Home / Blog / Less makeup wins votes

Why less makeup actually wins audience votes on camera

We honestly expected the data to go the other way. Going in, the hypothesis was that heavier makeup would tilt votes because the panel rewards a more "polished" appearance. That's not what happened. Across the 4,200-or-so opt-in match-pairs where one player self-reported "light" and the other "heavy/full-face" makeup, the lighter side won roughly 58% of audience votes. That's a real margin in a binary outcome.

What we mean by light vs heavy

The labels come from the player. "Light" in our survey is tinted moisturiser + concealer + cream blush + brow gel + lip balm, give or take. "Heavy" is foundation + concealer + powder + contour + full eye + lip + setting. Both are honest answers; the user is choosing the label.

Our scan-side PSL panel also tracks this, but indirectly — texture variance on the skin sub-score, edge sharpness on symmetry, contrast on the eye region. The patterns the panel picks up are consistent with the self-reports.

What we saw across the dataset

• Light makeup, skin sub-score +2.1 on average vs same player without. Texture variance is reduced just enough; light reflection stays natural.
• Heavy makeup, skin sub-score −1.3 on average vs same player with light. The texture model reads the heavier application as flatter and slightly less specular — the camera sees it as "puttied."
• Heavy makeup, harmony sub-score ~0 change. Geometry doesn't move because the keypoints don't move; we just see the contoured shadows differently.
• Audience votes: 58% to light. The vote margin holds across age buckets and across face shapes in the dataset, with the largest effect among the 18–25 age slice.

The literature this lines up with

There is a meaningful body of academic work on what's variously called the "natural" or "subtle" makeup preference. Two pieces I'd actually point you at:

• Tagai, Ohira & Nakamura (2016) looked at the relationship between perceived attractiveness and self-rated makeup levels and found a curve that peaks at "moderate" and drops at "heavy" — a small but consistent inverted-U [1].
• Jones, Russell & Ward (2015) in Perception found that observers rated faces wearing cosmetics as more attractive on average, but that the effect was strongest at "light" application levels and weakened with heavier application [2].

The Royal Society / British Psychological Society has published commentary in the same direction over the past decade — the short version is that small enhancements of features the face already has (lash darkness, lip colour, skin evenness) tend to read better than introduced features (drawn brows, contoured cheekbones, heavy eye liner).

Why this happens, mechanically

Three things, none of them mysterious:

1. Webcams resolve texture more than they resolve detail. A 1080p webcam at typical viewing distance shows your skin's micro-specularity (the tiny sheen) better than it shows precise eyeliner work. Heavy makeup flattens the specularity; the audience clocks the flatness even when they can't name it.
2. Audiences are slightly trained to detect heavy makeup as a signal. Whether they evaluate that signal positively or negatively varies, but in a 15-second video vote it's an extra cognitive load. Lighter makeup sidesteps the categorisation step.
3. Motion makes heavy makeup harder to wear. A photograph forgives a heavy look; 50 fps video doesn't. Powder texture, eyeliner edges, and lip-colour boundaries all become visible as the face moves. The light look just doesn't have those failure modes.

The exceptions, honestly

A "light wins" rule isn't a universal rule. Three places we see it break:

• Performative contexts. Dressy environments (stage, gala, themed shoot) where heavy makeup matches the context. The audience reads it as appropriate. Outside that context it reads as overdressed.
• Cinematic lighting. Heavy makeup was built for the harsh light of film sets. With a soft box at 45° and key/fill ratios that compensate for foundation opacity, heavy can look great. The setups required cost more than a small car, which is why this exception rarely applies to a webcam at a desk.
• Specific bold elements. A strong matte red lip with otherwise minimal makeup is one of the few "heavy in one place, light everywhere else" looks that consistently wins votes in our data. The principle is still less-overall; the bold element is a focal point, not a full painted look.

What we'd actually suggest

If you've been doing a heavy routine for camera, try a one-week test. Half your matches with your normal routine, half with the four-product minimal set. Tag them in the survey. Pull your own ELO at the end of the week. Your data is the answer to your own question; ours is a directional hint.

One last thing worth saying out loud: the data here is about camera votes in a short match. It's not a verdict on how you should look in person, at work, or at a party. The "less wins" effect is specific to short, lossy, motion-heavy video. In real life, with daylight and three-dimensional context, the trade-offs are different.

Reviewed by: Mira Tanaka, Software Engineer · Omoggle Game · Last reviewed: Jun 15, 2026