Makeup planning and troubleshooting
for your face, your light, and your actual day.
Start with the job in front of you: fix what looks wrong, build the look you need, or make better use of what is already in the drawer. The scoring and evidence stay underneath until you want the why.
Start with the symptom
Start with what you can see. Then we'll look at how the look is built.
You do not need to know whether the problem is formula, prep, film order, climate, coverage appetite or maintenance yet. Pick what is visibly wrong. Every route still opens the full edit, and every field stays editable.
Cakey / visible
The base looks like makeup before it looks like skin.
Work this problem →02Won’t last
It starts well and disappears, separates or transfers before the day is over.
Work this problem →03Too oily
Shine keeps turning into more powder, then more texture.
Work this problem →04Too dry
Foundation keeps finding flakes, tightness or dehydration.
Work this problem →05Settles into lines
Coverage is collecting where you wanted it least.
Work this problem →06Bad in photographs
The mirror looked fine. The lens disagreed.
Work this problem →07Too many products
The routine is longer than the result deserves.
Work this problem →09Event problem
You need polish and longevity without building a mask.
Work this problem →10Sensitive / reactive
The formula or the number of layers may be becoming part of the problem.
Work this problem →11Mature skin
You want flexible film, useful definition and less settling.
Work this problem →Guided walkthrough
Four questions, one at a time.
Answer in the order the desk reasons in and watch the options narrow. Nothing is saved until you choose the starting point you want to use.
Live profile tuner
Move one control. See what changed.
These are the four inputs that move the result most. Change one and the desk shows you what it changes — without asking you to learn the model first.
Live recommendations
The connected path
From what you want to what to check afterwards.
The Look Plan connects your goal to the look, the roles, the application sequence, the wear day and the reassessment — with confidence on every line. It is the desk's answer to 'so what do I actually do?'
Every line names whether it is known, inferred, suggested or uncertain — so you can tell what is yours from what is the desk's best guess. Product-role cards show where each object lives on the face. The wear timeline shows what should hold and what will move, hour by hour. And the reassessment checklist tells you what to look for at the end of the day.
Comparison builder
Put any two side by side.
Pick two product types and the desk marks every criterion that differs. What matches stays quiet — that is the whole point of a comparison.
How this desk works
Four questions, and they only work in this order.
Almost every bad makeup decision is one of these asked out of turn — buying an object before naming the job, or changing the object when the sequence was the problem.
Already wearing it, and it is not working?
Skip the score. Name what you can see.
Pilling, separation, patchiness, creasing, cakiness, oxidation, poor wear. The desk ranks plausible causes against your own profile, explains what would make each mechanism plausible, and gives you something to test rather than something to buy — and says so plainly when the next useful question belongs to skincare instead.
The standard this desk holds
Problem
Cake often becomes more visible when opacity, film count, surface texture and maintenance pressures stack up — not because you have failed to find the right brand.
Outcome
A ranked kit, a pancake index with every model weight named, and a list you can take to the counter.
Difference
We score how the look is built, not swatches. Desire is allowed. Opacity still has to earn the bag.
Proof
Every model score shows why. We don't make medical claims. Sources are dated. Unknown stays unknown.
See the product before the form
Anti-pancake everyday, already scored.
A lightweight preview of the current everyday example, kept out of the first-load scoring bundle. Open it to run the live model and see every reason, caveat and evidence receipt.
Example output · everyday path
Build performance minimal — 6 objects, 3 films, about 7 minutes.
This is a lightweight preview of the current everyday example, kept out of the first-load scoring bundle. Load the example to run the live model against the editable profile and see every reason, caveat and evidence receipt.
Pancake index
34
Skin-like
66
Products
6 / 6
Layers
3
Minutes
7
Tension
15
Start from something real
8 engines · one profile
Every answer explains why.
8 engines read one profile. None of them returns a number without the weight that produced it.
I
Look Planning
Goal, look, product roles, application, wear expectations and reassessment — connected as one path, with confidence on every line.
Ia
Makeup Match
Goal-first product-type scoring, with heavier films weighed against your own dryness and oil. Every reason is visible.
II
Alternative Finder
Pathways ranked live — sheer hybrid, spot-only, one-stick, mineral control, mature flex, and more.
III
Compatibility & order
How formula pairs are likely to behave in sequence, which variables can change the result, and what to test before declaring a pair incompatible.
IV
Undertone & colour family
Match rules by depth band, colour families by undertone, and corrector colour chosen for your depth rather than a fixed colour wheel.
V
Wear & Longevity
Hour-by-hour integrity, what the model says to watch, and Veil / Edit / Editorial pathways scored for the hours you actually live.
VI
Troubleshooting
Pilling, separation, patchiness, creasing, cakiness, oxidation and poor wear — ranked plausibilities, the mechanism being tested, and what to change.
VII
Claim literacy
SPF, treatment, hybrid and barrier claims on makeup — sourceable, evidence-graded, and explicit about what remains unknown.
The seeded desk
Houses held as examples, never as rankings.
Tier is access and positioning on this desk. It is never a quality input to a score.