New Manager at Google: Handling an Underperformer on a Remote Team

TL;DR

You do not let a remote underperformer drift; you compress the timeline, document the gap, and force a decision. The first mistake new managers make is confusing empathy with ambiguity, and remote work rewards ambiguity. The right response is a short reset window, one written standard, and a clean escalation path if the work does not change.

Who This Is For

This is for the new Google manager who inherited a remote engineer, PM, or analyst who looked stable under the previous lead and now looks ragged in your first month. Maybe the person still speaks well in meetings, but their docs arrive late, their handoffs are thin, and the team is quietly compensating. The real problem is not that you lack a script; it is that you are being forced to make a judgment in public before you have built a private evidence trail. I have watched new managers lose credibility in the first quarter by waiting for the pattern to become obvious to everyone else. By then, the team has already adjusted around the person, the manager has already softened the standard, and the eventual correction looks personal instead of operational.

How fast should I act when a remote underperformer starts slipping?

Act in days, not quarters, but do not act on irritation. In a Q3 performance debrief I sat in, the hiring manager kept saying the remote person was “probably just overloaded,” until we laid three artifacts on the table: a late doc, a missed partner update, and a planning decision that had to be redone twice. The room changed immediately because the issue was no longer temperament; it was pattern. That is the first mistake to avoid: not a coaching problem, but an operating rhythm problem. Remote underperformance is usually a breakdown in visible commitments, not a dramatic collapse.

The right move is a short reset window with a written bar. I would set the conversation within 48 hours, then give one week to show a different operating cadence: on-time drafts, proactive status updates, and no surprises on dependencies. Say this plainly: “I am not asking for reassurance. I am asking for a change I can inspect next Friday.” Then be specific: “By Friday at 4 p.m., I need the draft, the stakeholder update, and the decision log.” That is not harsh. It is the minimum standard for a manager who wants to stay credible.

📖 Related: Meta L5 Refresher Grant vs Google L5 Refresher Grant: Which Pays More?

What signals separate a bad week from a real performance problem?

The real signal is consistency of judgment, not one missed deadline. The first counter-intuitive truth is that remote underperformance often shows up as dependence, not sloppiness. The person is present in meetings, but they cannot close the loop without repeated nudges, they escalate after the window has closed, and they use “alignment” as a substitute for ownership. I have seen new managers blame workload when the actual failure was sequencing. The person was not drowning; they were not deciding.

Look for artifacts that survive your opinion. If the doc is thin, the handoff is vague, the follow-up is late, and the adjacent partner has to reconstruct the work, you are not looking at a bad week. You are looking at a judgment problem. That is the second counter-intuitive truth: the problem is not the output, but the judgment signal behind the output. Ask, “Walk me through the decision you made, the options you rejected, and the tradeoff you were optimizing for.” Strong performers answer cleanly. Weak performers hide behind activity. At Google, the people who last are the ones whose reasoning is legible when the manager is not in the room.

What should I say in the first conversation?

Say the gap plainly, or the conversation will become theater. In one manager conversation I watched, the lead spent ten minutes softening the message because the employee was remote and “seemed stressed.” The employee left relieved and unchanged. Two weeks later, the manager was angry because nothing moved. The problem was not kindness. It was clarity. Not a kindness problem, but a clarity problem.

Use language that names the standard and the timeline. “I need to be direct: your last two deliverables missed the bar on completeness and timing. This is not a personality issue. I want to see a specific reset over the next seven days: the draft by Tuesday, stakeholder notes by Wednesday, and no surprises.” Then add the hard line: “If you think the bar is wrong, say so now. If you agree, I need your plan by the end of this call.” The third counter-intuitive truth is that clarity feels less humane in the moment, but it is the only thing the person can actually use. The problem isn’t your answer — it’s your judgment signal. If your standard is vague, the employee hears permission.

📖 Related: 1on1 System vs Google Manager Check-In: Which Builds Better Teams?

How do I keep the team aligned without gossip or damage?

You protect trust by changing process, not by exposing the person. In a remote team, silence becomes rumor faster than weak performance becomes visible. If you say nothing, the team invents a story. If you overshare, you turn the problem into public punishment. I have seen both mistakes. The better move is narrow operational correction: smaller ownership slices, shorter deadlines, one point of contact, and less ambiguity in handoffs. That is not secrecy. It is privacy discipline.

Here is the script I use: “I am simplifying ownership on this workstream so we can reduce risk. Please route questions through me for now, and I will keep everyone updated on the milestones that matter.” That keeps the team focused on execution instead of speculation. It also prevents the underperformer from becoming a public project. In a remote environment, trust is built when people see that the manager can intervene without turning the team into a tribunal. Not transparency, but privacy discipline. The team does not need your drama. It needs a stable operating system.

When should I escalate to HR or start a PIP?

Escalate when the reset window has produced clean evidence of failure, not when the conversation feels uncomfortable. The second counter-intuitive truth is that a PIP is not the beginning of accountability; it is the paper trail after accountability already failed. If you have given two explicit checkpoints over 14 days, written the bar, reviewed the artifacts, and the gap remains, stop pretending the issue is ambiguous. At that point, you are no longer gathering sentiment. You are documenting reality.

Do not wait for a perfect confidence level. That is how new managers stall until the team has lost patience and the skip-level is already hearing a different story. I would align with HR or your manager once you know what standard you will enforce, not after you have emotionally committed to a conclusion. The cleanest sentence is: “Here is the standard, here are the dated artifacts, and here is the consequence if the standard is not met.” That is the evidence threshold. Not patience, but evidence threshold. If you cannot explain the issue in a way another manager could read without your voice, you are not ready to escalate.

Preparation Checklist

Your preparation should reduce ambiguity before you ever schedule the conversation.

• Write the performance gap in one sentence. If you need three sentences, you do not yet know the problem.
• Pull three recent artifacts: the last deliverable, the last status update, and one example of partner feedback.
• Separate output, responsiveness, and collaboration. Remote underperformance often touches all three, but you need to know which one is failing first.
• Decide your reset window before the meeting. Seven business days is usually enough to reveal whether the person can change the operating rhythm.
• Script the exact line you will use when the person pushes back, because they usually will: “I am describing the standard, not debating the existence of the gap.”
• Work through a structured preparation system (the PM Interview Playbook covers Google-style calibration, manager debriefs, and difficult-performance cases with real debrief examples).
• Align with your skip-level or HR partner on the evidence threshold before you speak, so you are not inventing the bar in the middle of the conversation.

Mistakes to Avoid

The worst errors are all forms of delay dressed up as leadership.

• BAD: “Let’s give it another week and see if it improves.” GOOD: “Here is the reset window, here is the standard, and here is the next checkpoint.”
• BAD: “I think the person is just going through something.” GOOD: “I have observed repeated misses in timing, completeness, and follow-through, and I am addressing the work.”
• BAD: “I don’t want to make it feel punitive.” GOOD: “I want the person to know exactly what success looks like, what failure looks like, and when I will decide.”

FAQ

1. Should I tell the team the person is on notice? No. Tell the team what changes operationally, not what you think of the person. People need clear ownership, deadlines, and escalation paths. They do not need a performance narrative that turns into gossip.

2. Do I go to HR before I give feedback? Usually no. Give the feedback first, document it, and then align with HR if the gap persists. HR is there to help you execute a clean process, not to substitute for judgment you have not yet made.

3. What if the person is strong in meetings but bad at execution? Treat that as a real performance problem, not a communication style difference. In remote work, execution is the job. If the person can speak well but cannot produce reliable artifacts, the team is still carrying the risk.amazon.com/dp/B0GWWJQ2S3).

Related Reading

• Navigating Your First PIP Process: Amazon vs Google for New Managers
• Google PMM vs Amazon PM: Role, Salary, Culture Compared