Five signals worth arguing about every week

Most weekly reviews try to cover everything and end up deciding nothing. We settled on five recurring signals, argued about them every week, and dropped anything that didn’t earn a place on the list. Here they are, in the order we check them.

1. Movement in usage

Not usage level, movement in it. An account sitting at the same activity for three months is not a signal; an account that dropped 40% in a week is. This should trigger a call within two business days, not a note for next month’s review.

2. Trials crossing the activation line

Every trial that crosses the activation behaviour this week goes on the list, and so does every one that was expected to and didn’t. This should trigger an owner assignment, not a percentage update.

3. Support concentration

One ticket from an account is noise. Four tickets from the same account in a week, especially one you’re counting on for expansion, is a signal that should trigger a customer success check-in before it turns into a churn conversation.

4. Expansion source

Where did this month’s expansion revenue actually come from: a handful of accounts, or a broad base? If it’s concentrated in a small number of accounts, that should trigger a conversation about what happens if one of them churns, not a celebration slide.

5. Silent accounts

An account that hasn’t logged in, opened a ticket, or replied to an email in three weeks is a signal even though nothing “happened.” Silence should trigger outreach, on a schedule, before the renewal date makes it urgent.

• Movement in usage — triggers a call within two business days.

• Trials crossing the activation line — triggers an owner assignment.

• Support concentration — triggers a customer success check-in.

• Expansion source — triggers a concentration-risk conversation.

• Silent accounts — triggers scheduled outreach before the renewal date.

Keeping the list at five

Every team we’ve talked to has tried to add a sixth signal at some point, usually after a bad quarter made one specific thing feel urgent. We let people propose a sixth, but only if they also nominate one of the existing five to retire for a month. Most of the time, the proposer realizes their new signal is a special case of an existing one, and drops it before the trade has to happen.

A sixth signal is free to propose and expensive to keep. Make someone pay for it by giving one up.

A few teams have asked whether the five signals should be the same for every business. They shouldn’t. A usage-based product cares more about movement in usage; a services-heavy business might replace it with billable-hours variance. What stays constant is the discipline of picking five and defending the list, not the specific five we use.

The support concentration signal caused the most internal debate when we introduced it. Support teams worried it would read as blame — “this account is generating too many tickets” — when the point was the opposite: concentrated tickets on an account you’re relying on for expansion is an early warning, not a scorecard entry.

Expansion source turned out to be the signal most often skipped in the first month, mostly because it required connecting billing data that nobody had gotten around to wiring up. Once connected, it was usually the fastest signal to produce an actionable finding, because expansion revenue concentration is rarely something anyone had actually checked before.

Silent accounts is deliberately the vaguest of the five, and that’s on purpose. It doesn’t point at a number crossing a line; it points at the absence of one. Teams that skip this signal tend to discover disengagement only when a renewal conversation forces the issue, which is later than anyone wants.

None of the five signals replaces judgment. They exist to make sure the argument starts from the same five places every week, instead of wherever the loudest voice in the room happened to look first.

Five is not a magic number. It’s the number our meeting can argue about properly in under an hour. Yours might be four or seven. What matters is that the list is short enough that everyone in the room can hold it in their head, and stable enough that the same five things get checked every single week.