it.

Second one: the viewer decides whether this is about them

The first second is not for your product. It is for the viewer’s recognition that the video concerns them.

Two things have to be true by the end of it.

Something is already moving. A static frame — a person about to speak, a product on a table, a title card — is a frame that gives the viewer nothing to process, and it is the easiest one in the world to leave. Start on an action: mid-pour, mid-unfold, mid-walk, hands already doing the thing.

The subject is legible at a glance. Whatever the video is about should be identifiable without the viewer parsing the frame. A wide shot of a kitchen is not legible. A close crop of a burnt pan is.

The most common repairable mistake here is starting too early. Almost every self-filmed product video has one to two seconds of set-up at the front — a breath, a reach, someone finding their eyeline. Delete it. Watch the video back starting at the first genuinely interesting frame, and it is nearly always better.

Second two: the claim lands

By the end of second two the viewer should have heard, or read, the one interesting thing.

That is about eight to ten words spoken at a normal pace. Not a build-up to a claim — the claim.

• Too slow: “So today I want to talk to you about something that’s been a bit of a problem in my kitchen…”
• Right: “Your knife block is what’s blunting your knives.”

Two rules for the line itself:

Front-load the surprising word. “Blunting” is the word doing the work, and it arrives at word five. If your interesting word arrives at word fifteen, the sentence is a paragraph.

No conditionals. “This could help if you’ve ever…” is three qualifications before a claim. Say the thing.

Second three: the reason to stay

Seconds one and two buy attention. Second three has to spend it on a promise — an implicit “and I’m about to show you”.

Usually that is one of:

• A visible unanswered question. Something is on screen that has not been explained yet: two products side by side, a cut-open version, a before shot with no after.
• A count. “There are three of these and only one is worth it.” The viewer now knows the shape of the video and how long it will take.
• A named payoff. “The bit that changed it is at the end” only works if the video is genuinely short. On a fifteen-second ad it is fine; on a ninety-second one it is a promise you will not be forgiven for.

What is on screen while all this happens

Text. Put the hook on screen as well as saying it — a large proportion of views start muted, and text lets the line land before sound is on. Keep it to one short line, high contrast, and out of the bottom third and right edge where the platform’s own interface sits. Native captions cover the same ground but you control the timing yourself.

Sound. A voice starting on frame one beats music starting on frame one. Music is atmosphere; a human voice is information. If the first thing a viewer hears is a beat drop, the first second told them nothing.

Cuts. One cut inside three seconds is plenty; two is the practical maximum. The point of an early cut is to prove the video goes somewhere, not to look busy.

Framing. The subject wants to be large. Phone screens forgive close-ups and punish wide shots. If you can see the whole room, the viewer can see nothing.

The four ways the window gets wasted

The greeting. “Hey guys, welcome back.” Nobody has come back. This is a convention borrowed from a platform where people subscribe to a person, applied to an ad where they have not.

The logo card. A branded intro on a product ad spends the entire attention budget on a name the viewer has no reason to remember yet. Brand at the end, if at all.

The slow reveal. Holding the product just out of frame to build suspense assumes attention you have not earned. Show it.

The pitch that starts with the company. “We spent two years developing…” Nobody has agreed to care about your process. Get to what it does for them, then earn the process story later.

A test you can run on any cut

Play the first three seconds with the sound off and ask someone who has not seen it: what is this about, and who is it for?

If they cannot answer both, the opening is not working, and no amount of improvement after second four will rescue it. That test costs nothing and is brutally reliable.

Then write more than one

The opening window is too small to iterate on by intuition. Write several different first lines for the same product — one per objection — and choose between them side by side rather than sequentially. Choosing is a different skill from writing, and it is where most of the gain actually is: testing hooks before you spend covers the cheap ways to do it.

Related

• How to write a hook for a product video — the method for producing the line
• Hook formulas that stop the scroll — the shapes that reliably fit the window
• TikTok hook ideas for product videos — thirty openers to adapt