Solana token metrics explained
The nine figures on a standard Solana token dashboard, what each one counts, which inputs it inherits, and the order to read them in so a strong number cannot hide a weak one.
Filed in Metrics
What each token metric actually counts
A token page shows you nine or ten numbers and implies they describe the same thing. They do not. Each figure has its own definition, its own inputs, and its own way of being wrong, and almost none of them are audited before they reach your screen.
This desk takes one metric per page. We state the formula, name the inputs it inherits, show what the number looks like when someone has an incentive to move it, and give you the check that settles the question against chain data.
The console linked above is a third-party product. It appears here because readers who study reported volume usually want to see what software that produces volume actually does with wallets, venues and scheduling. We do not audit it and we take nothing from your use of it.
rule 01
A metric is a definition, not a fact
Two dashboards can print different volume for the same token in the same hour and both be internally consistent. The difference is what each one decided to count.
rule 02
Every number inherits a weaker one
Market cap inherits circulating supply. Liquidity inherits a price. If the inherited input is soft, the headline figure is soft, however precisely it is displayed.
rule 03
The check is usually cheap
Most figures on a token page can be reproduced from mint state, pool accounts and transaction history in a few minutes. We publish the procedure, not the conclusion.
The nine figures on a standard token page
Read them in this order. Supply-derived numbers first, because everything priced depends on them; flow numbers next; distribution last, because it explains the other two.
01 Trading volume Sum of the notional value of trades in a window, counted per venue and often double-counted across them. 02 Transaction count A tally of swap events, blind to size, so one whale and one thousand dust trades look similar. 03 Market cap Circulating supply multiplied by last traded price, where circulating supply is an editorial judgement. 04 Fully diluted valuation Total supply multiplied by the same price, ignoring whether the supply can reach the market. 05 Holder count The number of token accounts holding a non-zero balance, not the number of people. 06 Unique wallets Distinct addresses that traded in a window, which measures addresses rather than participants. 07 Liquidity The value sitting in pool reserves right now, usually reported as both sides added together. 08 Top holder share The percentage of supply held by the largest accounts, before any of them are classified. 09 Buy / sell ratio A comparison of buy-side and sell-side activity that changes meaning depending on what is being counted.
Field guides
Four longer pages that carry most of the method: how the figures relate, how they are inflated, and how to record what you checked.
Verifying dashboard numbers
A repeatable procedure for taking one figure from an aggregator and reproducing it from mint state, pool accounts and transaction history, including what to do when the numbers refuse to agree.
Filed in Verification
Top holder concentration
A top-ten percentage is only meaningful once you have classified each address. Pool accounts, program vaults and exchange wallets all sit in that list and none of them mean the same thing.
Filed in Signals
Building a token scorecard
A scorecard template built from checks you can perform yourself, with pass conditions written before you look at the token, so the result is a record rather than a rationalisation.
Filed in Verification
Three sections, one discipline
Definitions come first, then the patterns people read into them, then the procedures that turn a reading into a record.
Metrics
Each page takes a single figure from a token dashboard, states its formula, names the inputs it depends on, and shows what the same figure looks like when the inputs are manipulated.
Open metricsSignals
Distribution, flow direction and wallet counts are pattern data, not verdicts. These pages set out what each pattern can support, what it cannot, and the alternative explanation you must rule out first.
Open signalsVerification
Procedures for reproducing a dashboard figure from primary sources, recording what you checked, and scoring a token on evidence you gathered yourself rather than on a badge someone else applied.
Open verificationThe four-step check we apply to every figure
It is deliberately short. A check you will not repeat is not a check, and the point is to run this on a token before you form a view rather than afterwards.
Name the definition
Write down what the number claims to count before you look at its value. Volume over which window, on which venues, counted once or twice.
Find the input it inherits
Trace it back to the softest thing it depends on. Usually that is circulating supply, a price source, or an address label somebody chose.
Reproduce it from the chain
Pull mint state, pool reserves or transaction history and recompute. Agreement to a few percent is normal; an order of magnitude is a finding.
Record the gap
Note what you could not verify. An honest scorecard has blanks in it, and the blanks are usually where the risk lives.
Where the numbers come from
Nothing on a token dashboard is measured by the dashboard. Every figure is assembled from a small number of upstream sources, and knowing which source produced a number tells you most of what you need about its reliability.
Reported volume is an assembled figure
An aggregator reads swap events from indexed transactions, prices each swap against a reference, sums them per venue, and applies its own rules about which venues to include and whether a route through two pools counts once or twice. Every one of those decisions changes the headline.
Automation changes the picture again, because software can produce genuine on-chain trades on a schedule. Those trades are real transactions with real fees, and they are indistinguishable from organic ones at the event level. That is exactly why the derived ratios on this site matter more than the headline totals.
If you want to see what a volume console actually controls, the one we reference in examples exposes venue selection, wallet counts and pacing.
Explore volume toolingThird-party software, opens in a new tab. Named for illustration, not endorsed.
How this desk works
Pump Metrics Desk is an editorial project, not an analytics vendor. We do not run an indexer, we do not sell data, and we do not publish live market figures because we cannot stand behind them a day later.
What we publish
Definitions, decompositions and procedures. If a page shows arithmetic, the inputs are chosen by us to make a mechanism visible and are labelled illustrative. Protocol facts, such as the nine decimals on native SOL or the structure of an SPL mint account, are stated as facts because they are checkable.
What we do not publish
Live prices, market capitalisations, ratings, holder counts for named tokens, or any figure that would be stale before you read it. We also do not publish reviews, star scores or testimonials, because we have no defensible way to produce them.
How to argue with us
Every method page is written so you can run it yourself and get a different answer. If you do, the contact page takes corrections, and the editorial policy explains what we change and how a correction is recorded.