Half a second into my Chapagetti video, the packet is already in my hands. By 2.1 seconds the noodles are up on chopsticks. That’s the job of the first 3 seconds of a food reel: tell a stranger what they’re looking at, and get to the good part before they leave. That means opening on the food, or the most striking thing about it, rather than the sign or the logo. One line on screen that works with the sound off. And the best shot early, not saved for the end. Below are the first 3 seconds of four reels I made, frame by frame, and the checklist I use before posting.
What is the 3-second rule for reels?
It’s shorthand for the first few seconds of a reel, when someone scrolling decides whether to stay. My working version: by the 3-second mark, a stranger should be able to guess what the video is about. If they can’t, the good footage later on doesn’t get seen.
So I treat the first frame and the first line of text as the trailer. If someone is watching with the sound off, the words on screen are the only preview they get. Whatever question the opening raises, I want the video to answer it on screen within about 7 to 15 seconds. And I don’t finish with a recap of what we learned.
What should the first shot of a food reel be?
The strongest shot you have of the food. Not the shopfront, not the sign, not the logo. If you only had one second, it’s the shot you’d use.
When I cut my hotteok video in May, I opened on the most striking moment I had: syrup going onto the hotteok. Then the crunch of the first bite, then a before and after, and only then the how-to.
Before any of that, the subject has to be strong enough. When a hook feels weak to me, I go back and look for a stronger subject: something people stop for without thinking, the way they stop for a cat or a dog.
What does a food reel opening look like, second by second?
Here are four reels I made, paused frame by frame. Each one has words on screen inside the first second, and each shows the food, the drink or the coffee being made before the 3-second mark.
Chapagetti, @waitwhatmj
The frames at the top of this page. In the first half second the packet swings into frame and into my hands. Within 2 seconds I’ve asked “Do you know Chapagetti?”, and the words come up on screen one at a time. At 2.1 seconds it cuts to the cooked noodles lifted on chopsticks, still steaming, and by 2.8 seconds I’m lifting them to eat.
A question, then the answer straight away. If I open with a question, it’s a real one, asked the way I’d ask it off camera, and the payoff comes right after it.
3.1M views, 252.9K likes and 64.9K saves on TikTok. More on my creator page.
The Art, FAVE Specialty Coffee
A latte art reel for FAVE Specialty Coffee in Fortitude Valley.
The camera looks straight down from the barista’s point of view. At 0.5 seconds the label says “The Hands”, over a hand above the drip tray. At 1.8 seconds it’s “The Tools”, a portafilter of tamped coffee. At 3 seconds it’s “The Canvas”, a cup of espresso. Three of the reel’s five labels have appeared by 3 seconds. “The Milk” and “The Art” follow, and the finished pattern is on screen at about 6 seconds. Each label names what’s in the frame, so it reads with the sound off.
1M+ plays across Instagram and Facebook, and 18K likes on Instagram. More in the FAVE case study.
Grape Tea Slushie review, @whereamimj
My review of Jian Cha’s Grape Tea Slushie w/ Cream Cheese, on my other account. It wasn’t a commission.
This one doesn’t follow my own rule about the shopfront. For the first 1.2 seconds it’s the Jian Cha shop on George St, with people queued at the door and the words “new thai tea brand in brisbane” on screen. At 1.3 seconds the cup is in my hand with the shop’s name on screen, and the drink’s name follows at 1.7. At 2.7 seconds a wooden spoon is in the cream cheese.
62K plays on Instagram and 31.9K views on TikTok. More on the Jian Cha page.
Thai Milk Tea Cocoa, @waitwhatmj with @jiancha.au
A Halloween reel for Jian Cha’s Thai Milk Tea Cocoa, on my account as a collab with @jiancha.au. The line “this thai tea is a bit too spooky” is on screen from the first frame. At 0.25 seconds I’m holding up the empty cup, with mochi ghosts inside it. At 1.9 seconds the topping goes on, with the cup on a scale. At 2.8 seconds the finished drink is handed across the counter to me.
What should the on-screen text and the caption say?
Give them different jobs. A reel has three places to hook someone: the cover and on-screen text, the first line of the caption, and the first comment. If the same noun turns up in two of them, you’ve used one slot twice.
The first line of the caption should say something the screen didn’t: a regret, a mistake, a verdict or a twist. On the Chapagetti video the screen asks “Do you know Chapagetti?” and the caption says “No more soggy noodles. You’re welcome”. The screen names the food. The caption points at the mistake most people make with it.
Then read the last line of your caption and the first line of your first comment out loud. If they say the same thing, cut one. How I set the line on screen is in the note on reel captions.
What I check before a drink shoot is in the note on filming drinks. If you’d rather hand over the filming and the edit, that’s a Single Reel.
Numbers as of 27 Oct 2026.
