Strong reaction is the fastest beauty ad hook to test
Several of the top-performing beauty hooks in the data opened with nothing but a raw emotional interjection before any context:
- “Omg look at that, that is so gross.”
- “Wow…”
- “Ok, wait. I was not expecting to love this as much as I do.”
The reaction reads as involuntary, and that’s exactly what turns it into a pattern interrupt in a feed full of scripted pitches. It doesn’t require the viewer to know anything about the product yet, which makes it product-agnostic and the fastest format to test on a new item.
Each line also implies something off-screen. “Omg look at that” is a reaction to something, and the natural response is to stay long enough to find out what. But because the reaction feels genuine rather than promised, it doesn’t read as a tease.
Notice too that one of the three is negative. “That is so gross” performing alongside “I was not expecting to love this” tells you the mechanic isn’t about enthusiasm. It’s about the reaction being unguarded.
Audience call-out hooks create instant self-identification
Some of the strongest performers named the viewer or the occasion in the opening line:
- “You girls keep asking what I’m using on my hair to help it grow seven inches.”
- “You got to see what I just got for my husband.”
- “If you don’t know what to get your mom for Mother’s Day, this is it.”
It works because it creates instant self-identification. The right viewer knows within a word or two that the video is speaking to them specifically, not to a general beauty audience.
The first example is doing double duty. “You girls keep asking” doesn’t only name the audience. It establishes that other people have already asked, which is social proof delivered before any claim is made. The creator isn’t recommending the product; she’s answering demand, an easier position to be believed from.
The gifting examples name a moment rather than a person. A viewer with Mother’s Day coming up is already carrying an unresolved task, and a hook that names it lands as a solution, though occasion call-outs have a shelf life the identity ones don’t. Either way, this is best for beauty brands with a defined segment or moment to speak to.
If you’re briefing this one: give the creator the exact person or occasion to name in the first sentence. A call-out that has to be inferred isn’t a call-out.
Curiosity gap hooks withhold the product until the reveal
Other top performers teased the payoff and withheld the product entirely:
- “I tried everything, really everything, for my skin that the internet talked about.”
- “I didn’t notice this area aging until I couldn’t unsee it. Then I found this.”
- “I kept seeing this video on social media.”
It forces the viewer to keep watching for the reveal, since the hook deliberately doesn’t name what actually fixed the problem.
This works better in beauty than in most categories because the withheld thing is usually a problem the viewer already has. “I didn’t notice this area aging until I couldn’t unsee it” is a gap the viewer fills in themselves, and whichever area they think of is the one they’re most self-conscious about. The hook doesn’t have to be specific because the viewer supplies the specificity.
The third example works differently. “I tried everything, really everything” isn’t withholding a product so much as establishing exhaustion, and exhaustion is credibility in skincare. Someone who has already failed repeatedly is a more trustworthy narrator than someone who found the answer on the first try.
The failure mode is a gap with nothing behind it: if the reveal is an ordinary product presented as a discovery, the viewer feels the switch. It’s best suited to problem-solution beauty products, especially skincare concerns where the viewer already recognizes the problem and just wants the answer.
Demo hooks let the result carry the weight
The rest of the top performers just showed the product and let the result carry the hook instead of talking around it:
- “Ok stop. This side has SPF sunscreen on it, this side does not.”
- “This lip oil actually adjusts to your lips.”
- “Base is done, this is my favorite part.”
Beauty is a visual category, so a result the viewer can see is simply more convincing than one that’s only described. The deeper advantage is that a demo can’t be disbelieved the way a claim can. A split screen isn’t asking for trust. It shows an outcome and lets the viewer draw the conclusion, side-stepping the scepticism reflex most beauty claims run into.
“Base is done, this is my favorite part” is the quietest line of the three. It makes no claim at all. It just signals that something worth watching is about to happen, from someone clearly mid-routine and not mid-pitch. In a category where viewers watch application videos for their own sake, that’s enough.
This works best for beauty products with a payoff you can actually show: a color-adapting formula, a texture change, a visible before-and-after.
The best hooks stack multiple mechanics
Most of the top performers in the beauty data weren’t pure examples of a single mechanic, and that’s the point. A couple of the strongest hooks combine two or three mechanics in a single line:
- “You girls keep asking what I’m using on my hair to help it grow seven inches” is a call-out (“you girls”) stacked with withholding (the product itself is never named) and a bold, specific claim (seven inches), all in one sentence.
- “Ok stop, this side has SPF sunscreen on it, this side does not” opens with a strong reaction (“ok stop”) layered directly onto a demo. The split-screen comparison is the proof and the pattern interrupt at the same time.
- “I didn’t notice this area aging until I couldn’t unsee it, then I found this” combines a relatable problem with a curiosity gap built around a single withheld word: “this.”
Stacking works because the mechanics do different jobs in different seconds. The reaction buys the first half-second, the call-out decides who stays, and the withheld payoff carries them to the reveal. A single-mechanic hook has to do all three with one move, and usually manages one, which is also why hooks plateau.
The takeaway for a beauty creative brief: don’t treat these four mechanics as separate boxes to pick from. When one is performing, the next test is usually to layer a second on top rather than look for an entirely new hook.
A great hook isn’t always a great ad
Hook rate and ROAS measure different things, and in the beauty data they pulled in opposite directions often enough to matter. A scroll-stopping hook can still be a money-loser.
The two highest hook rates in the beauty set converted at 0.86 and 0.28 ROAS. Both were excellent at stopping the scroll and unprofitable at the spend level tested. Meanwhile, a handful of beauty ads with meaningfully lower hook rates had ROAS above 5.00.
Read those two figures together: picking scaling winners off the hook-rate leaderboard would have put budget behind two unprofitable ads while the ads clearing 5.00 ROAS sat further down the list.
The takeaway: use hook rate to diagnose the first three seconds, but never pick your scaling winner on it. Judge the ad on ROAS, or you’ll pour budget into creative that’s great at being watched and bad at selling.
How to apply this to your next beauty creator brief
Start with strong reaction on any new beauty product. It’s the fastest mechanic to brief, it doesn’t require the creator to know the product yet, and it works across categories inside beauty.
Layer in audience call-out and curiosity gap once you know your segment or your product’s problem-solution story. Call-out needs a real person, role, or occasion to name in the opening line; curiosity gap needs a real, withheld payoff, not a vague tease. Reach for demo whenever the product has something genuinely visible to show. Once a hook is converting, stack a second mechanic onto it rather than starting from scratch.
Never green-light a scaling decision on hook rate alone. An ad that stops the scroll but doesn’t sell is a sign the audience, the product match, or the offer needs a second look, not a reason to spend more against it.
If there’s one habit worth changing, it’s the order of the two numbers. Look at ROAS first and hook rate second. Hook rate then stops being a verdict and becomes what it’s good for: telling you whether a profitable ad has room to grow, and whether an unprofitable one is failing at the top or further down.
