A holder count needs a dust line

To measure dust in a token’s holder count, count addresses above a stated balance cutoff alongside all addresses with a positive balance. The unadjusted figure answers how many addresses hold some of the token; the adjusted figure shows how many hold more than the chosen minimum. Neither count tells you how many people own it, since one person can control several addresses and one address can represent a service or contract.

What does a token holder count measure?

A holder count measures addresses with a positive current token balance. Etherscan’s token-page guide describes its holder figure as the total number of addresses holding a token. For ERC-20 tokens, the standard defines balanceOf to return an address’s balance and requires a Transfer event for transfers, including transfers of zero tokens. That means counting transfer recipients alone can overstate holders: an address may have received tokens and later sent them all away, or received a zero-value transfer.

A data query can rebuild balances by adding incoming transfers and subtracting outgoing ones, then counting balances above a threshold. Dune’s public ERC-20 holder query illustrates this approach and uses a minimum balance of 0.1 token in one example. That is a query choice, not a universal definition of dust; the same token amount can mean very different things across assets. When comparing token prices across markets, Poocoin’s guide to comparing prices across trading pairs covers the pricing detail that a holder-count calculation only touches on.

How should you set a dust threshold?

Set a cutoff that matches the question, then report it with the result. A token-unit cutoff is easy to reproduce, but says little about economic value when tokens have different prices. A dollar-value cutoff is easier to interpret across assets, but depends on a price source and timestamp; for a thinly traded token, a quoted price may not represent a price at which a small holder could actually sell.

For a useful comparison, keep the raw count and show adjusted counts at more than one cutoff. Dune’s example demonstrates why the threshold belongs in the method: changing it changes which addresses qualify. A small set of cutoffs also makes it easier to see whether growth comes from balances accumulating beyond dust or from a growing tail of tiny balances.

• Record the chain, token contract or mint, and the block or snapshot time.
• Define whether a qualifying balance must be greater than zero or at least the cutoff.
• State the cutoff in token units or dollars, and name the price source and timestamp if using dollars.
• Keep the raw and adjusted counts together so readers can see what the filter removes.

How can you tell whether holder growth is meaningful?

Compare the adjusted count over time using the same chain, token, snapshot method, and cutoff. If the raw count rises while the adjusted count stays flat, more addresses have small balances, but the headline increase does not show that more addresses crossed the stated threshold. If both rise, the data supports a narrower claim: more addresses hold at least the chosen amount. It still does not establish how many independent owners there are.

Check the largest balances and any known contract or service addresses before treating the count as a picture of individual ownership. Etherscan lists addresses, while ERC-20 defines balances at the address level; neither definition converts addresses into people. Also check that the data reflects current balances rather than a cumulative list of recipients. The ERC-20 standard’s transfer events help reconstruct balances, but a simple recipient count cannot account for tokens later sent away.

The most useful signals to watch are the raw count, counts at consistent dust cutoffs, and whether balances are moving above those cutoffs over time. If a dollar-denominated threshold is used, track its price input and timestamp too. Together, these show whether reported growth reflects more nonzero addresses, more addresses above a meaningful balance, or both.