CoachMe · Coach mobile · July 2026 · Exploration

Rounds — a coach's pocket.

Drawn only from what the coach mobile app is contractually for: work the task inbox, read the priority ranking, handle risk flags, close the loop with a message. Full builder parity stays on web.

One answer to one question — how does a coach spend four minutes between clients? — committed to a single interaction model rather than hedging. Two other directions (Ward, a full-roster glance grid; Line, tasks-as-messages) were explored alongside it and retired in the July 2026 cleanup; git history holds them.

Update: The screens below are drawn on Priority view, which the scope contract defers to Post-V1 — so they mock a surface that isn't in scope. Rounds has been rebuilt on the coach task inbox with a proper app shell: Rounds, revision B →

Selected direction

Rounds

The app is a session, not a screen. It deals you one client at a time and ends when the deck is empty.

A coach who opens a list decides what to do next forty times a day. Rounds removes the choosing. The priority projection already ranks every relationship and names a primary action — so the app plays that ranking as a sequence and asks one question per card: handle it, defer it, or skip it.

You start a run, you finish a run. There's a beginning and an explicit end, which is the thing a list can never give you. The reward is the empty state.

Character: editorial, high-contrast, near-black. Serif display numerals against a mono spine. It feels like a briefing document, not a dashboard.

Reads
coach-priority ranked list, top reasons, primary action
Writes
/viewed, /snooze, resolve with linked message
Bet
Coaches want to be told the order, not to pick it
Risk
Hostile to browsing — "just look up one client" is a detour
7:41▮▮▮ ⌁ ▊
Thu 16 Jul · Morning run
123

clients need you today

Three are critical. Estimated 8 minutes at your usual pace.
3
Critical
5
High
4
Due today
Start rounds
Snoozed until tomorrow · 6
01 · Open — the run, not the inbox
7:43▮▮▮ ⌁ ▊
Critical · 4 of 12
Nadia
Haddad
Active · 14 weeks · High-touch package
+70 Reply gap, 4 days You asked about her knee. No answer.
+30 Adherence 48% this week Down from 91% — dropped Tuesday
+10 Check-in due today
Logs · last 14 days
Message Nadia
Snooze 1hOpen profile
← skip · hold for history
02 · The card — score made legible
7:51▮▮▮ ⌁ ▊

Rounds clear

Everything that needed you this morning has been handled or deferred.

9
Handled
3
Snoozed
9:12
Minutes
Next run when Farah replies, or 4:00 PM.
03 · The end — a list can't do this

Why Rounds

The priority projection already does the ranking work and already names a primary action — Rounds spends that asset instead of re-displaying it, and the finite session is the one thing a coach can actually fit between clients. It's also the cheapest to build: one card component, three actions, no dense grid to make accessible.

One thing worth carrying in from the retired directions: an inline composer. Nearly every coach task resolves in a message — a nudge, a follow-up, a risk check — so replying should close the card without leaving the run.

The density idea that Ward was built on isn't dead either, it's just on the wrong device. A full-roster glance grid belongs on the coach web dashboard, where scrolling costs nothing.