Hook Anatomy of Beauty Devices
576 skincare and haircare device ads, classified frame by frame. Who is holding the device, what they say is wrong with your skin, and what they never bother to prove.
The device is a supporting character in its own ad.
Across every format in this audit, from red-light panels to curling irons, the same casting choice repeats: a person fills the opening frame. The gadget shows up a beat later, usually already in a hand. On its own, without a person attached, the device leads the ad in one out of every twenty-five.
The device rarely opens the ad, and almost never opens it alone.
When the device does appear at 00:00, it is nearly always in someone's hand. A clean product shot, with nothing else in frame, is the rarest way to open a beauty device ad in the entire sample.
Add "presenter to camera" and "presenter and product" together and a person is on screen in 86% of openings. Of the 45% of ads where the device is visible at all in the first frame, it is in someone's hand 92% of the time. It essentially never gets to introduce itself.
This holds regardless of device type. Red-light panels, cleansing brushes, hair stylers and microcurrent wands all default to the same opening move: a person talking, device optional. The category has decided that a face builds more trust in the first second than a product shot does.
Half the category names no problem. When it does, the problem is too specific for a standard list.
Nearly half of all ads (48%) state no concern whatsoever in the opening. The device simply appears, someone likes it, the ad moves on.
Only 6% of all ads name a concern that lands cleanly in a generic bucket like lack of time or a safety worry. The remaining 46% of the category names something real, but it is granular enough (a specific patch of skin, a specific frustration with a specific tool) that no shorthand category catches more than a sliver of it.
Read together, this says something simple: beauty-device buyers aren't shopping a general anxiety. They're shopping a named, narrow annoyance, and the ads that convert best tend to name it just as narrowly rather than reaching for a category-wide claim.
When a concern is raised, voice carries it almost every time. Spoken word alone accounts for 45% of the category, and spoken word paired with on-screen text or a visual demonstration accounts for another 41%. Text on screen without a voice behind it is rare.
The category has cast a single kind of person, over and over.
Presenter type, gender, and framing all point the same direction. This is a one-person, one-format category.
The presenter is a woman in 83% of ads and a man in 14%, roughly six to one. She's most often under 30 (41%) or between 30 and 49 (37%), speaking straight to camera with her face visible in 91% of ads overall. A voice with no face on screen, or hands with no face, together make up less than 8%.
What the presenter is selling is rarely a persona. In 62% of ads, the narrative archetype is best described as general wellness with no specific story attached, someone simply likes the way it makes their skin or hair look. A recovery or mobility-seeker narrative shows up in 3%, and a genuine before-and-after transformation story in 2%. The category leans on relatability over storyline.
Nearly nine in ten ads are built to look homemade.
Whatever the device costs, the footage rarely looks expensive. The category has standardized on a phone-in-hand aesthetic, even for brands running a studio budget.
UGC outnumbers studio-produced creative more than 11 to 1. The look of a phone selfie has become the category's default trust signal, not a budget constraint. Even brand-employee and founder-led ads (03) are shot in the same handheld style rather than a produced setup.
The setting backs this up. 68% of ads are filmed in a living room or bedroom, spaces that read as unstaged and personal. A studio or plain backdrop appears in 9%, and outdoor settings in 3%. Text overlay is present in 83% of ads, but like the dog food category before it, that text is doing branding and captioning work, not carrying the claim, since 91% of the time it's the presenter's own voice that's driving the message.
Most ads promise a result. Most don't show one.
The category leads overwhelmingly with outcome language, then rarely backs it with anything the viewer can independently verify.
When proof does appear, it usually isn't proof of the outcome being promised. A spec number, like an intensity setting or an LED count, shows up in 14% of ads. An actual before-and-after visual, the closest thing to evidence for a transformation claim, appears in just 11%.
Competitive framing is nearly absent. 73% of ads make no comparison at all, and another 26% gesture at "other" devices in general terms without naming one. A specific, named competitor is called out in exactly one ad out of 576. Financing and subscriptions are equally absent: 92% of ads carry no price or offer in the hook beyond an occasional promo code, and every device in this sample sells as a one-time hardware purchase, never a subscription.
The category has agreed on a formula. That's exactly where a different one stands out.
Across 576 ads and 25 brands, the same choices repeat: cast an ordinary person, shoot it like a selfie, film it in a bedroom, lead with a result, and don't bother proving it. None of that is wrong on its own, and it clearly performs, or the category wouldn't have converged on it this hard.
But a formula this consistent also leaves specific gaps sitting open. A device that gets to lead its own opening frame, a claim followed by an actual before-and-after, a named concern instead of a vague "other," any one of these breaks the pattern in a category where almost nothing does. In a market this uniform, doing the thing everyone else skips is the fastest way to stop looking like everyone else.
How this was measured.
The sample. 576 beauty and personal-care device video ads running in the United States, across 25 advertisers spanning red-light and LED therapy, microcurrent and facial toning, hair styling tools, cleansing devices, and gua sha or lymphatic tools. The roster is weighted toward the category's most active advertisers: PMD Beauty, Nood and medicube alone account for 35% of the sample. Ads were sourced from the Meta Ads Library.
Classification. Each ad was processed with Recharm's AI tagging system, which reviews the asset frame by frame and assigns structured values against a fixed schema covering the opening, the presenter, the concern raised, the claim made, and the evidence offered. Every ad carries the same field set, so any two ads are directly comparable on any dimension.
Reading the figures. Percentages are a share of all 576 ads unless stated otherwise, rounded to whole numbers. A handful of fields returned a near-universal single answer for this vertical (workout intensity, session length, movement pattern) and carried no signal worth reporting, so they are omitted here rather than padded into a chart.
Concern classification. A concern is counted as named when the ad explicitly states what's wrong that the device is there to fix. Concerns that didn't match one of the audit's standard categories are grouped as "other" rather than forced into the nearest fit.
Reuse. Every chart and figure on this page is free to reproduce with attribution to Recharm and a link to this page.