HOW UNMASQUE WORKS

Empowering shoppers to Unmasque the truth behind the brands and products they love.

Social commerce has a problem. Your feed is packed with viral products, glowing "reviews," and sponsored posts pretending to be recommendations. Somebody is always getting paid to tell you it's amazing.

Unmasque is built for the other side of the table — yours. We do not take money from brands. We do not let companies open accounts or edit their scores. We read what REAL people are actually saying, and we tell you plainly whether a brand is worth your money.

WHERE OUR VERDICTS COME FROM

Every verdict is built from two sources of REAL, public discussion:

Reddit — posts and threads from the last 4 years. Reddit accounts have karma, comment histories, and communities that call out fakes. It is hard to fake a track record.
YouTube — reviews and haul videos from the last 2 years. Creators show the product on camera and have to answer to their audience in the comments.

We do not use press releases. We do not use brand-submitted testimonials. We do not run sponsored placements. If a brand is loved, you will hear it from actual buyers. If a brand is trouble, you will hear that too.

THE GRADE SCALE

Grade	Score	Label	Meaning
A	85–100	TRUSTED	REAL buyers overwhelmingly back this brand. Buy with confidence.
B	70–84	RELIABLE	Mostly positive with small caveats. A safe bet for most people.
C	55–69	MIXED	REAL disagreement out there. Read the reviews before deciding.
D	35–54	RISKY	Serious recurring complaints. Proceed carefully.
F	0–34	AVOID	Widespread negative experience. We would walk away.

Two situations do not get a grade. If the name is ambiguous (a word that also means something unrelated) or if this search has many options that could populate, we ask you to pick a category so we score the right brand. And if there simply is not enough real discussion out there yet, we say so instead of making up a number.

THE FIVE TRUST SIGNALS

SENTIMENT STRENGTH

30% OF SCORE

How positive or negative the overall conversation is. Are people happy, or are they warning others away?

EVIDENCE CONSISTENCY

20% OF SCORE

How much reviewers agree. Strong consensus is a signal. When people contradict each other, we say so instead of picking a side.

REVIEW DEPTH

15% OF SCORE

Whether reviewers give specific details or vague one-liners. Detailed posts with REAL use cases weigh more than "it's great!" throwaways.

COVERAGE VOLUME

15% OF SCORE

How much REAL discussion actually exists. A brand with 30 posts is telling you something a brand with 3 posts cannot.

RED FLAG SEVERITY

20% OF SCORE

The seriousness of complaints. Minor gripes score differently than scam reports, safety issues, or fraud claims.

The five signals are weighted and combined into a single 0–100 score, then mapped to a letter grade.

HOW A VERDICT IS BUILT

01
You search a brand.
If the name is ambiguous, we ask you which kind of brand you meant.
02
We pull REAL discussion.
Reddit posts from the last 4 years and YouTube reviews from the last 2 years, in parallel.
03
AI reads the reviews.
Every post and video is analyzed against the five trust signals.
04
We calculate the score.
The five signals combine into a weighted 0–100 score, then a letter grade with green flags and red flags called out.
05
We show you the verdict.
Grade, summary, and the actual posts and videos it came from — so you can read the source yourself.
06
Verdicts stay fresh.
We cache each verdict for up to a week; after that, the next search pulls new discussion and re-scores from scratch.

WHAT WE DON'T DO

Unmasque is built for you. Not for brands. So the rules are simple:

Brands cannot pay for a better score, better placement, or any kind of promotion.
Brands cannot open accounts on Unmasque. There is no dashboard, no submissions, no "claim your listing."
Brands cannot edit or remove their verdict. If real people are unhappy, the score reflects that.

This is the long version of our promise: built for you, not for brands.

OUR LIMITATIONS AND HOW TO FLAG A MISS

We would rather be honest about what we cannot do than pretend to be perfect.

AI can misread sarcasm and jokes. We tune for it, but it happens.
Very small or brand-new companies may not have enough real discussion yet. We say "not enough data" instead of guessing.
A verdict is a snapshot of real opinion right now, not a personal guarantee.
Accounts get 3 brand searches a day (three unique brands, so going back to a verdict you already pulled today never costs you a search). Your count resets at midnight UTC. Unlimited searches are coming with Unmasque Pro.

Think something looks wrong or have any critiques? Get in touch with us. Contact us and we will get back to you.