Bridges onchain volume is a great metric but it should not be taken for granted. A few month ago, I shared how a few addresses amplified Orbiter volume with non-organic usage: There is a similar situation with two other bridge providers appearing on the top of the @DefiLlama bridge page 1- First case on @mesonfinance 90% of their bridge volume is from two addresses doing back-and-forth transactions between Arbitrum and BSC. Those addresses are sending each other 0.5 or 1 BTC approx. 500 times per day resulting in $30m of daily volume. This does not seem to relate to any kind of organic rebalancing activity. On Arbitrum: On BSC: In May alone we count $450m of volume done from the two addresses and only $18m from any other addresses ($18m). 2- Second case on @hyperlane: Described by @PrimordialAA in this tweet: This address-0x1eeaaf572d1dd107054bedb65e1bdf05556a43ae- is doing back and forth $20,000 transactions with USDT on Celo and openUSDT on Unichain approximately 2500 times per day resulting in $40-50m of daily volume in the last few days Overall, among the 6700 addresses that have used Hyperlane in the last 9 days, 6 (0.09%) represents 93% of the volume. The one above is #6. #3, #4, #5 presents similarities in activity (0x734ada6c042f97f4780468383085a4e3f527bb08, 0x6a688e505720eb06857aa6e7ce31afad7bf51afb, 0x2deee85bc05055b070515f041148384d3d1dd49a) #1 and #2 (0xbba1938ff861c77ea1687225b9c33554379ef327, 0x4a8149b1b9e0122941a69d01d23eae6bd1441b4f) seems to be linked to openUSDT protocol ( and also same back and forth patterns than the other 4 - even if the activity seems to have changed in the last few days from $20,000 transactions to more random amounts. 3 - Conclusion Always question volume data even on a chad website like DeFillama and question even more when there is a large variation/increase of this volume. I do not think that the protocol themselves controls those addresses. However advertising volume growth while knowing about those limitations does not help creating trust. @DefiLlama in order to improve checking if the volume is legit, I propose two things. Creating a volume concentration ratio i.e. how many addresses and the proportion in terms of total addresses that are responsible for 50% and 90% of the protocol's volume. The smaller the number of addresses and proportion (i.e. below 0.1% and 1%), the more likely the volume is inorganic. Something harder to do -> looking at the pattern of volume and number of transactions for addresses with the more various number of transactions and the biggest volume. If a lot of addresses are sharing the same stats it is likely a farming strategy and volume is non-organic. This is what we unveiled with @Orbiter_Finance.