Base growth dilemma: everything is right and the users are still leaving
Thejaswini M A
Photo by Chopper, Foresight News
A few days ago, I read about a concept in Japanese philosophy: the basho. The rough translation is "place," but the philosopher Doro Nishida gave it more than one geographical location, and more like a situation: a place in which everything becomes its own. In other words, human beings are not incidentally present, but are shaped by their place. Today I'm going to use this theory to read Base。
Last month, the number of active addresses fell to 18 months of a new low. In reflecting on this phenomenon, I realized that Base built a place but never created conditions for things to grow and form。
In 2023, when Coinbase launched Base, it was rare to have a faith in encrypted primitive circles. It is believed that it can finally solve the oldest problem of the Ethera: infrastructure is everywhere but there is no real user. Coinbase has a unique advantage in its ability to distribute 100 million users with unparalleled power. As soon as the gate was opened, the user was waiting outside。
For some time, that confidence seemed to have been tested. Base grew faster than all Layer2 before. In October 2025, its total lock value (TVL) amounted to $5.6 billion, and the revenue from handling fees was beyond reach in the entire L2 area. So, in September 2025, Base confirmed the issuance of a token, as if it meant an experiment that would necessarily succeed. Yes, a place is becoming a scene。
Then the user left。
See more intuitively: Base's active address returned to July 2024. The distribution of tokens is expected to meet the air-dropping party ' s needs: get the last payment and go。
Base's bet on the creator economy in 2025 did not work either. The core is the Zora protocol, which by default monetizes content. By the end of the year, 6.52 million creators and content tokens were issued through Zora, of which only 17,800, or 0.3 per cent, remained active throughout the year. The remaining 99.7 per cent are no longer consulted。
Base's live address peaked at 1.72 million in June 2025. By March 2026, there were only 458,000 left, and the higher point fell by 73 per cent. After the announcement in September 2025 that Base was considering issuing currency, the active address fell by 54 per cent within a short period of six months, which meant that speculative funds were completely out of business。
The sociology Ray Oldenburg has studied: what brings people back to a place without pay. He called it third space, like bars, barbershops, city squares. They are not efficient production space, but give people a reason for return that is unrelated to incentives. At the core is that the will to return cannot be created by human beings, but can only grow naturally from the long-term possibilities offered by the site. The design site for the encrypted money industry was designed to extract users and then to wonder why no one was left behind。
this is the location of the no-basho: people pass by, take what they need, and then leave because leaving is costless. there is no identity here, no ability that cannot be repeated for three weeks elsewhere, and nothing makes departure a loss. is there a unique relationship in this chain? we never built anything like that, did we
you can't build the grounds with financial incentives. incentives can, of course, drag people into the door, but they cannot be kept. the legacy of desire must come from the possibility of long-term motherhood in the place. todochiro nishida called it "the logic of the place", which means how the relationship field shapes the things that emerge. the encryption industry designed a field for extraction, and finally found it surprising that only the extraction was born。
Brian Armstrong publicly stated that Base App is now focused on becoming Coinbase's own hosting, trading version。
The vision of the social and creative person that was designed to create social viscosity and to allow users to establish a protective identity has disappeared. In terms of data, this is a rational decision, but it is also acknowledged that this vision has never really materialized. Base has a location, and it is now focused on serving past users, because that is what it can provide。
A chain, a track
Base is the most visible microcosm of the entire L2 mode。
Since June 2025, the overall usage of L2 by small and medium-sized has declined by 61 per cent. Most of the chains outside the first three have been turned into zombie chains: active to the point where they cannot be stopped, but cold to the point. The L2 ratio of daily life to L1 fell from 15 times in mid-2024 to 10-11 times today. Most of the new L2s collapsed at the end of the incentive cycle. The whole L2 ecology is cooling, not just Base。
The Rollup-centred road map used to be a set of theories for users: reducing the cost of participation and the influx of users into ecological formation and compounding growth. This year, the ETA Foundation released a 38-page vision paper on the future direction of the ETA. And the largest L2 activity bottomed and left OP Stack, and the second largest L2 growth stalled。
Reducing the cost of entry does not create the conditions for things to take shape. The industry solved the "entry" problem, but naturally thought that the sense of belonging would follow. It does not appear automatically, as a sense of belonging is not a function that can be accessed online。
Farcaster is the closest product of the encrypted world to building a basho. Because a particular group of people built a specific culture on it: developers shared their work, discussed the Ethera, and developed a view of each other over the months. This takes time and competitors cannot replicate it with higher rewards. Friend.tech tried to do the same thing with an incentive, a week top, a month dead. The same mechanisms have not developed culture. The difference is not in the product, but in whether someone stays long enough to make something really form。
What's keeping people
The chain of retention in the cold winter is not based on better incentives。
Arbitrum's daily address peaked at 740,000 in June 2024, and now stands at 157,000, with a similar drop of 79 per cent. Both chains are falling, but the bottom logic is completely different。
Base's users go online for the transaction, and they leave when the volume drops. Arbitrum users are not affected by the level of fees, and the correlation between the number of users and the fee income is almost zero. Base attracts tourists, while Arbitrum somehow retains users。
Hyperliquid was able to stand firm because of its unique trading experience, where communities formed identities that did not exist elsewhere. Incentives for tokens are almost irrelevant and have become part of their behaviour and identity. Things shape users, who in turn shape things。
The encryption industry is still optimising "how to make people come," and the "how to make things work" question is never remembered until data collapses, never considered at the beginning of the chain design。
I believe that Base has the strongest distribution capacity in history and could have solved this problem better than any chain。
now it's a trade application. that's the right direction, but it's what more than 40 products do already. the transaction application does not produce the field (basho), but only the session: the user enters when the transaction is required and leaves after completion。
To be truly a successful application, a continuous link is needed. Users are required to establish a relationship between each visit, so that the next visit feels like a return rather than a mere arrival。
The transformation of Armstrong is largely based on the lessons learned by Base from data. Social layers, creators' economies, chain identities, things that should have turned Base from being "used" to "sheltered" needed patience, and the system did not reward patience。
It's not just a place to trade. The whole L2 narrative is rooted in the fact that chains can be the infrastructure around which people build their lives. If the strongest distribution of L2 in the history of encryption is ultimately willing to do a faster Coinbase, then the narrative itself cannot stand。
in the view of tohiro nishida, the deepest ground (basho) is where the border with the site begins to melt. you can't separate "who you are" from "where you are shaped." it sounds abstract, but putting it on a public chain means that a user cannot imagine a financial life after leaving a chain; that all the kits of a developer are based on an ecosystem; and that their identities are almost impossible to exist elsewhere。
AS FAR AS I KNOW, SOMETHING LIKE THIS HAS NEVER BEEN BUILT ON ANY L2. IT MAY NOT BE BUILT UNDER AN INCENTIVE SCHEME AT ALL。
Even if you hold 100 million potential users, as long as there's nothing left to keep, you'll end up empty. Base understands now。
