Votes, Money and Facts: Three Separate Signals

Open almost any "best of" page and you will see an order. First place, second place, third. The order feels like a single verdict, as if someone weighed everything and announced a result. Often, though, it is the product of several very different inputs stirred together. When they are stirred, the reader loses the ability to ask the most important question: why is this at the top?

The approach we recommend, and the one this site follows, is to keep three kinds of signal separate and visibly labelled. Each answers a different question, and each can mislead if it is mistaken for another.

Signal one: community votes

A vote says "I liked this". Visitors who take a moment to support a product tell you something about its appeal, its launch and its audience.

What votes are good for:

• Showing which products caught people's attention in a period
• Surfacing newcomers that critics have not noticed yet
• Giving makers a direct, human response

What votes are poor at:

• Measuring long-term quality
• Reflecting people who never visit or never vote
• Resisting coordinated campaigns unless protected

Writing on the wisdom of crowds points to conditions under which a group's judgement is best: diverse opinions, independent individual judgements and a sizeable group. Vote counts are most meaningful when those conditions roughly hold. They are least meaningful when the same prompt reaches everyone, when voters copy each other or when a small group casts a large share of the votes.

For this reason a credible voting system has rules: one account per person, limits on how many votes each person may give, checks on suspicious patterns and a policy against paid or traded votes. On this site, votes can never be bought. That sentence is not decoration. It means no price list, no package and no arrangement exists in which money turns into votes.

Signal two: money contributed

A paid placement says "this business chose to spend on visibility". That is honest advertising. People understand it from newspapers and search results, and they do not mind as long as it is labelled.

What paid placement is good for:

• Giving a business a way to reach an audience quickly
• Funding the running of the platform
• Letting small advertisers compete for a visible position

What it cannot do:

• Say anything about quality
• Replace reviews, votes or proof
• Guarantee results

Google's guidance on links makes the same distinction for the web at large: advertisements and paid placements should be marked as sponsored, so search systems and readers can tell them apart from editorial references. A leaderboard page applies the same logic to its own layout. If a position was purchased, the page says so.

Signal three: verified and factual data

Some information can be checked against a source. A site's backlink strength, as measured by a named tool on a named day. A launch date. A confirmed company registration. A listing's age. These facts do not depend on opinion or budget.

What facts are good for:

• Giving readers anchors that do not shift with mood
• Letting people verify claims for themselves
• Providing a neutral way to compare entries

What they cannot do:

• Capture everything about quality
• Stay true forever, because data ages
• Settle questions of taste

Take Domain Rating as an example. Ahrefs describes it as a metric showing the relative strength of a website's backlink profile, and says clearly that it is purely link-based, leaving out search traffic, domain age and the popularity of a parent brand. It is a fact about links on a given day, useful for what it is. Treating it as a verdict on a business would be a mistake.

What goes wrong when signals are blended

Imagine a board whose score is 50% votes, 30% money and 20% a link metric. A product appears at the top. Why?

• It might have strong community support.
• It might have a large advertising budget.
• It might have a high link score from older, established sources.
• It might be a mixture of two or three of these.

You cannot tell, and neither can the businesses on the board. The blend hides information. It also invites gaming: if money contributes to the score, a rich business can climb without earning anything. If a link metric contributes, someone can buy links. And if votes count only in part, voters feel their voice is diluted.

Charles Goodhart's observation that a measure becomes a poor measure once it is a target applies fully. A blended score is a target with several handles, each of which can be pulled. Keeping signals separate means that gaming one does not corrupt the others.

How separation looks on a page

A well-designed leaderboard page makes the separation obvious without making the reader work.

• A labelled paid strip at the top or side, with the word "Paid" or "Sponsored", not just a colour.
• A community list ordered by votes, with a note on the voting rules.
• A facts column or tab showing checkable data with its date, such as "DR checked 3 March".
• A method link that explains each system in plain words.
• Consistent symbols that do not rely on colour alone: arrows and words for movement, text for labels.

If a reader can answer "is this paid, voted or measured?" for every entry in a second, the page is doing its job.

What you can do as a business

Treat the three signals as three separate opportunities.

For votes

Earn them honestly. Build something people like, explain it clearly, make it easy to try, and let people who have used it say so. Do not ask for votes in exchange for rewards, and do not use vote-trading groups. Not only does it break the rules, it leaves you with a number that means nothing.

For money

Decide if paid visibility fits your plan. Treat a placement as an experiment with a goal, a link you can track and a way to measure the result. The pricing page lists current options, and there is a minimum contribution for leaderboard entries, currently $5.

For facts

Keep your own details accurate and let verifiable data speak. Make sure your listing carries correct dates, your site is reachable and your profile is complete. Strong, checkable facts are the quietest and most durable form of credibility.

What you can do as a reader

When you see a ranking:

1. Look for labels. Which entries are paid?
2. Find the method note. What do the votes count? How are they protected?
3. Check dates on any figures.
4. Compare the signals. A product with strong votes and strong facts is more convincing than one with only one.
5. Make up your own mind by opening the listings.

If you cannot find labels or a method, assume the page mixes things together and treat the order as entertainment.

Why separation is good for everyone

For readers, separation gives clarity. For honest businesses, it means a good product can win votes without being outspent. For advertisers, it means a paid placement is seen for what it is, with no suspicion of cheating. For the platform, it means the trust that gives the board value is not traded away for short-term gain.

It also makes mistakes easier to correct. If a vote anomaly is found, it can be investigated within the vote system, without touching the paid placements or the factual columns. If a data source changes, the facts can be updated without disturbing votes.

An illustration

Consider three entries on a week's page.

• Entry A has the most votes, no paid placement and a modest link score.
• Entry B has fewer votes, a paid placement at the top and a strong link score.
• Entry C has the highest link score, few votes and no paid placement.

On a blended board, B might rise to first by combining all three. On a separated page, a reader sees A as the community favourite, B as the one that chose to advertise and C as the best linked. Each tells a different story, and the reader chooses which matters. That is a better result than a single misleading number.

A note on language

The words a page uses can undo the separation that its design achieves. Phrases such as "top rated", "most popular" and "best" are each shorthand for a particular signal, and each can be used loosely. "Most popular" should mean most votes. "Sponsored" should mean paid. "Highest rated" should mean a named measure. When you write about your own result, borrow the discipline: say "ranked third by community votes in the week of 3 March" rather than "a top product". Precision costs a few extra words and buys a great deal of trust, and it makes your claim easy to verify, which is the whole point of keeping the signals apart in the first place.

Frequently asked questions

Does separating signals make the page more complicated? Slightly, but clear labels keep it simple, and readers benefit from the honesty.

Can a paid listing also be popular? Of course. The two systems are independent. A product can earn votes and also pay for placement.

Where can I see this in action? The leaderboards page shows each system with its own label and method.