ufindMarket demand
Mercury specialist · 25%
Do people clearly need this?
Looks for recurring pain, urgency, adoption, and willingness to pay. Existing competitors alone prove supply, not demand.
Live every launch week
How voting works at ufind.best
Every eligible launch starts equally, real member support can move the ranking, and six strict AI specialists keep a large audience from becoming the only path to winning. Every product receives all six ratings plus a launch-week snapshot of independent public evidence.
The weekly result
One board, three clear point sources
1
Owner vote
Every eligible launch begins with the same 1 owner point. It can never be multiplied.
1·2·3
Member streak vote
Three active weeks unlock 2-point votes; eight unlock 3. One missed week is protected, and earlier valid activity is included.
9·7·5
AI Jury vote
The Jury's first, second and third choices receive 9, 7 and 5 points. Fourth, fifth and sixth receive 3, 2 and 1, so a well-supported product just off the podium is not flattened to nothing.
Quality gets a fair start
A strong launch does not begin at zero just because its maker has a smaller audience.
People still have a say
A community vote is meaningful social proof and can change the live ranking.
Harder to game
Flagged votes are excluded, voter identities stay private, and popularity never changes a judge's score.
Consistency earns more say
Your current streak, vote value and next milestone stay visible on your dashboard.
1
Start
1 point
Vote for another maker
2
3 weeks
2 points
Three active weeks
3
8 weeks
3 points
Maximum vote power
What counts: one valid vote for another maker in a launch week. Extra votes that week help products but do not inflate the streak.
What does not: owner votes, flagged votes, payments or repeatedly voting for the same product.
When power changes: the vote that reaches week 3 is worth 2 points; the vote that reaches week 8 is worth 3.
If a week is missed: one grace week keeps the streak alive. A second missed week starts the next streak from week 1.
Why we chose this
We wanted better launches, not louder campaigns
We keep voting deliberately strict. A person must visit the product's website before voting, suspicious votes can be flagged, and only valid votes from the launch week count. That makes the signal cleaner, but naturally means there are fewer votes than on a frictionless popularity board.
The six specialist functions fill that gap without replacing the community. Their shared Top 3 vote creates a quality baseline, while real member support can still move any launch up the board.
Our judges are strict on purpose. Modern visuals and basic AI features are expected now, not automatic signs of quality. A score around 65 already means a product is convincing, while scores above 85 should be rare.
A score compares a product with the rest of its own launch week. We keep making the judges' questions harder as products change, so a score given in an earlier week was set against a different measure. Every score on the site carries the date it was given, and a product's standing inside its own week is the part that stays comparable.
Six independent perspectives
Six questions, one clearer picture
Each specialist looks at one part of the same product. They work independently and meet only in the shared score.
Uniqueness
Uranus specialist · 30%
Is there a meaningful reason to choose it?
Compares the product with real alternatives, not with an empty market.
Execution
Mars specialist · 15%
Does the product deliver its promise?
Rewards a working product, not a polished claim.
User experience
Venus specialist · 10%
Is it easy to understand and use?
Treats modern visual polish as a baseline, not a breakthrough.
Trust
Saturn specialist · 10%
Does it feel credible and transparent?
Checks claims, pricing, ownership, privacy, and support.
Practical value
Jupiter specialist · 10%
Does it solve the promised job well?
Looks for a useful, complete outcome without rewarding launch polish twice.
One shared verdict
Six perspectives meet in one score
The specialists stay separate while they review. Their views come together only at the end, then the product is compared with the rest of its weekly cohort.
Market demand64
Uniqueness72
Execution66
User experience74
Trust68
Practical value78
Shared score
69.5
out of 100
Compared with the weekly cohort
1st place
9Jury
4Community
13Weekly total
The score compares product quality. It does not drop straight into the board. Only the resulting weekly place becomes Jury points — 9, 7, 5, 3, 2, 1, then 0 — and valid community points are added once. The top three still take the week; the tail only decides the order below it.
Independent market research
Claims are a starting point, not the evidence
The product page tells us what to investigate. It cannot prove its own demand, originality, or trust.
- 01 Neutral market question We remove the brand name and promotional language, then identify the concrete problem, audience, solution category, and supported global, regional, or local market scope.
- 02 Five evidence paths Separate searches look for recurring pain, visible alternatives and their solution approaches, purchasing intent, adoption friction or workarounds, and independent quality or trust reports.
- 03 Relevance and independence checks The product's own domain is excluded. Generic pages are rejected, duplicate URLs are merged under their strongest supported research intent, and local-language market terms remain matchable.
- 04 Frozen weekly snapshot We limit repeated domains, seek honest coverage across all five research angles, publish only the strongest relevant sources, and freeze that evidence after the first review.
Not every launch has evidence on every platform, and more links do not automatically mean a better score. The first pass prioritizes established review, community, store, and developer sources. When it finds fewer than 12 relevant links or five platforms, a broader public search runs and the same relevance and diversity filters are applied. After publication, routine design or deployment changes do not run that research again.
Reading a public report
Every view explains the same six ratings
The expanded report separates the decision from its supporting context. No chart, evidence count, alternative count, or coverage bar creates another score.
Market signal
A compact reading of published demand, differentiation, and execution-related ratings.
Evidence mix
The number of retained pain, demand, alternative, adoption, and quality sources, not a quality grade.
Scope and alternatives
Whether the product is global, regional, local, or unclear, plus verified direct and adjacent tools and supported approaches.
Specialist profile
All six ratings on one 0-100 scale so strengths and the weakest constraint are visible together.
Rating details
Each function shows its explanation, strengths, concerns, cited source IDs, and any evidence-based score cap.
Public research trail
Every retained public source remains openable and uses nofollow; weak or mismatched pages are not kept to fill a quota.
Open where transparency creates trust
The six criteria and weights, evidence links, score explanations, update timing, member vote rules, and final point calculation are public. Anyone can inspect how a result was formed.
Private only where disclosure invites gaming
Exact prompts, anti-fraud thresholds, anomaly rules, and internal abuse signals stay private. They do not introduce hidden scoring criteria; they protect the published rules from being reverse-engineered.
During launch week
Scores arrive early, winners wait until the end
The live standing can move all week. Nothing is final until voting closes.
- 01 Launch day Shortly after the cohort goes live, all six specialists publish their scores, explanations, and one frozen evidence snapshot.
- 02 All week Valid member votes add 1, 2 or 3 points and update the live ranking immediately.
- 03 Evidence remains fixed Routine product, design, or deployment changes do not rerun market research. A review changes only after a real methodology update or a confirmed source problem.
- 04 Winners and rewards When voting closes, the final Top 3 unlock a place-specific Product of the Week badge and receive a congratulations email. Their verified badge keeps the winner placement and dofollow reward active.
What the specialists can consider
The public product, website, store page, and working features. Public pricing, policies, product claims, and launch materials. The audience, category, and problem the product claims to solve. Independent public evidence of demand, complaints, alternatives, and willingness to pay captured for that launch week.
What never buys a better score
A Premium launch or any other paid placement. The size of the founder's audience or where we discovered the product. Community votes, popularity, founder identity, or personal relationships.
FAQ
A few fair questions
Scoring & fairness
How the six ratings become one fair weekly signal.
Why combine the AI Jury with community votes?
Small launch communities can produce very few votes, and audience size is not the same as product quality. Six isolated specialists create a quality baseline, while every valid member point is added directly and can still change the order.
How is the overall Jury score calculated?
Each rating uses its published weight: Market demand 25%, Uniqueness 30%, Execution 15%, User experience 10%, Trust 10%, and Practical value 10%. A score around 65 is already convincing; modern visual polish is treated as a baseline, while Practical value asks whether the product solves its promised job directly and completely.
How does the Jury score become weekly points?
The weighted 0-100 score orders launches inside the cohort. Only first, second, and third Jury placement becomes 9, 7, and 5 points; every other launch receives 0 Jury points, and valid community points are added separately. Specialist category-leader labels are explanatory and add no points.
Can Premium, popularity, or another specialist influence a rating?
No. Paid placement, founder reach, community votes, and Premium status are invisible to the judges. Each specialist evaluates only its own measure without seeing another specialist's score, and Premium does not compete for the free weekly Top 3.
How are exact ties resolved?
Equal overall Jury scores first favor the more balanced profile with the higher weakest specialist rating, then higher Uniqueness and Demand ratings. Equal final weekly points favor the higher Jury score, then community points, the earlier launch time, and finally a stable product ID.
Evidence & reviews
What the Jury can trust, retain, and update.
Where does independent evidence come from, and is there a source quota?
Five research paths look across public communities, reviews, app stores, developer discussions, alternatives, complaints, and adoption signals. A broader pass runs when the curated result is too narrow, but twenty sources is only a maximum, never a target; fewer strong sources are better than padded counts.
How are demand, similar tools, and market scope checked?
Mercury looks for recurring pain, purchasing intent, and adoption friction. Uranus uses verified named alternatives, separates direct from adjacent tools, and compares supported approaches. Global, regional, or local scope comes from product signals such as language, regulation, payments, data, or geography, never the maker's nationality. These summaries inform published ratings but add no hidden points.
What happens when a source matches the topic but not the task?
It is rejected. A page about broken YouTube thumbnails is not evidence for a thumbnail downloader unless it also discusses downloading or saving them. Broad category overlap cannot substitute for the concrete job the product performs.
Can a published review or its evidence change?
The first weekly market snapshot and its concrete task definition stay frozen. A documented methodology update can rejudge the stored snapshot, and a confirmed mismatch can be removed without adding a replacement. Fresh market research runs only for an explicit evidence exception; routine design changes do not refresh the score or alter member votes.
What is public, and what happens when proof is missing?
Criteria, weights, source links, explanations, timing, point math, and the stored model and rubric version are public. Exact prompts and anti-abuse thresholds remain private. The Jury never invents inaccessible evidence: unclear or missing proof lowers confidence, and winner selection waits if any eligible launch lacks a complete valid review.
Voting & rewards
How member support moves the live board and what winners unlock.
How does member vote power grow?
A verified member starts at 1 point. Supporting another maker in three consecutive active weeks unlocks 2-point votes and eight weeks unlocks 3. One missed week is protected, a second resets the streak, and account age alone never increases vote power.
What must happen before a member vote counts, and are voter names public?
Voting unlocks only after the member visits the product website, while the current score stays visible beforehand. We never publish who supported a specific product; the streak leaderboard shows activity without revealing the vote, and flagged or invalid votes are excluded.
Why does every eligible launch begin with one point?
Every eligible launch receives its maker's point automatically when it goes live. Everyone gets the same start, the owner point always stays at 1, and it never receives a streak multiplier.
Why can the live ranking change before voting closes?
Member points can move the board all week. A documented model or methodology update may also rejudge the same frozen evidence while voting is open; the stored version changes transparently and existing member votes remain untouched. Nothing is final until the weekly close.
What do the final Top 3 receive?
After voting closes, first, second, and third place unlock matching Product of the Week badges in light and dark themes, with HTML, React, and downloadable SVG options. One congratulations email links to the reward. Makers may replace the launch badge or keep it; either verified badge keeps the winner placement and dofollow reward active.
The result stays inspectable
Every reviewed launch shows one clear overall score first. The expanded report adds one consistent signal summary, a six-specialist score profile, research coverage, cited evidence notes, and a fully expandable public source trail. These views explain the published ratings without creating hidden points.