You Bought the LED Mask. How Would You Know if It Was Working?

An LED mask is now one of the most common things in an Australian bathroom cabinet that nobody can tell you the results of.

The devices sell for a few hundred dollars. The instructions ask for around ten minutes a day, most days, for several months before you assess anything.

That is a long time to run on faith. It is also long enough that most people quietly stop, or keep going without ever deciding whether it did anything.

The research is more interesting than the marketing

A multicentre, randomised, double-blind trial tested a home-use LED neck device against a sham device in 70 participants (Park et al., J Cosmet Dermatol 2025;24(1):e16599). Participants used it nine minutes a day, five days a week, for sixty sessions.

Improvement in the study group reached statistical significance at weeks 8, 12 and 16. Nothing meaningful separated the groups before then.

Worth noting the study looked at the neck rather than the face, in an Asian population, using one specific device. It is not a verdict on every mask on the market.

But one finding in it should change how anyone shops for this category.

The participants could not see what the investigators measured

Both the investigators and the participants rated results on the same measure, the Global Aesthetic Improvement Scale. The investigators recorded a significant difference at weeks 8, 12 and 16.

The participants, rating their own necks, did not.

The change was measurable enough for trained observers to agree on it, and invisible to the person who had shown up for sixty sessions.

This is not a story about the device underperforming. It is a story about the limits of looking at yourself in a mirror.

Your memory of your own face is the weakest instrument you own

You see your face every day, which is precisely the problem. Gradual change is the hardest kind for a daily observer to detect, because there is never a morning where anything looks different from yesterday.

Then add the money. Once you have spent several hundred dollars, you have a quiet stake in the answer being yes.

So people reach for a photo from last winter, taken in different light, at a different angle, in a different mood. That comparison cannot settle anything, and most people know it while they are doing it.

What the evidence base actually says right now

A dermatology review of light-based devices notes their growing home use and a favourable safety profile, alongside a blunt caveat: "Future clinical trials are necessary to produce standardized recommendations across conditions and skin types" (Mineroff et al., 2024).

The same authors advise that clinicians counsel patients on home-device use to manage safety and expectations. They also note that effects vary across different skin types and ethnicities.

Put plainly: there is no settled protocol, and results are expected to differ between people. Yours is the only record that answers the question for you.

It is also worth knowing where the category is heading. L'Oréal showed a flexible LED mask at CES 2026, but it is a prototype the company expects to launch in 2027, and its United States release is subject to the FDA 510(k) process. The claims being made for it are the company's own, about a device nobody has assessed independently yet.

How to run a fair test on a device you already own

Record where you are starting before you begin. A record made after four weeks of use is not a starting point, it is a guess wearing a timestamp.

Hold the conditions still. Same room, same time of day, same distance, no makeup — the variables that ruin bathroom comparisons are all controllable.

Change one thing at a time. Starting a device the same fortnight you start a new serum leaves you with two suspects and no way to separate them.

Then give it the timeline the research used. If a controlled trial needed eight to twelve weeks to separate a real device from a sham, three weeks of watching your own face will not tell you much.

And decide in advance what would count as no. A test you cannot fail is not a test, it is a purchase justification.

Where Stella fits

Stella scans your skin and keeps the readings, so the comparison is against a record rather than a memory.

A scan takes about thirty seconds. Do one before the device arrives, keep the rest of your routine unchanged, and let the trend accumulate across the twelve weeks the research suggests you will need.

At the end you will have something better than a feeling. You will have a trend, and the honest possibility that it did not move.

That last part matters. A tool that can only confirm your purchases is not measuring anything.