The standard response to a scattered-information problem is a platform: an account, a cloud database, a team plan, a welcome sequence, and eventually a pricing page with three tiers where the one you want is the middle one.
For a personal research habit, that is a heavier answer than the question deserves.
Accounts add friction at exactly the wrong moment
Every login is a barrier between noticing an artist and recording them. Sync delays, session expiries and password prompts all arrive during the two seconds when you were willing to act. Two seconds is the entire budget, and it is easy to overspend.
Shared does not mean better
A team database is genuinely useful for teams. For one person keeping their own research, it adds permissions, seats and the mild self-consciousness of knowing that a half-formed shortlist is visible to colleagues.
Your research is not particularly anyone else’s business
A roster of artists reveals what you are planning, who you rate, and what you can afford. It is ordinary professional material, and it is also a reasonably complete map of your intentions. There is no strong reason for it to sit on someone else’s server.
Local-first is not a philosophy
It is mostly a practical position: fewer moving parts, no outage, no migration when a service is acquired, nothing to export in a hurry when pricing changes. The privacy benefit is real and largely a side effect of the simpler architecture.
The honest version of the trade
Local storage means no sync between machines, and that is a genuine cost worth stating plainly rather than hiding. For most people keeping a research roster, on one working machine, it is a cost worth paying to skip the account entirely.
And nothing to unsubscribe from
Which, for a tool whose job is to reduce the number of things demanding your attention, seems like the minimum standard.