The XRP Ledger crossed 5 billion lifetime transactions this week, but new research shows the milestone rests on a very narrow base. In August, 793 sender accounts produced 93.2% of the ledger’s 81.56 million transactions, and 767 of those accounts were classified as machines.
The data comes from Bitquery, a blockchain analytics firm that examined more than 5 billion XRPL transactions dating back to 2013. Its August breakdown found that 26 exchange hot wallets contributed another 1.1% of activity on top of the bot share. After filtering out automated traffic and spam, Bitquery estimated that real, human-scale payments made up just 0.8% of ledger traffic, a share it said has stayed roughly flat since 2018.
The finding landed badly with parts of the XRP community, which has long pointed to ledger throughput as evidence of adoption. The 5.06 billion validated transactions figure has circulated for years as a talking point in the XRP-versus-Bitcoin debates, and this week’s milestone gave it fresh air. The Bitquery data undercuts the way that number gets used.
One Address, 13 Million Transactions
The concentration at the account level is extreme. A single address generated almost 13 million transactions in August, including 12.79 million decentralized exchange orders, yet only 882 of those orders settled into actual trades. Across the whole ledger, 2.97 million trades settled from 12,153 accounts, and nearly half of all active accounts sent a single transaction and then went quiet.
DEX trading bots accounted for 48.1% of all August transactions, with dust-sending accounts adding another 23%. The remaining machine share came from NFT bots, spam accounts, and other automated traffic. Bitquery noted that its research only covers activity recorded directly on the ledger. XRP trading on centralized exchanges does not show up as individual ledger transactions, so the bot-heavy picture does not capture every form of XRP use.
Volume Up, Users Down
A second dataset points in the same direction. Evernorth Research found that XRPL order-book trading volume rose 79% year over year in the second quarter even as the number of accounts initiating those trades fell about 40%, from 1,864 daily trading accounts to 1,111. Volume per account climbed from 1,072 XRP to 3,217 XRP. Evernorth said part of the shift reflects professional traders taking a larger share of flow, which is not necessarily a bad sign on its own.
Capital on the network did grow. The average value held on XRPL reached $4.26 billion, a six-quarter high, and balances of Ripple’s RLUSD stablecoin averaged $539 million, up 642% year over year. But daily transacting accounts and new account creation both fell about 25%, and 18% of settled trades never touched XRP itself.
Ghost Chain Criticism Meets Schwartz Defense
The findings drew an immediate reaction from XRP critics. A commentator known as Scam Daddy highlighted the numbers on X, writing that 793 accounts did 93% of August transactions and calling the network a ghost chain. He added that less than 1% of traffic was real human payments and that it had been this way for eight years. The claim spread quickly through crypto social media.
Ripple CTO Emeritus David Schwartz pushed back, asking on X: “If it were more expensive and fewer people did low-value things on it, would that somehow make it better?”
Schwartz argued that low fees are a feature, not a liability, since cheap transactions let users run both useful and low-value activity without penalty. He called the criticism odd, noting that a network being inexpensive to use is not evidence that it fails at its job. His response drew its own replies, with some users pointing out that a network built on bot traffic has little to show for its throughput.
What the Numbers Actually Mean
The dispute matters because transaction counts are often cited as a headline adoption metric. The Bitquery and Evernorth data suggest throughput on XRPL is a weak proxy for real adoption. A ledger can post record totals while the number of people actually using it shrinks, as long as a few hundred automated accounts keep firing.
That pattern is not unique to XRPL. Bot traffic and wash-style ordering show up on most low-fee chains. Analysts made a similar observation about Solana’s token minting surge earlier this month, where more than 263,000 new SPL tokens were created in a single day with almost none of them trading. The difference is that XRPL’s human share, at 0.8%, is unusually low and has stayed low for eight years, through the NFT boom, the stablecoin rollout, and the addition of an automated market maker.
Why It Matters for Ripple’s Institutional Push
For Ripple, the optics are awkward at a time when the company is pushing institutional use cases for the ledger, from RLUSD settlement to tokenized assets. Institutional flows do show up in the reserve data, and the six-quarter high in average value held suggests real money is parked on the network. The Besu client Ripple competes against in enterprise deployments is in use at Citi, BNY Mellon and other banks, so the adoption argument matters commercially.
The gap between value stored and transaction activity is the part both sides will keep fighting over. Critics read it as proof that the ledger is a ghost town with a few machines inside. Defenders read it as evidence that activity has professionalized, with fewer but larger participants. The August data supports both readings, which is probably why neither side is going to drop the argument soon.