New Manager Dealing with an Underperformer on a Remote Team at Google

Quick Answer

This is not a morale problem; it is an evidence problem with a manager credibility cost. In a Google remote team, the mistake is to treat a slipping IC as a personality puzzle when the room is waiting for a performance judgment. The right move is a short written reset, a tight check-in cadence, and a decision path that does not drift for 90 days because the manager wants to avoid discomfort.

TL;DR

Not sure what to bring up in your next 1:1? The SRE Interview Playbook has 30+ high-signal questions organized by goal.

Who This Is For

This is for the first-time Google manager who inherited a remote IC, saw the misses start in docs, deadlines, or ownership, and now feels the team quietly reorganizing around the gap. It is also for the manager who keeps hearing, “They are trying hard,” while the artifacts say otherwise. If the person is liked, senior, or time-zone disconnected, the pressure to delay gets louder. That delay is usually the wrong call.

Is this actually an underperformance problem or a role-fit problem?

It is usually a role-fit problem before it becomes a character problem. In one remote calibration, the manager kept describing the engineer as “very thoughtful,” while three launch slips and two missing written handoffs sat in the packet like evidence nobody wanted to name. The room did not need more empathy. It needed a judgment.

The distinction matters because managers confuse effort with output when the evidence is ambiguous. Not a motivation issue, but a mismatch between what the role demands and what the person can consistently deliver. Not a remote-work issue, but a clarity issue exposed by remote work. The person may be busy, responsive, and pleasant, and still be failing at the job that matters. That is the trap.

At Google, I have seen this pattern get misread because the team protects smart people for too long. A high-status IC can sound credible in meetings and still be structurally unreliable in execution. The organization then starts paying a hidden tax: extra review cycles, manager shadow work, and downstream distrust. That tax is bigger than the discomfort of calling the gap early.

The key signal is not whether the person works hard. The key signal is whether their work can be trusted without the manager stitching it together. If the answer is no for more than one cycle of feedback, this is no longer a coaching note. It is a performance decision in slow motion.

What should I do in the first 30 days?

The first 30 days should produce written clarity, not emotional relief. I have watched new managers spend their first two weeks trying to “build rapport” with the underperformer, only to discover in week three that the team still cannot tell what good looks like. That is backward. The manager is not there to process the situation. The manager is there to define it.

Start with one clean reset conversation and follow it with a written recap the same day. Name the exact misses, the expected standard, the deadline for change, and the next check-in date. Use days, not feelings. If the gap is in execution, give a 7-day or 14-day checkpoint with visible artifacts. If the gap is in ownership, require one owner, one deliverable, and one written status note that someone else can read without a meeting.

The judgment here is simple: the first month is about proving whether the person can recover inside the role. Not a motivational talk, but a documented reset. Not a vague “let’s improve,” but a narrow definition of what improvement means. Not a private hope, but a manager-owned record that survives calibration.

A good 30-day sequence is boring on purpose. Week 1, define the gap. Week 2, inspect the evidence. Week 3, see whether the person can hold the line without the manager’s rescue. Week 4, decide whether the trend is real. If the pattern is still fuzzy after 30 days, the manager has probably been too polite to make a call.

How do I manage a remote underperformer without turning into a surveillance manager?

You manage remote underperformance with artifacts, not frequency. The worst version of this job is a manager who starts pinging every morning, then calls it accountability when it is really anxiety. That is not control. It is noise.

Remote work changes the evidence standard. In person, a manager can infer a lot from proximity. Remotely, inference is weak and written proof matters more. Not more pings, but better receipts. Not constant availability, but predictable outputs. Not “Are you online?”, but “Where is the draft, the handoff, the decision log, and the blocker?” That difference is everything.

I have sat in reviews where the manager thought the issue was communication style, but the actual problem was unreliable closure. The underperformer could talk fluently in 1:1s and disappear when the work needed to be finished asynchronously. In a remote team, that pattern is expensive because it forces other people to become reconstruction agents. The team starts spending cognitive energy on translation instead of execution.

The psychological principle is simple: people optimize for what the manager inspects. If you inspect chat volume, you get chatty people. If you inspect written handoffs, ownership notes, and on-time deliverables, you get a cleaner signal. That is why remote management has to be evidence-based. A manager who relies on vibes in a distributed team is already behind.

The practical standard is visible work, not visible busyness. If the person says progress is happening, it should show up in a document, a merged change, a customer-facing draft, or a decision memo. If it only shows up in conversation, it is not progress. It is narration.

When do I move to a formal performance plan at Google?

You move when the same miss survives multiple documented resets and the person still cannot own the gap. In a calibration conversation, the room usually goes quiet when a manager keeps asking for “one more week” after two written check-ins have already failed. At that point, the issue is no longer whether the person can improve. It is whether the organization is willing to keep pretending the pattern is unclear.

A formal plan is not a punishment. It is the point where the company stops letting ambiguity protect everyone. The manager who delays this step to avoid being seen as harsh usually looks less humane, not more. People can tolerate a hard decision. They do not tolerate a manager who drifts for months and then acts surprised.

At Google, the cleanest judgments I have seen came after the manager had already documented the expectation, the support, the misses, and the follow-up date. That is what makes the discussion credible. Not a surprise move, but a consequence. Not a sudden personality shift, but a stable record. Not an emotional reaction, but an organizational one.

The counter-intuitive truth is that faster clarity is usually kinder. It gives the underperformer a real chance to recover or exit with dignity. It also protects the rest of the team from the message that performance standards are optional for people with enough seniority, charm, or internal sponsorship.

If you wait until the evidence is overwhelming, you waited too long.

What do I say in calibration and to HR?

You say the expectation, the evidence, the support, and the recommendation, in that order. In a real debrief, the strongest manager I saw brought a one-page summary, not a speech. The page listed the role expectation, the concrete misses, the feedback already given, and the exact next step. The room moved quickly because there was no room for spin.

Calibration is not a place for adjectives. It is a place for decisions that can survive scrutiny. If you show up saying someone is “not senior enough” or “not thriving,” you invite a political discussion. If you show up with three specific misses, two written resets, and one recommendation, you invite a judgment. That is the difference between being persuasive and being correct.

The better script is plain. “The role requires independent closure on X. Over the last 30 days, the person missed Y and Z after written feedback on dates A and B. Support was provided through check-ins and clear examples. My recommendation is to move to the formal path unless we see a measurable change by the next checkpoint.” That is not aggressive. It is legible.

In an HC-style debate, managers lose when they speak in abstractions. They win when they can separate effort from impact, friendliness from reliability, and potential from current role fit. The organization does not need a story that makes everyone comfortable. It needs a story that can be defended.

Preparation Checklist

Prepare around evidence, cadence, and decision rights, not around being nicer.

Write the role expectation in one paragraph before you talk to the person. If you cannot describe the standard cleanly, you are not ready to manage the gap.
Gather three concrete examples of the miss, each with a date, an impact, and a prior ask. The conversation gets weaker every time it relies on memory.
Set a 7-day or 14-day review cadence and define what artifact you expect to see at each checkpoint. Remote management needs receipts, not reassurance.
Write the follow-up summary the same day as the conversation. If it is not documented, it will be remembered as a vibe.
Align with your skip manager and HR partner before the situation becomes a surprise in calibration. Political safety comes from early alignment, not from urgency.
Work through a structured preparation system (the PM Interview Playbook covers performance conversations, evidence logs, and Google-style calibration narratives with real debrief examples). Use it for the same reason strong managers use templates: it keeps the narrative clean.
Decide in advance what “improvement” means and what failure looks like. Ambiguity is how managers accidentally extend bad situations for another quarter.

Mistakes to Avoid

These are the mistakes that make the manager look vague, reactive, or politically naive.

Mistake 1: Treating the issue as a vibe. BAD: “I just don’t feel great about the pace.” GOOD: “The person missed the handoff on Tuesday, then missed the revised deadline on Friday after a written reminder.”
Mistake 2: Managing remote work like a surveillance problem. BAD: “Send me updates every morning so I know you are on it.” GOOD: “Send one written recap by 4 p.m. PT with blockers, decisions, and next steps.”
Mistake 3: Waiting for the person to self-correct because they are nice. BAD: “Let’s give it another month and see.” GOOD: “We already completed two documented resets, so the next step is a formal performance path unless the next checkpoint shows a clear change.”

FAQ

The right answer is to move on evidence and sequence, not emotion or hope.

Should I wait if the person is popular on the team? No. Popularity buys patience, not exemption. If the miss is already visible to peers or stakeholders, delay only makes the manager look political.

What if the underperformer says they are overwhelmed? That may be true, but overwhelm is not a free pass. If the scope is too large, resize it. If the output is still weak after that, the issue is capability, not bandwidth.

Can a remote underperformance issue just be bad communication? Sometimes. If the person improves once the expectations are written down and the artifacts become reliable, it was a communication problem. If the same misses continue after clear correction, it was performance all along.

Ready to build a real interview prep system?

Get the full PM Interview Prep System →

The book is also available on Amazon Kindle.