What 1 million requests taught us about follow-ups
The second nudge works better than the first. The third rarely helps. More from our request data.
Amara Nwosu
·
·
3 min read

Follow-ups are the quiet engine of operations. A request gets stuck, someone nudges, and the work moves again. Most teams do this by hand and by instinct: a message on Tuesday, another on Thursday, an escalation on Friday.
Metrivo has now sent follow-ups on more than a million requests, which gave us a rare chance to check those instincts against data. Some of what we found confirmed what experienced ops people already knew. Some of it surprised us.
The second nudge does most of the work
First reminders work less often than you’d expect. Plenty of people see the first nudge, mean to act on it and get pulled into something else. The second nudge, sent after a sensible gap, is consistently the one that gets the most responses.
The third nudge rarely helps. By that point the request is usually stuck for a reason a reminder can’t fix: the person is the wrong owner, a detail is missing or the request needs a decision they can’t make. Sending a fourth or fifth message just trains people to ignore the agent.
So after two nudges, the agent changes tactics
Instead of a third reminder, Metrivo asks a different question. Is this the right owner? Is something missing? Should this go to someone who can decide? That one change did more for stuck requests than any amount of extra nudging.
Timing beats frequency
When a follow-up lands matters more than how many are sent. Three patterns stood out across the data:
Mornings win. Nudges delivered early in the recipient’s working day get answered noticeably faster than ones that arrive late in the afternoon.
Respect time zones. A reminder that lands at 7pm local time is mostly ignored until the next morning anyway, and it annoys people.
Avoid the Friday cliff. Requests nudged on Friday afternoon often slip to the following week. Nudging on Thursday is far more effective.
Specific beats polite
“Just checking in on this!” is the most common follow-up in the world, and one of the least effective. Follow-ups that say exactly what is needed perform much better: “Can you confirm the cost center for Dana’s laptop? It’s the only thing blocking the order.”
The best follow-ups share three traits. They name the single action needed, they explain what it unblocks, and they make that action possible in one click or one reply.
Where follow-ups should go
A reminder in a busy channel is easy to miss. A direct message to the owner is much harder to ignore, but sending every nudge privately hides the request from everyone else. The pattern that worked best was a quiet first nudge in a direct message, then a visible update in the team thread if the request was still stuck.
Escalation is not a punishment
Teams often hesitate to escalate because it feels like telling on a colleague. We found the opposite: when escalation is automatic, consistent and polite, people stop taking it personally. It becomes part of the process rather than a judgment about them.
What we changed in Metrivo
Based on this data, Metrivo now sends at most two reminders, times them to the owner’s working hours, writes each one around the specific missing piece and, after the second, switches to diagnosing why the request is stuck. Teams that adopted the new defaults saw fewer reminders sent overall and faster turnaround at the same time.
The rule of thumb
Nudge twice, then ask a better question. More reminders won’t unstick a request that is stuck for a reason.
Keep reading.
More notes from ops teams on routing, approvals and the follow-ups nobody wants to send.
Untangle your ops this week.
Most teams are live in an afternoon. The first 200 runs are on us.



