Holder counts are one of the few new things a public chain gives you about a stock-tracking instrument. Nobody publishes the number of people holding a particular ETF in real time. For these tokens, anybody can count.
The number is real. What it means is narrower than it looks, in four ways.
One: holders are wallets, not people
One person can hold a hundred wallets, and some do. A wallet is free to create and carries no identity.
A holder count is therefore not a count of people. The reverse also happens: one exchange wallet can hold for many people. Treat it as a measure of wallets, which is what it literally is.
Two: much of the supply is not held by anybody
A large share of a token's supply typically sits in protocol contracts: liquidity pools, bridges, lending markets, the issuer's own treasury.
Those are addresses, so they count as holders, and they usually hold a lot. A token where the top address holds most of the supply is very often a token whose liquidity pool holds most of the supply, which is a completely different fact from one person holding most of the supply.
This is why our holder tables split wallets from protocol contracts rather than ranking everything in one list. Putting a liquidity pool at the top of a list headed "largest holders" invites exactly the wrong conclusion.
Three: a count is only as good as its method
This site counts a holder only when it holds a balance above zero. On Solana, it reads every token account for the mint at one moment. On the EVM chains, it publishes a figure only when the balances it counted match the token's own total supply, within 0.5 percent. A token that fails that check shows no figure.
Every figure is stamped with the time it was read.
Four: it says nothing about liquidity
A token can have a lot of holders and no depth, or few holders and a working pool. Holder count measures distribution, and tradability is a separate question.
If what you want to know is whether a quoted price means anything, pool depth answers it and holder count barely does. We wrote about that in why the token price differs from the share.
What the number is good for
Two things.
Checking a token is not deserted. A token with a handful of addresses and a tiny supply is telling you something, and it is usually that almost nobody is involved.
Watching concentration over time. The share held by the ten largest addresses, and the split between people and protocols, are published with a read date.
What we will not do with it is rank anything. There is no leaderboard of most-held tokens here, no list of what is moving, and no score. An ordered list of tokens by holder growth is a trading signal delivered by layout, whatever the disclaimer under it says, and we would rather publish the numbers and let you do your own arithmetic.
One thing we deliberately do not print
Explorers often show self-assigned labels on addresses. Those labels are set by whoever controls the address, which makes them attacker-controlled text.
We do not render them. A contract's own name can appear after checks, and a name bought through a name service never does. A label somebody chose for themselves is not a fact about them.
Where to look
Each stock page shows its holder distribution where this site has read it, and the whale pages show how concentrated each token is and whether holders are people or protocol contracts, listed alphabetically. They name no person. Methodology explains how each figure is read, and the glossary defines holder, supply and transfer.