Management by Proximity Is Dead: The People Science of Leading Hybrid and Distributed Teams

The old management model ran on watching people work. The evidence says the problem was never location — it was structure.

For most of the last century, a manager could walk the floor and feel the state of the team: who came in early, who looked stuck, who was huddled with whom. None of it was measurement, but all of it was information — ambient, free, and constant. Then the floor emptied, and an entire management model quietly lost its sensory input. The managers didn’t get worse. They went blind, and many responded the way blind systems do: with anxiety, surveillance, and gut-feel evaluation of whoever happened to still be visible.

Key takeaways

1. Hybrid work, done deliberately, does not damage performance: a randomized controlled trial of 1,612 employees published in Nature found two days per week of home working had no significant effect on performance reviews or promotions, while cutting quit rates by one-third. [2]
2. The foundational WFH randomized experiment found home-based workers performed 13% better than their randomized office counterparts over nine months, with attrition cut roughly in half. [1]
3. The same experiment documented proximity bias inside an RCT: despite higher performance, home workers’ promotion rates conditional on performance fell roughly by half. [1]
4. Observers spontaneously infer character traits from passive face time — “dependable” from being seen during business hours, “committed” from being seen after hours — without any information about actual output. [6]
5. Microsoft’s 2022 Work Trend Index quantified the trust gap it named “productivity paranoia”: 85% of leaders said hybrid made it harder to be confident people were productive, while 87% of employees reported they were productive. [3]
6. The evidence-backed replacement for proximity is deliberate structure: explicit priorities, structured 1:1s, and outcome-based evaluation — 81% of employees told Microsoft that manager help prioritizing workload was important, yet fewer than a third said they had ever received clear one-on-one guidance. [3] [4]
7. Hybrid is not an edge case to wait out: Gallup’s tracking shows hybrid is now the dominant arrangement for remote-capable U.S. employees. [5]

What did managers actually lose when the office emptied?

Not productivity. Visibility.

The pre-2020 management model was built on ambient proximity, mostly without anyone naming it. Seeing people work stood in for knowing what they produced. Hallway collisions stood in for status updates. Long hours at a desk stood in for commitment. A manager could go weeks without an explicit conversation about priorities because the office itself was a low-resolution monitoring system running in the background.

When that background system vanished, organizations discovered how much of their “management” had been passive observation rather than deliberate practice. The tell is what happened next: instead of building explicit structure, many leaders reached for surveillance software, mandated badge swipes, or fell back on what Cadence’s research hub calls management by archaeology — reconstructing what happened to a person from fragments in scattered tools after the outcome is already decided. Proximity’s disappearance didn’t create that failure mode. It exposed it.

Does remote and hybrid work actually hurt performance? What the RCTs say

This is the rare management question with genuine randomized-controlled-trial evidence, and it comes from the same research program: Nicholas Bloom and colleagues at Stanford, running experiments inside Trip.com (formerly Ctrip), a NASDAQ-listed Chinese travel firm large enough to randomize.

The 2015 experiment (fully remote, call center). Call-center employees who volunteered to work from home were randomly assigned to work from home or the office for nine months. Home working produced a 13% performance increase — about 9 percentage points from more minutes worked per shift (fewer breaks and sick days) and 4 from more calls per minute in a quieter environment. Home workers reported higher work satisfaction, and their attrition rate roughly halved. [1]

The 2024 experiment (hybrid, professionals). A second RCT randomized 1,612 university-educated employees — engineers, marketing, finance — to hybrid (working from home Wednesday and Friday) or full-time office for six months. Hybrid work reduced quit rates by one-third and improved job satisfaction, with no significant effect on performance reviews, performance grades, or promotions over two years of follow-up. Retention gains were largest for non-managers, women, and employees with long commutes. [2]

Two details from the 2024 study deserve their own sentence. First, the outcome measures were the company’s real performance reviews and promotion decisions — not self-report. Second, managers surveyed before the experiment predicted hybrid would hurt productivity; after living through it, they revised their estimates upward. [2] The anxiety was measurable, and it was wrong.

The reframe that matters: the best causal evidence available says hybrid and remote arrangements are not inherently a performance problem. Which means that when a distributed team underperforms, the actionable hypothesis is usually a management-structure problem, not a location problem — and structure, unlike geography, is something a leader can fix this quarter.

Why did so many leaders predict the opposite?

The most instructive data point in the Nature study is not the null effect on performance. It is the before-and-after on managers’ beliefs. Managers surveyed before the experiment expected hybrid work to hurt productivity; after six months of running it, the same managers revised their estimates upward, into positive territory. [2]

That reversal is worth sitting with, because it explains the last several years of return-to-office whiplash. The managers were not being irrational. They were reasoning correctly from the wrong instrument. Under the proximity model, a manager’s confidence in the team was built from observation — and when observation went away, confidence went with it, regardless of what output was doing. The absence of visible work felt like the absence of work.

Three mechanics make this predictable rather than a personal failing:

• Managers experience inputs; systems record outputs. A manager sees effort, presence, and responsiveness in real time, but sees results only at review checkpoints. Remove the real-time channel and the manager is flying between checkpoints with no instruments.
• Asymmetric error costs. For an individual manager, wrongly trusting an unproductive remote employee feels riskier than wrongly distrusting a productive one — the first failure is attributable, the second is invisible. Paranoia is the locally safe error.
• Anecdotes fill the vacuum. Without structured signal, one unresponsive afternoon generalizes into a narrative about “people slacking at home,” even when the team’s shipped work says otherwise.

The fix is not exhorting leaders to trust more. It is giving their confidence something better than line-of-sight to run on — which is exactly what the structured practices below provide. Trust built on visible outcomes survives distribution; trust built on visibility never could.

What is “productivity paranoia” — and why is it so corrosive?

Microsoft’s 2022 Work Trend Index put numbers on the leadership anxiety: 85% of leaders said the shift to hybrid work made it hard to have confidence that employees were being productive. Meanwhile, 87% of employees reported that they were productive. Microsoft named the gap “productivity paranoia.” [3]

Paranoia is the right word, because the anxiety is about missing evidence, not negative evidence. Leaders lost the input their confidence used to run on — sight — and most organizations never replaced it with anything better. The corrosive part is what paranoia does downstream: monitoring theater, presence-performing employees who calibrate their green status dots instead of their output, and evaluation drifting toward whoever is most visible. Gallup’s hybrid-work research lands on the same operating conclusion from a different data set: as flexibility increases, so does the requirement for managers to be explicit — about priorities, about progress, about handoffs — because implicit coordination no longer happens by osmosis. [4]

And this is not a transitional condition to wait out. Gallup’s ongoing tracking shows hybrid has settled in as the dominant arrangement for remote-capable U.S. employees, with only a small minority fully on-site. [5] The proximity model is not coming back at scale. The replacement has to be built.

Is proximity bias real, or just a talking point?

It is one of the better-documented perception effects in workplace research, and it predates the pandemic.

Elsbach, Cable, and Sherman studied what they call passive face time — the time an employee is merely seen at work, without any interaction or information about what they are doing. Observers spontaneously inferred traits from it: being seen during normal business hours produced inferences of “dependable” and “responsible”; being seen outside normal hours produced “committed” and “dedicated.” The inferences were spontaneous — observers made them without being aware of it, from presence alone. [6]

Now put that mechanism next to the RCT evidence. In the 2015 experiment, the home workers were measurably more productive — and their promotion rates, conditional on performance, still fell roughly by half relative to office peers. [1] That is proximity bias with a randomized control group: same firm, same jobs, better measured output, worse career outcomes, and the only systematic difference was visibility.

For evaluation, the implication is blunt. Any performance process that runs on manager impressions will silently import a visibility premium. In a hybrid organization that premium is not just unfair — it is selectively unfair, because Gallup and the Nature RCT both find that flexibility is disproportionately valued by women, caregivers, and long-commute employees. [2] [4] A promotion process that rewards face time is a retention risk aimed at exactly the people hybrid work helps keep.

What replaces proximity? The deliberate-structure playbook

The research points at a consistent answer: everything proximity used to provide implicitly must now be provided explicitly. Four practices carry most of the load.

1. Explicit priorities, written down. Microsoft found 81% of employees said it was important for managers to help them prioritize workload — and fewer than one-third said their manager had ever given clear one-on-one guidance on it. [3] That is the largest documented gap between what distributed employees need and what they get. The single highest-leverage act for a hybrid manager is stating, in writing, what matters this week and what can wait — because in a distributed team, an unstated priority is indistinguishable from no priority.

2. Structured, recurring 1:1s. The hallway is gone; the 1:1 is the channel that remains. It only substitutes for ambient context if it is engineered to:

• run on a shared agenda both sides contribute to before the meeting;
• carry memory — prior action items and commitments resurface instead of evaporating;
• cover priorities and blockers explicitly, since neither will surface in a hallway that no longer exists;
• survive busy weeks, because for a remote employee a cancelled 1:1 is not a rescheduled meeting, it is a week of zero manager signal.

See How to Run Effective 1:1s for the practice in detail.

3. Goals visible between conversations. When progress lives in a shared system rather than a manager’s line of sight, “how is the project going” stops being a surveillance question and becomes a coaching question. Visible goals also protect the employee: they create a running, timestamped record of contribution that does not depend on being seen making it.

4. Outcome-anchored evaluation. The countermeasure to proximity bias is documented evidence: goals met, work shipped, recognition received, commitments kept — so calibration debates start from a record rather than from who was most visible. Gallup’s operating advice for hybrid management points the same direction: explicitness about progress and handoffs, not inference from presence. [4] For the fuller treatment, see Managing Remote and Hybrid Teams Without Visibility.

Where does Cadence fit?

This is precisely the gap Cadence is built for — it is, in effect, a management operating plane for teams that can no longer manage by line of sight. Labeled by current availability:

• Live today: structured 1:1 workspaces (agendas, action items, meeting notes) and goal/OKR tracking with check-ins, goal trees, and at-risk views — the explicit-priorities and visible-goals layers above.
• Preview: AI-generated 1:1 summaries and manager coaching prompts; survey-based listening.
• Roadmap: 9-box talent calibration (Coming Q3) — the room where outcome-anchored evidence would formally feed promotion discussions — and audio capture.

The AI’s role in all of it is bounded the same way the research would bound it: the system prepares the manager, remembers the commitments, and surfaces the patterns a distributed manager can no longer see ambiently. It does not rate people from keystrokes, and it does not decide who gets promoted. AI develops managers, not replaces them — and in a hybrid organization, “develops” specifically means replacing a manager’s lost ambient awareness with structured awareness, rather than replacing their judgment.

One distinction matters enough to state plainly: structure is not surveillance. Monitoring software watches inputs — screens, keystrokes, status dots — and thereby rebuilds the worst property of the old model, evaluation by observed busyness. Deliberate structure documents agreed outcomes — priorities, goals, commitments — and evaluates against those. The first tells employees they are not trusted; the second tells them exactly how to succeed from anywhere. The evidence on productivity paranoia suggests the surveillance route deepens the trust gap it was meant to close. [3]

How do you know if your organization is still managing by proximity?

A short diagnostic for People leaders. Each “yes” is a place where the old model is still running without its inputs:

• Can employees state their manager’s top priority for them this week — in writing, from a system, not from memory? If not, priorities are still traveling by hallway, and the hallway is gone.
• When a 1:1 gets cancelled, does anything notice? If skipped 1:1s are invisible, the organization’s only remaining structured channel to remote employees can silently go dark.
• In your last promotion round, what evidence was actually cited? If the file is impressions rather than documented outcomes, the face-time research says visibility is being graded, whether anyone intends it or not. [6]
• Do office-based employees get promoted faster than equally rated remote peers? This is measurable in most HRIS data, and the RCT evidence says the burden of proof sits on the process, not the remote employees. [1]
• Is leadership’s answer to distributed-work anxiety more monitoring or more structure? Monitoring signals distrust of people; structure signals distrust of the old model. Only one of these is warranted by the evidence. [2] [3]

None of this requires believing remote work is universally superior. It requires only the conclusion the data forces: presence is no longer a valid proxy for contribution, so contribution has to be made legible on its own.

The honest objection: doesn’t the office actually matter?

Steelman the skeptic, because parts of the skeptic’s case are well-founded:

These RCTs come from one company class — a large tech firm with strong measurement systems and, in 2015, volunteer call-center workers doing individually measurable work. The hybrid study kept everyone in the office three days a week; it says nothing about fully remote knowledge work. Mentoring, onboarding, and the apprenticeship of junior employees may genuinely degrade without co-location. And “no significant effect on performance” is not the same as “proximity had no value” — maybe the office three days a week is exactly what preserved performance.

Most of this deserves agreement, not rebuttal:

• Generalization is a real limit. Two RCTs at one class of firm do not prove hybrid is costless everywhere. The 2015 study’s subjects volunteered to work from home; effects for employees compelled into an arrangement may differ. What the studies do establish is an existence proof with causal identification: hybrid and even fully remote arrangements can run at equal or better performance — so location alone cannot be the default explanation for underperformance. [1] [2]
• The hybrid finding is a finding about hybrid. The Nature result covers roughly two days at home, not zero days in office. Bloom’s own reading is that deliberately structured hybrid — coordinated in-office days, not randomized emptiness — is where the evidence is strongest. [2] Organizations extrapolating to fully distributed models are extending past the strongest data, and should say so.
• Onboarding and mentoring concerns survive. Neither RCT was designed to measure the multi-year development of junior employees. A defensible hybrid design concentrates co-location where tacit learning happens — which is itself an argument for deliberate structure rather than for nostalgia.
• What the objection cannot rescue is proximity as an evaluation instrument. Whatever the office is worth for collaboration, the face-time research shows that presence-based impressions measure visibility, not contribution. [6] The skeptic can keep the office; the promotion committee still has to give up the window seat as evidence.

What Cadence should not claim

• Do not claim “Cadence makes hybrid teams X% more productive” — Cadence has no customer outcome data, and says so plainly.
• Do not claim the RCT results as Cadence results. Bloom and colleagues studied hybrid work arrangements, not any software product. The studies validate the practice pattern Cadence operationalizes; they do not test Cadence.
• Do not overstate the evidence’s reach: the causal results cover specific firms, job types, and hybrid intensities. “Hybrid done deliberately does not damage performance in the best available RCTs” is defensible; “remote always wins” is not.
• Do not claim Cadence eliminates proximity bias. Outcome documentation gives calibration better inputs; humans still make the judgments, and bias requires ongoing human vigilance, not a dashboard.

The defensible claim: the practices that replace proximity — explicit priorities, structured 1:1s, visible goals, outcome-anchored evaluation — are independently supported, and Cadence’s contribution is running them in one connected system instead of scattered tools. That integration argument is Cadence’s own synthesis, labeled as such in the People Science hub.

The argument in one paragraph

Proximity was never a management practice; it was a subsidy. It let organizations skip writing down priorities, skip structuring conversations, and skip anchoring evaluation to outcomes, because ambient observation papered over the gaps. The randomized evidence shows that when the subsidy disappeared, performance did not — hybrid work managed deliberately holds performance and improves retention. [1] [2] What actually broke was leaders’ confidence (productivity paranoia) and evaluation fairness (proximity bias), both of which are visibility problems, not output problems. The organizations that thrive distributed are not the ones that recreate the office or install surveillance; they are the ones that replace an implicit management model with an explicit one — written priorities, structured 1:1s, visible goals, outcome-based evaluation — running on a rhythm instead of a hallway.

How to cite this document

Suggested citation: Cadence, “Management by Proximity Is Dead: The People Science of Leading Hybrid and Distributed Teams” (2026). https://cadencehr.ai/resources/management-by-proximity-is-dead

Methodology and provenance. This synthesis was drafted in July 2026 by Cadence as part of the People Science research pillar. It prioritizes causal evidence where it exists — two randomized controlled trials of remote and hybrid work — supplemented by large-sample workplace research programs (Microsoft Work Trend Index, Gallup) and peer-reviewed field research on face-time perception. Every citation below was verified to resolve to the named source as of 2026-07-27. Claims about Cadence’s product carry current availability labels (live / preview / roadmap); the integration argument is identified as Cadence’s own synthesis.

References

1. Bloom, N., Liang, J., Roberts, J., & Ying, Z. J. (2015). “Does Working from Home Work? Evidence from a Chinese Experiment.” The Quarterly Journal of Economics, 130(1), 165–218. doi:10.1093/qje/qju032
2. Bloom, N., Han, R., & Liang, J. (2024). “Hybrid Working from Home Improves Retention Without Damaging Performance.” Nature, 630(8018), 920–925. doi:10.1038/s41586-024-07500-2
3. Microsoft WorkLab, “Hybrid Work Is Just Work. Are We Doing It Wrong?” 2022 Work Trend Index Pulse Report. microsoft.com/worklab
4. Gallup, “The Future of Hybrid Work: 5 Key Questions Answered With Data.” gallup.com/workplace/390632
5. Gallup, “Hybrid Work” indicator (ongoing tracking of remote-capable U.S. employees). gallup.com/401384
6. Elsbach, K. D., Cable, D. M., & Sherman, J. W. (2010). “How Passive ‘Face Time’ Affects Perceptions of Employees: Evidence of Spontaneous Trait Inference.” Human Relations, 63(6), 735–760. doi:10.1177/0018726709353139

This is a research synthesis, not a Cadence customer-outcome claim. Module availability is labeled because this page covers both live and roadmap concepts.

This article is part of Cadence’s People Science research pillar.

See how Cadence turns people science into operating rhythm at cadencehr.ai/product, or check plans at cadencehr.ai/pricing.