Streamlining the Purchase Account picker

The current picker: one native dropdown with every active account for the provider, sorted default-first then alphabetically. Open it and scroll — this is the pain being solved.

Replace the dropdown with a search input. Type any fragment of the email, nickname, or owner and the list narrows instantly with match highlighting. Arrow keys + Enter select. The single cheapest change that makes 260 accounts feel like 5.

Keep a picker list, but order it by usefulness: Default → Used by you this week → Used by the team this week → everything else A-Z. In practice fulfillers reuse a handful of accounts, so the right one is almost always in the first screenful. Timestamps show freshness.

Don't make the fulfiller pick at all. The sheet opens with the best account already selected (provider default → your most recent for this provider → least-recently-used) and shows why. Zero clicks in the common case; "Change" expands a search only when needed. Pairs naturally with the provider auto-select that now reads the lot.

Most of the 535 accounts belong to other people. Segmented tabs scope the list to Mine (accounts you created or use) first, expanding to Team (used in the last 30 days) or All only when needed. Defaults to Mine, remembers your last scope.

Accounts aren't interchangeable: some are mid-purchase, on a velocity cooldown, or missing a stored card. Track state per account and surface it as chips, with the list pre-filtered to Ready. Stops the "picked an account that couldn't check out" retry loop entirely.

A ⌘K-style overlay: big search box, rich result rows (provider, owner, last used, notes), recent picks on top, full keyboard flow. The sheet shows just the chosen account; the palette handles the choosing. Feels fast at any account count and doubles as account lookup anywhere in the portal.

Adding an account today means leaving the completion flow for a separate sheet. Instead: "+ New account" expands inline, pre-fills the provider from the lot, and suggests the next free alias from the team's plus-addressing scheme so nobody has to invent an email or check for collisions. Save and it's selected immediately.

Attack the 535 directly: pin the workhorse accounts to the top (star), and auto-archive anything unused for 90 days into a collapsed section that never renders in pickers unless expanded. The active list shrinks from ~260 to the ~25 that actually get used; nothing is deleted.

Treat accounts as a managed pool instead of a menu. The sheet deals the top 3 candidates with the reason each was chosen (least-recently-used, no active rental at this venue, stored card on file). One tap takes a card; a purchase automatically sends that account to the back of the rotation. Spreads volume across accounts by design, which also protects them.

Remember which account bought at this venue (and metro) last time, per provider, and default to it with the evidence shown. Great when accounts are effectively "assigned" to markets — Wrigley orders keep landing on the Wrigley account without anyone remembering which one that is.

"Find" = speed of locating an existing account · "Add" = speed of creating one · "Scale" = stays usable as accounts grow · "Errors" = prevents picking an unusable account.