How we choose what to recommend
The criteria, the ranking method, and — the part most sites omit — everything we do not do.
Last updated October 2, 2026
This page exists because a recommendation is worth nothing without a method behind it, and because most review sites in this category do not publish one. 102 pages on this site recommend products from a registry of 69. Here is how those choices are made.
Read the section on what we do not do as carefully as the one on criteria. It is the shorter one and it is the more useful.
The criteria we actually apply
Eight criteria, in roughly the order they carry weight on mature skin.
1. Formula and finish suitability. Whether the formula has slip — enough silicone or oil to move across a crease rather than sink into it — and whether the dried film stays flexible rather than setting rigid and cracking along a fold. These two properties decide more about how a base behaves on mature skin than anything else, and they are almost never stated on a label. We infer them from the formula type, the published ingredient list and the finish the product is sold on.
2. Documented behavior on texture. Whether a formula is documented and reported to sit on texture rather than settle into it. Satin and demi-matte finishes are favored; longwear and full-matte formulas are consistently marked down, because the hard continuous film they require is what cracks along expression lines.
3. Ingredient profile, read from the manufacturer's published list. Humectants and emollients together rather than a humectant alone; silicones as a positive rather than a negative; fragrance and drying alcohols as a negative, particularly on a thinned barrier.
4. Shade range breadth, including depth and undertone. A range that gets darker without changing undertone is treated as a short range regardless of how many shades it lists. This matters most for concealer, where a mismatch sits as a visible patch with nothing to blend it into.
5. Ease of application and packaging. Dispensing mechanism, cap type, barrel width, grip required, applicator size and whether the product can be used one-handed. This criterion is unique to this site and it is the reason the site exists. Every product page states it under How it handles.
6. Availability. The product must be currently sold. A discontinued product presented as current is useless, and a product you cannot re-buy is a problem in itself.
7. Price band, stated qualitatively. Drugstore, mid or premium. We do not print numeric prices, for the reason in the next section.
8. Reported experience of people with mature skin, where it can be sourced and cited. Every page lists its sources.
How we rank
Rankings are editorial judgments applied consistently against the criteria above. The first pick on a page is the one we would hand to the largest number of people with that specific concern, weighing formula behavior, shade range, handling and availability together.
There is no numeric scoring system on this site, deliberately. A published score implies a measurement, and a weighted formula — 40% formula, 30% shade range, 30% handling — would imply a precision we do not have. Inventing one would make the site look more rigorous and be less honest. So pages make a call and explain the reasoning instead.
Every pick carries a stated limitation. A review that lists only benefits is an advertisement, so each product on this site has a drawback written next to it. Several of them are drawbacks in the product we ranked first.
Every roundup names something we would not buy, with the reason. That section is the cheapest trust signal available and almost nobody in this category spends it.
And several pages end by telling you to use less of what you already own, or to buy nothing. A page that cannot honestly recommend a purchase says so.
What we explicitly do NOT do
Stated without hedging, because this is the section that matters.
- We do not operate a laboratory and we conduct no clinical or instrumental testing. No wear-time measurement, no transfer testing, no colorimetry.
- We do not run a tester panel. Nobody is paid or asked to trial products for this site.
- We do not claim professional cosmetology, esthetics or medical credentials. There are none behind this site, and no page implies otherwise.
- We do not claim to have personally used products we have not used. Where a page describes how a formula behaves, that is reasoned from composition, format and documented and reported behavior — and the page says so, in a scope note, every time.
- We do not accept payment, product or placement in exchange for a ranking or a position. No brand on this site has been given the opportunity to influence its position.
- We do not give medical advice. Pages touching arthritis, tremor, low vision, dry eye and rosacea describe product design and refer the condition itself to a doctor.
- We do not publish prices we cannot verify. There is no live price data for this site, so no page prints a number. "Check price on Amazon" means the live price is on Amazon, not that we are concealing it.
- We do not publish ratings, review counts or star scores, because we have no first-party rating data and would not invent any.
Why no prices
Prices change constantly, and a stale number on a review page is worse than no number — it misleads, and it makes every other figure on the page less trustworthy.
This site has no live price feed. The mechanism to carry one is built and shipped, and it reads from a single price file that a daily automated refresh would populate. Until that data exists, the file is empty by design and every price slot renders "Check price on Amazon" instead of a number.
That is also why there is no numeric price in any of our comparisons or scoring. Price bands are stated qualitatively — drugstore, mid, premium — which is a claim we can stand behind from the retail channel a product sells through.
If the live price data is switched on in future, prices will appear with a visible "as of" date and will disappear automatically once they are more than 48 hours old. That expiry is deliberate: a price with no date is not information.
How we handle being wrong
If something on this site is factually wrong, tell us: contact@primerandpoise.com or the contact form.
Factual errors are corrected promptly, and the page's updated date changes to reflect it.
If a correction materially changes a recommendation, the page says what changed rather than quietly editing the ranking.
If a product is discontinued or reformulated, the page is updated or the pick is removed. A reformulation is treated as a new product, because it is one.
We will not remove a stated limitation because a brand asks. A brand is welcome to tell us something is factually wrong and we will check it.
How often we revisit
Every page carries a real last updated date, and that date means the page was actually reviewed rather than that a script touched it.
Product roundups are reviewed on a rolling basis, with priority given to pages where a pick may have been discontinued or reformulated.
Any page is reviewed immediately when a product on it is discontinued or reformulated, when a reader reports an error, or when something we wrote turns out to be wrong.
Guides are reviewed when the underlying advice changes, which for technique content is rarely — the mechanics of how product behaves on creased skin do not move quickly.
If a page has not been revisited in a long time, the date will tell you. We would rather you see an old date than a false recent one.