Weekly and monthly 1:1s that ask different questions, not the same questions at different lengths. Then a way to run them so six months of them can be queried rather than remembered.
Set up your reports, take the two templates, and read how to make it recur. Nothing you type leaves your browser.
Who, and what number is theirs
The number is not decoration. If a person has one number that is genuinely theirs, the weekly agenda writes itself and there is nothing left to status-update. If you cannot fill this column, fix that before you fix the meeting.
How a 1:1 ladders up
A 1:1 is the bottom rung of the ladder. If the rungs above a person are missing, the meeting cannot fix it, and an hour a week gets spent discovering that repeatedly.
One rule per rung. The domain charters should add up to the north star, which is the Goal Ladder's job. Each person's number is a slice of exactly one domain, which is this one's.
Why bother with the format
What this buys you six months later
A 1:1 held in a shared doc is a conversation. The same 1:1 held as one ticket per person per period is a record, and a record can be queried. Everything below is a consequence of that one decision, and the Make it recur tab is how you get there.
What becomes queryable
Give every ticket the same fields and these stop being questions you have to remember.
Question
Because
What did I commit to and not do?
Commitments live in a field, with the person who made them and the period they were made in.
Which blocker has come up three times?
A blocker raised repeatedly is a systems problem being handled as a conversation.
Whose scores are sliding?
The monthly ratings become a series per person rather than a feeling.
Who has not had a monthly?
A missing ticket is visible in a way a missing meeting is not.
What goes in the review?
Six months of records written at the time, instead of a recency-weighted reconstruction the week it is due.
Reports that fall out of it
Once the tickets exist and share a shape, these are queries rather than projects.
Report
Cadence
Built from
Commitment close rate
Monthly
Action items marked done against action items created, by person and by you.
Repeat blockers
Monthly
Focus areas appearing in more than two consecutive weeks.
Motivation trend
Quarterly
The monthly ratings plotted per person over time.
Coverage
Weekly
Tickets created against tickets filled in. Low coverage is a signal about the format, not about the people.
Review pack
Half-yearly
Every ticket for one person over the period, in order.
The weekly exists to remove status updates
That is the whole design. If the number is visible and owned, nobody needs to narrate what happened. The half hour goes to what is off track, unclear, or waiting on a decision. Everything else is a document.
Prep is due the day before, not at the meeting. A 1:1 where both people read the doc for the first time in the room is a status meeting with extra steps.
Weekly template
—
The monthly is a different conversation, not a longer one
Most advice describes the monthly as the weekly with more depth. That framing is why monthlies decay into long weeklies and then get cancelled. The weekly is tactical. The monthly asks how the person is doing, which never fits in a week that has a blocker in it.
Monthly template
—
Make it a ticket, not a document
A 1:1 in a shared doc is a conversation. A 1:1 as one ticket per person per period is a record, and the difference is that a record can be queried. This is the part that pays off six months later.
Four recurring jobs, not one
Job
When
What it does
Create the weeklies
Start of the week
One ticket per report, titled by person and date, pre-filled with the weekly template and assigned to them.
Nudge the empties
Midweek
Checks which prep docs are still blank and messages just those people. Nagging everyone trains everyone to ignore it.
Create the monthlies
First of the month
One ticket per report with the monthly template, referencing the month just finished.
Audit the jobs
Monthly
Checks the other three are still registered and still producing tickets, against a written inventory.
Any scheduler that can create an issue will do. The important part is that they are separate jobs and that the fourth one exists at all, for the reason in the warning below.
The failure worth designing against
A scheduled job is two things stored separately: the instruction, and the registration that makes it fire. They fail independently and silently. The common version is that something disturbs the schedule, someone restores "the weekly routines", and the monthly quietly never runs again. Nothing errors. Nobody notices until a quarter later when there are no monthly records to review.
Keep a written inventory of every job and check it against what is actually registered, on a date in the calendar. Recover from the inventory, never from memory.
Take the four jobs as skills
One file per job, filled in with the reports and numbers from the Setup tab. Drop them in your agent's skills folder, or hand them to whoever owns the scheduler. They are plain Markdown, so they read as instructions even if nothing ever runs them.
The fourth one is the inventory check, and it exists because of the failure above. Three jobs that nobody audits become two jobs and then one, silently, and the first anyone notices is a quarter with no records in it.
The setup, as a written spec
—
The meeting is downstream of the goals
If the number column on the first tab was hard to fill in, the 1:1 is not the problem you have. Working out what each person is actually accountable for, and whether those add up to anything, is the sort of thing the practice does.