The Holdout Column

Automate the Coordination Ritual

Jun 14, 2026 · 4 min read

The standing cross-team meeting exists to stop campaigns from colliding. Replace it with shared events, attributes, and auto-suppression — and let the calendar slot go.

What the weekly sync is really for

Most lifecycle teams have a recurring meeting whose real job is collision-avoidance: making sure the promo team and the retention team and the product team don’t all message the same person in the same week. People call it alignment. Strip away the language and it’s a manual lookup — humans in a room, reading each other’s calendars, hoping nobody forgot to mention a send.

It works, sort of, at small scale. It does not survive growth. Add channels, markets, and teams and the meeting gets longer, the coverage gets thinner, and the failures move to the gaps between meetings — the campaign that launched Thursday that nobody flagged on Tuesday. A standing meeting is a workaround for a system that can’t coordinate itself. The fix isn’t a better meeting. It’s to make the workaround unnecessary.

Move the coordination into the data

Everything that meeting does by hand can be done by the system, if the system shares three things. Events: every send and every significant user action written to one place every team can read, so “did anyone message this person this week?” is a query, not a question you ask a colleague. Attributes: a contact’s current state — recently messaged, in an active flow, opted out of promos — stamped on the profile, not held in someone’s head. Auto-suppression: rules that read those events and attributes and quietly hold a message back when it would collide, fatigue, or break a frequency cap.

With those three in place, coordination stops being a conversation and becomes a property of the platform. The retention email simply doesn’t fire at someone who got a promo this morning, because the rule saw the event and suppressed it. Nobody had to remember. Nobody had to meet.

A recurring meeting is a workaround for a system that can’t coordinate itself. Build the system; cancel the meeting.

Rules don’t forget, and they don’t scale linearly

The reason to prefer the system isn’t that meetings are unpleasant. It’s that a rule is consistent and a room isn’t. The meeting depends on the right people being present, paying attention, and remembering to raise the one send that mattered; it degrades exactly when you’re busiest. A suppression rule applies identically at two campaigns a week or two thousand, at 3 a.m., in every market, without anyone in the loop.

I’ve run this at a scale where manual coordination simply wasn’t possible — roughly two thousand automations across seventeen countries. You cannot hold that in a meeting; you can only hold it in logic. Once the rules carry the coordination, the human time goes to the decisions worth a human — what to say, to whom, and whether to send at all — instead of policing overlap on a calendar.

Retire one meeting this quarter

Pick the standing sync that exists mainly to avoid collisions. Write down what it actually decides, line by line. For each line, ask: could a shared event, a profile attribute, or a suppression rule do this instead? Most can. Build those first — a global frequency cap and a “recently contacted” suppression rule alone will cover the bulk of it — then shorten the meeting, then drop it.

Keep a meeting for the things that genuinely need judgment; automate the ritual that was only ever bookkeeping. The aim isn’t fewer meetings for their own sake. It’s a system that coordinates itself so reliably that the calendar slot becomes obviously redundant — and the people who used to spend it cross-checking sends spend it on work only people can do.

This is the thinking. The Diagnostic is where I point it at your stack.