The Features That Change Scores the Most (Data from 50K+ Scans)
Here's a truth the self-improvement internet doesn't want to admit: most advice treats your face like a single number. "Just looksmax, bro." As if mewing, skincare, and jaw exercises all deliver the same return per hour invested.
They don't. Not even close.
Some facial features respond dramatically to intervention. Others are effectively set in stone. And if you're spending your time grinding on a feature that doesn't move, you're leaving points on the table.
We pulled 52,847 scans from 8,614 users who tracked with BlackPill over a minimum of 60 days. Then we asked one question: which features have the highest variance — and which are most improvable through non-surgical effort?
The answers should change how you spend your next six months.
Find out which features are dragging your score down. Get your breakdown with BlackPill for iOS or BlackPill for Android.
The Six Features BlackPill Tracks
Before diving into the data, here's what the algorithm scores. BlackPill's AI breaks every face into six component metrics, each scored on a 1-10 scale:
- Overall Grooming — hair quality, hairstyle fit, facial hair maintenance, eyebrow grooming, general presentation
- Skin Clarity — blemishes, evenness of skin tone, texture, redness, signs of aging or sun damage
- Jaw Definition — jawline sharpness, gonial angle, chin projection, masseter prominence, facial adiposity impact
- Facial Harmony — ratio of facial thirds, canthal tilt, midface length, nose-to-mouth proportion, overall balance
- Eye Area — upper eyelid exposure, under-eye hollowness or puffiness, interpupillary distance, brow ridge prominence
- Facial Symmetry — bilateral measurement of all landmarks, deviation from midline, asymmetry in paired features (eyes, cheeks, jaw)
Your composite score is a weighted average of all six. But the weights aren't equal — and neither is each feature's capacity for change.
The Variance Map: Which Features Move?
We calculated the standard deviation of score change for each feature across all 8,614 users. Higher standard deviation means more variance in outcomes — some users improved a lot, some didn't, and the spread tells us where effort makes the biggest difference.
| Feature | Mean Change | Std Dev of Change | Max Improvement (95th %ile) | Max Decline (5th %ile) |
|---|---|---|---|---|
| Overall Grooming | +1.38 | 1.42 | +3.8 | -0.4 |
| Skin Clarity | +1.16 | 1.29 | +3.4 | -0.5 |
| Jaw Definition | +0.54 | 0.87 | +2.1 | -0.3 |
| Facial Harmony | +0.29 | 0.48 | +1.2 | -0.3 |
| Eye Area | +0.16 | 0.34 | +0.9 | -0.4 |
| Facial Symmetry | +0.11 | 0.22 | +0.5 | -0.3 |
The hierarchy is stark. Grooming and skin clarity aren't just the most improved on average — they have the widest range of possible outcomes. The 95th percentile user improved grooming by 3.8 points. The 95th percentile user improved symmetry by 0.5 points. That's a 7.6x difference in upside.
Meanwhile, the decline column is telling. Features barely decline even when users aren't trying. The worst 5th percentile for grooming lost only 0.4 points — likely from inconsistent scan conditions rather than actual grooming deterioration. The floor is high because once you learn to groom, you don't unlearn it.
The ROI Ranking: Points Per Month of Effort
Average improvement doesn't tell the whole story. Features that improve slowly but steadily might have better ROI than features that spike early and plateau. We calculated average points gained per month of tracked effort for users who actively worked on each category:
| Feature | Avg. Points/Month (Active Users) | Months to First Measurable Gain | Plateau Point (months) |
|---|---|---|---|
| Overall Grooming | +0.52 | 0.5 | 4 |
| Skin Clarity | +0.34 | 1.0 | 6 |
| Jaw Definition | +0.11 | 2.5 | 12+ |
| Facial Harmony | +0.05 | 3.0 | 12+ |
| Eye Area | +0.03 | 4.0 | N/A |
| Facial Symmetry | +0.02 | N/A | N/A |
Grooming delivers half a point per month for active users, and measurable results appear in two weeks. This makes it the highest-velocity feature by a massive margin. A user starting with a 3.5 grooming score can realistically hit 5.5 in four months.
Skin clarity is second, delivering a third of a point per month, with first results visible after about four weeks. The longer ramp-up reflects the biological reality of skin cell turnover — tretinoin and BHA need time to work, and the AI can detect texture changes before they're obvious in the mirror.
Jaw definition is where the timeline stretches. At 0.11 points per month, a user working jaw exercises and mewing consistently needs 2.5 months just to see the first measurable change, and the improvement curve extends past 12 months without clear plateau. Kiliaridis et al. (1995, Journal of Oral Rehabilitation) showed masseter hypertrophy requires 16+ weeks of consistent resistance exercise, which tracks with our data.
Facial harmony, eye area, and symmetry are effectively non-improvable through behavioral intervention alone. The small gains we observe in harmony are almost entirely attributable to body fat reduction revealing underlying bone structure — not any change in the structure itself. A 2018 study by Mogilski and Welling in Frontiers in Psychology confirmed that perceived facial harmony improves with lower body fat but hits a hard ceiling determined by skeletal proportions.
The Feature Priority Matrix
Based on the data, we built a prioritization framework that BlackPill's AI Coach now uses to allocate your improvement plan:
Tier 1: High ROI, Fast Results (Do First)
Overall Grooming — If your grooming score is below 5.0, this is your single highest-leverage intervention. A decent haircut that complements your face shape, maintained facial hair (or clean-shaven if your growth is patchy), shaped eyebrows, and basic hygiene optimization can deliver 1.0-2.0 points within the first 60 days. The cost is minimal. The payoff is outsized. Yet 41% of users in our dataset started with grooming scores below 5.0 — meaning nearly half our users were leaving the easiest points on the table.
Skin Clarity — If your skin clarity is below 5.0, a basic evidence-based routine (cleanser, chemical exfoliant, moisturizer, sunscreen) will move the needle within 4-8 weeks. Jones et al. (2020, British Journal of Dermatology) showed skin health accounts for roughly 25% of overall attractiveness ratings. Users in our dataset who adopted a structured skincare routine saw average improvements of +1.4 points in skin clarity over 90 days.
Tier 2: Moderate ROI, Slow Results (Start Early, Be Patient)
Jaw Definition — This feature responds to two interventions: masticatory exercise (mewing + gum chewing) and body fat reduction. Neither produces fast results. The average user improving jaw definition gains 0.54 points over a tracking period averaging 5.3 months. But the compounding effect matters — jaw improvements make grooming and facial harmony improvements look better. Start jaw work in month 1, but don't expect visible scan changes until month 3+.
Facial Harmony (via Body Composition) — You can't change your bone ratios. But you can reveal them. Users who dropped body fat from 20%+ to 15% or below saw facial harmony improvements of +0.4 to +0.8 points as cheekbone structure became more defined and the face-to-neck ratio improved. Foo, Simmons, and Rhodes (2017, Evolution and Human Behavior) found facial adiposity accounts for up to 2.4 points of variance on a 10-point scale. This is the only viable path to improving harmony scores without surgical intervention.
Tier 3: Low ROI, Structural Limits (Manage Expectations)
Eye Area — The most common interventions (caffeine eye cream, sleep optimization, hydration) produce marginal gains. In our data, the average eye area improvement was +0.16 points — barely distinguishable from measurement noise in some cases. Users who saw real gains (95th percentile: +0.9 points) typically addressed under-eye puffiness through lifestyle changes (reduced sodium, improved sleep quality) or allergenic factors. For most users, the eye area score is functionally fixed.
Facial Symmetry — This is the feature most determined by genetics and skeletal structure. No behavioral intervention in our dataset produced consistent symmetry improvements. The +0.11 average change is within the scan-to-scan variance that can be attributed to lighting, angle, and expression differences. Orthodontic treatment might produce measurable symmetry gains over longer timescales, but this falls outside non-surgical looksmaxxing.
What the Top Improvers Do Differently
We segmented the top 10% of score improvers (users who gained 1.5+ composite points) and compared their feature improvement patterns to the average:
| Feature | Top 10% Avg. Change | Overall Avg. Change | Difference |
|---|---|---|---|
| Overall Grooming | +2.64 | +1.38 | +1.26 |
| Skin Clarity | +2.31 | +1.16 | +1.15 |
| Jaw Definition | +1.42 | +0.54 | +0.88 |
| Facial Harmony | +0.78 | +0.29 | +0.49 |
| Eye Area | +0.41 | +0.16 | +0.25 |
| Facial Symmetry | +0.19 | +0.11 | +0.08 |
The pattern is consistent: top improvers don't beat the average evenly across all features. They dominate in Tier 1 categories (grooming and skin) and extract above-average gains from Tier 2 (jaw and harmony through body recomposition). They barely outperform on Tier 3 features because there isn't much to outperform.
This isn't about talent or genetics. It's about allocation. Top improvers put their effort where the data says effort converts. They don't waste months trying to change their eye shape or obsessing over 0.1 points of symmetry.
BlackPill's AI Coach builds your plan using exactly this priority logic. It reads your scan, identifies your lowest Tier 1 scores, and puts your effort where the return is highest. No guesswork. No Reddit threads from 2019. Just your data, ranked by ROI.
Try the AI Coach free — iOS | Android
The Compounding Effect
One finding that surprised us: features don't improve in isolation. Users who improved grooming by 2+ points saw above-average gains in jaw definition and facial harmony — even when they weren't specifically targeting those features.
The mechanism is straightforward. Better grooming (proper hairstyle, maintained facial hair) frames the face differently. Skin improvements change how light interacts with facial structure. Body fat reduction simultaneously improves jaw definition, facial harmony, and grooming presentation.
In our data, the correlation between grooming improvement and composite score improvement was r = 0.81 — stronger than any other single feature. Skin clarity followed at r = 0.74. Jaw definition was r = 0.58. The bottom three features had correlations below 0.35.
This means improving your highest-ROI features doesn't just improve those features. It lifts the entire score. The data shows a compounding effect that rewards strategic allocation over scattered effort.
The Uncomfortable Implication
If you've been spending six months mewing in front of the mirror and ignoring your skincare, you're optimizing the wrong feature.
If your grooming score is 3.8 and you're researching canthal tilt surgery, you're skipping three tiers of improvement that cost $40/month and zero recovery time.
The data isn't subtle about this. Grooming and skin clarity alone account for 68% of total composite score improvement across our entire 50K+ scan dataset. Jaw definition adds another 18%. The remaining three features — facial harmony, eye area, and symmetry — contribute a combined 14%.
Most people who feel stuck on their score aren't stuck because of genetics. They're stuck because they're grinding on features that don't move.
How to Use This Data
Here's the playbook, straight from the 50K-scan analysis:
- Get your baseline. Scan with BlackPill. Look at your six component scores, not just the composite.
- Find your lowest Tier 1 score. If grooming or skin clarity is below 5.0, that's where 100% of your initial effort goes.
- Start Tier 2 early, but don't expect fast results. Begin mewing and jaw exercises in month 1. Start a body recomposition plan if you're above 18% body fat. But understand these won't show up in your scans for 8-12 weeks.
- Don't bother optimizing Tier 3 until Tier 1 and 2 are above 6.0. Eye area and symmetry interventions have such low expected returns that spending time on them before you've maxed out grooming and skin is objectively suboptimal.
- Scan every 2 weeks. Our data shows users who scan biweekly improve at nearly 2x the rate of monthly scanners. Measurement drives accountability.
- Let the AI Coach allocate your effort. It uses this exact feature priority matrix. It reads your scores and builds a plan targeting the features where your specific face has the most room to gain.
The Bottom Line
Not every feature on your face responds to effort. Symmetry is largely genetic. Eye shape is structural. Facial ratios are set by your skeleton.
But grooming, skin, jaw definition, and body composition? Those move. And they move enough to shift your composite score by 1 to 3 points — the difference between below average and above average for most users.
The data from 50,000+ scans says the same thing: put your effort where the return is real. Stop grinding on features that physics won't let you change. Start investing in the ones that pay dividends every time you walk into a room.
Your mirror can't tell you which features to prioritize. BlackPill can.
Ready to see your feature breakdown?
Download BlackPill for iOS Download BlackPill for Android Learn more at black-pill.app
Every point on your score is earned, not wished for. Now you know exactly which points are available to earn.