7 Data Signals That Show Whether a Cryptocurrency Network Is Built for Everyday Transactions

Most of the numbers a blockchain publishes about itself look designed to be admired rather than used. Throughput records and market capitalisation say very little about whether a network can carry a bus fare or a grocery bill without embarrassing the person paying. 

Seven measurements come closer to answering that. They tend to sit in methodology footnotes and incident reports rather than press releases, and the networks making the loudest claims appear to fail at least three of them.

The Distance Between Raw Volume and Adjusted Volume

Begin with the gap between what a network reports moving and what it plausibly moved on behalf of a human being, because nothing else exposes as much so quickly. Visa’s onchain dashboard reports both figures, and the distance between those two lines is where a good deal of this industry’s credibility has gone to die. The adjustment strips out bot traffic, redundant internal contract cycling, exchanges shuffling value between their own wallets, and any address that has pushed more than a thousand transactions or ten million dollars through itself inside thirty days. What survives that filter resembles people paying people. What does not survive is usually the number in the headline.

Scale is where the confusion gets expensive. Stablecoin transaction volume across all chains ran to roughly $28 trillion in 2025, a figure quoted constantly and understood rarely. Set against the largest American wholesale settlement systems, it works out at under three business weeks of traffic, and netting out transfers between wallets owned by the same party pushes the honest total lower still. None of which makes the technology unserious. It does make the comparison against card networks, trotted out at every conference, close to unusable.

How Much of the Traffic Is Small

The second measurement follows straight from the first. Everyday commerce is small, repetitive and dull, so the share of surviving volume that is retail-sized tells you more than the total ever will. Transactions under $250 are counted as their own category, a blunt instrument that happens to be the right one here. A network settling vast sums in vast blocks may be a fine wholesale rail and still be hopeless at a coffee. Those look like unrelated capabilities, and treating them as a single achievement has muddled this argument for a decade.

For anyone assessing a chain professionally, the ratio worth recording is retail-sized count as a percentage of adjusted count, tracked monthly rather than sampled once. A single month tells you almost nothing. A rising line across two quarters is one of the few pieces of evidence that a network is being adopted for payments rather than for positions.

What It Costs to Get In and Back Out

Third, and almost nobody publishes it, comes the all-in cost of entering and leaving. The network fee is the smallest component and the only one anyone quotes. The spread charged when national currency becomes tokens, and charged again when tokens become national currency, routinely dwarfs it.

Retail on-ramps such as Paybis, where users buy litecoin with a card and convert holdings back into dozens of national currencies, put that arithmetic somewhere a person can see it, which block explorers do not. A sub-cent network fee sitting underneath a conversion spread of a percent or two is a payment system priced by the spread, whatever the ledger says.

The incumbent comparison is the one that keeps this honest in both directions. Sending money across a border still costs a global average of 6.36 percent of the amount sent, a figure that has drifted down for years without ever getting close to cheap. Against that benchmark a two percent all-in crypto route is a genuine improvement rather than a rounding error, which is precisely why the measurement has to include both ends. Quoting the network fee alone flatters the chain. Quoting only the incumbent’s headline rate flatters the incumbent. The practitioner’s version is a full round trip on a real corridor, priced on the day, including the exit.

How Far the Fee Can Move Before Tomorrow

Fourth is fee variance rather than average fee. Averages tend to be the wrong statistic for a merchant, who prices against the worst plausible day and not the median Tuesday.

Ethereum’s fee mechanism is instructive here because it was built specifically to address this, and only partly did. The base fee is set by the protocol from the preceding blocks, which makes the next block’s cost knowable in advance, and it can move by at most 12.5 percent per block. That sounds reassuring until the arithmetic compounds: eight consecutive full blocks roughly double it. Predictable one block ahead is not the same as predictable one hour ahead, and an hour is the horizon a shop actually operates on.

Two details from the same documentation deserve more attention than they get. The fee is charged whether the transaction succeeds or fails, which means a merchant integration with a badly set limit can burn money producing nothing. And wallets quote a recommended fee rather than a guaranteed one, so the number a customer sees at the moment of paying is an estimate that the protocol is free to overtake. Protocol reforms have improved predictability across several networks, though the underlying inefficiencies have proven stubborn, and congestion still converts into longer confirmations and higher costs exactly when volume peaks, which is exactly when a payment network is being asked to prove itself.

Anyone who has watched a checkout queue stall knows that a fee which is cheap on average and unpredictable in practice can be worse than a fee which is expensive and fixed. Predictability is the product. Cheapness is a bonus.

Who the Fee Ends Up Paying

Fifth, whether that fee pays for anything that lasts. Fees on these networks are not prices in the ordinary retail sense. They are incentive payments to whoever keeps the ledger honest, and the structure of that incentive goes a long way to deciding whether a chain survives a quiet decade as well as a loud one. This falls under token economy design and auditing, a discipline unglamorous enough that most coverage skips it entirely.

A chain paying its validators largely through token inflation is running a subsidy rather than a business. Subsidies end, usually at the least convenient moment. The question worth asking of any network is what happens to validator economics if the token price halves and stays there, because that scenario is neither exotic nor hypothetical. Where the answer is that operators would leave, the security and the fee stability both degrade together, and they degrade fastest under exactly the market conditions that send people looking for a payment rail in the first place.

The Machines Behind the Validator Count

Sixth, and the measurement that tends to separate the serious networks from the theatrical ones, is infrastructure concentration hiding behind the validator count. Validator counts get quoted because they are easy to quote. They are also far less informative than they appear.

On 12 August, around ninety Solana validators dropped out of consensus at the same moment after a routing fault at one hosting provider, pulling 28.83% of staked value offline against the 33.34% threshold at which the network stops finalising transactions. Blocks kept being produced. Few outside the industry noticed, and the public status page logged no incident at all. What makes the episode genuinely instructive is not the near miss but the aftermath: a further tranche of stake, hosted elsewhere, went dark in the same few minutes, and the analysts picking through it could not establish whether they were looking at a shared dependency or a coincidence.

Encouragingly, this is one of the few signals with a public measurement standard attached. Solana’s foundation now requires validators receiving its delegation to sit on a network operator holding under 25% of total stake, with a 15% ceiling on any single data centre, thresholds it sets out in its published delegation criteria alongside a plain statement that concentration inside one provider creates systemic risk. Those thresholds govern delegation eligibility rather than consensus, so they are a floor rather than a guarantee. They are still the most useful thing on this list for a practitioner, because the underlying concentration data is published at validators.app and can be checked in a few minutes by anyone. Before trusting a chain with settlement, look up which providers carry the top decile of stake and how much of it sits behind one of them. Hundreds of independent operators renting from the same handful of providers can behave like one operator with better branding, and a dashboard reporting the count alone will not distinguish the two cases.

The Version of This Problem That Moves Between Chains

A smaller relative of the same problem catches people out constantly, and it belongs here because it also masquerades as a solved issue. A token issued by the same company on two different chains is not one asset. It is two, sitting on ledgers that do not natively communicate, and sending value from a wallet on one to a wallet on the other can simply fail, with the funds unrecoverable.

Every bridge built to close that gap adds a component that can break, and the record is not encouraging. Chainalysis counted around $2 billion stolen across thirteen cross-chain bridge attacks during 2022 alone, accounting for most of the funds stolen in the sector that year, and its structural explanation still holds: a bridge concentrates the collateral backing the wrapped asset into one place, and that place becomes the target. Newer designs have improved on the worst of it. None has removed the basic shape of the risk. Fragmentation is a cost line, not a cosmetic one.

Whether the Money Ever Leaves

Seventh, and last, is whether value ever exits the ecosystem. This is the measurement that embarrasses everyone. A network can score respectably on the previous six and still operate as a closed loop, where tokens circulate among people who hold tokens and never touch a till. The number to hunt for is merchant settlement, and specifically what happens to a small business that starts accepting crypto payments rather than what happens to a trading desk. Most surveys of onchain activity keep finding the same thing, which is that trading dominates and payments trail some way behind.

There is a crude test for this that works better than it should. Take any network, find the wallets that receive funds regularly and spend them down to near zero on a rhythm, and count them. That pattern is what a wage looks like, or a supplier being paid. Wallets that only ever accumulate are savings, or speculation, or somebody’s cold storage. The ratio between those two populations arguably says more about a chain’s payment credibility than any throughput figure yet published, and very few projects report it, which is itself informative.

The Awkward Part About All Seven

Which is where the argument turns uncomfortable, and it should. Judged by these seven numbers, most networks available today fall short as general payment systems. Judged by whether people use them anyway, several are doing enormous business. Both statements hold at once.

The resolution seems to be that the seven signals describe suitability for a market with functioning banks, stable prices and cheap domestic transfers, and a substantial share of the world does not live in that market. If the local currency loses a third of its value in a year, a two percent spread stops being a defect. It starts being a discount. Adoption across emerging markets has rarely been driven by technical merit, and there is little sign of that changing. The networks winning on usage are not obviously the ones winning on these seven measurements, and the gap between the two lists looks wider than it did.

What to Run Before Trusting a Network With Settlement

So here is the practical version, and it takes an afternoon rather than a quarter. Pull the adjusted-versus-raw volume series and the retail-sized share for any chain being pitched to you, and record both monthly rather than once. Price a full round trip on a real corridor, entry spread and exit spread included, on an ordinary day and again on a volatile one. Check the concentration tables to see who actually hosts the stake. Ask what validator economics look like at half the current token price. Then ask the only question that carries commercial weight, which is whether anyone is being paid on this network rather than merely positioned on it.

Run those checks on any network sold as ready for daily use and it will fall short somewhere, and that shortfall will predict remarkably little about whether it succeeds. The data tells you what a network is capable of. It has much less to say about what people will tolerate, and people have historically tolerated a great deal in exchange for being able to move money on a Sunday. Which leaves the possibility that these are the right seven numbers aimed at the wrong question, and that the question worth asking is not whether a network is built for everyday transactions, but whose everyday it was built for.