Raw exchange volume is the wrong metric to trust on its own

I’ve been looking at exchange-ranking data recently, and the thing I keep coming back to is simple: Raw 24h exchange volume is a bad standalone metric. That doesn’t mean exchange rankings are useless. It means reported volume needs context. The questions I care about are: Does reported volume line up with visible liquidity? Does the order book look deep enough for the claimed activity? Does the venue have real user traffic? Do the most active pairs look natural, or do they look mechanical? If I tried to route size there, would the book behave anything like the ranking suggests? This is where current ranking systems are more useful than old raw-volume tables. CMC is one example. Its current exchange-ranking methodology does not rank venues only by self-reported 24h volume. It uses traffic, liquidity, trading volume, confidence in the legitimacy of reported volume, and qualitative factors. At the market-pair level, CMC also uses reported volume, Liquidity Score and Web Traffic Factor as ranking inputs, with a Confidence Indicator that labels markets as High, Moderate or Low confidence. That still doesn’t make CMC “truth.” I wouldn’t treat any aggregator that way. But it’s a lot more useful than a plain volume table, because at least you’re getting liquidity, traffic and confidence signals in the same view. The way I’d use it: 1. Reported volume is an input, not the answer. High volume with weak liquidity or weak traffic is a red flag. 2. Confidence labels are filters, not proof. High confidence doesn’t mean clean. Low confidence means be careful. 3. Liquidity matters more than rank position. A top-ranked exchange still needs a book that can actually handle your size. 4. On-chain flows are useful, but limited. Exchange-wallet deposits and withdrawals can sanity-check activity, but they are not the same as centralized exchange trading volume. Most CEX trading settles internally. 5. Long-tail exchanges deserve extra skepticism. The smaller the venue, the noisier the traffic and liquidity signals can get. 6. Execution still requires checking the book. Rankings are useful for discovery. They are not a substitute for spreads, depth, slippage, and actual routing tests. My current view is: Raw volume is noisy. Rankings that include liquidity, traffic and confidence signals are useful filters. None of them replace due diligence. For practical use, I’d use CMC exchange rankings as a first-pass filter, then verify the venue with order-book depth, spread behaviour, market-pair quality and, where possible, exchange-wallet flow context. The biggest mistake is treating any aggregator ranking as final truth. The second biggest mistake is dismissing every ranking because raw volume used to be worse. The data is better than it was, but it still needs to be used carefully. Sources: – CMC exchange rankings: https:// coinmarketcap. com/rankings/exchanges/ – CMC exchange ranking methodology: https:// support.coinmarketcap. com/hc/en-us/articles/360052030111-Exchange-Ranking – CMC market-pair ranking and Confidence Indicator: https:// support.coinmarketcap. com/hc/en-us/articles/360043675052-Market-Pair-Ranking-Confidence-Indicator – CMC Liquidity Score: https:// support.coinmarketcap. com/hc/en-us/articles/360043836931-Liquidity-Score-Market-Pair-Exchange Curious how others here sanity-check exchange rankings. Do you use rankings as filters, or do you mostly ignore aggregator data and go straight to the order book?