WFH.teamOpen appStart with the problem, not the attendance rule
Return-to-office decisions often bundle several different questions together: Are projects falling behind? Are new employees getting enough coaching? Do teams resolve complex issues more effectively face to face? Is remote work affecting retention? These questions are related, but they are not interchangeable - and none can be answered by counting building entries alone.
Begin by describing the problem in a way someone could observe. "People need to collaborate more" is a hypothesis, not a diagnosis. "Design reviews are repeatedly delayed because product and engineering decisions happen in separate channels" is more useful: it identifies a process and suggests evidence to examine. Likewise, if the concern is onboarding, specify what is going wrong: new hires may be waiting too long for answers, missing practice opportunities, or taking longer than expected to handle routine work independently.
Then name the outcome a policy is meant to change. For delayed reviews, that might be time from request to decision, the number of reopened decisions, or rework caused by missed requirements. For onboarding, it could be time to complete defined tasks independently, access to coaching, or new-hire feedback about clarity. Choose measures that reflect the problem rather than a convenient proxy such as desk occupancy.
Research can help frame the question, but a study's conclusion only applies as far as its population, methods, and setting support. Coverage of return-to-office decisions and reports on remote-work patterns point to relevant evidence to investigate; neither, by itself, establishes what will work for a particular employer. Before relying on a headline, check what was measured, who was included, what was compared, and whether the finding is relevant to the work at hand.
- Describe the operational issue without using location as the explanation.
- Identify the team, task, or workflow affected and the outcome that should change.
- Write down what result would count as improvement - and what result would not.
A useful workplace policy begins with a testable problem, not a preferred answer.
Measure work outcomes - not visibility
Select a small set of measures that reflect how the work creates value. Depending on the role, these might include delivery against agreed deadlines, customer resolution, quality checks, error rates, or rework. Use more than one measure when a single metric could reward the wrong behavior. A team that closes more tickets, for example, may appear faster while leaving difficult cases unresolved; a team that delivers a project sooner may have traded away quality or maintainability.
For a coordination problem, look for signals in the workflow. How long do handoffs take? How often does work wait for a decision? Are requirements being missed and later restored through rework? Are cross-team dependencies resolved at the rate the plan requires? These measures can establish that a problem exists, but they do not prove that an office mandate is its remedy. Unclear ownership, overloaded reviewers, or poor documentation may explain the delays more directly.
Include outcomes that can reveal costs beyond day-to-day output. Retention, hiring acceptance, absenteeism, employee feedback, and access to development may shift even when delivery measures do not. Compare like periods where possible, and record factors that could distort the comparison: staffing changes, workload, leadership changes, a product launch, or seasonal demand. If performance improves after a policy change, timing alone does not establish that the policy caused the improvement.
Set the baseline before changing the arrangement. Decide who will review the results and how often, and avoid changing several major practices at once if you need to understand what made a difference. Numbers should be paired with context: a rise in missed deadlines might reflect a new, more complex workload rather than weaker coordination. Ask teams to explain patterns and check those explanations against project records and customer or quality outcomes.
Why the right arrangement differs by task and team
Location needs can vary inside a single role. A software developer may need uninterrupted time to investigate a problem, a scheduled session to resolve an architectural decision, and occasional in-person support while learning an unfamiliar system. A customer-support team may handle routine cases independently but benefit from coordinated coaching or rapid escalation for complex issues. The useful question is not whether an entire occupation is "remote-friendly"; it is which activities need which conditions.
Some work has practical constraints: equipment, secure facilities, physical materials, in-person service, or hands-on instruction. Other tasks can be completed effectively using digital tools, provided that responsibilities, access, and communication are well designed. Even where work can happen remotely, teams may need planned overlap for decisions or learning. Conversely, requiring people to commute on days when their collaborators are elsewhere may add cost without addressing the stated need.
Team experience and management practices matter. A newly formed group may need explicit onboarding, clear decision owners, and deliberate relationship-building. An established team with strong documentation may coordinate well across locations. Time-zone differences, tool reliability, meeting habits, workload, and access to suitable workspace can also shape results. These factors are reasons to investigate differences - not reasons to assume that one arrangement will always succeed or fail.
Different expectations can be reasonable when they follow meaningful differences in the work. For instance, employees responsible for maintaining specialized equipment may need scheduled on-site coverage, while colleagues handling documentation may not. Explain the operational reason, define who makes the decision, and provide a consistent way to ask questions or request an adjustment. A policy that uses job titles as shortcuts without checking actual duties risks being both inefficient and difficult to trust.
Match the work arrangement to the task, the team's coordination needs, and the support employees actually have.
Compare the options - and make in-person time purposeful
"Hybrid" can mean a team-selected day, fixed days tied to particular activities, or a minimum number of office days per week. "Remote" can still include occasional travel, training, or planned gatherings. Define the options before comparing them: who chooses where to work, how much notice is expected, whether teams need overlapping hours, and what work is supposed to happen in person. Otherwise, employees and managers may use the same label for different arrangements.
Compare remote, hybrid, and office-based options against the same goals and costs. Consider coordination, quality, access to talent, workspace expense, commuting time, employee preferences, and the effort required to manage each arrangement. Then look for a targeted change before adopting a broad rule. If new hires wait days for answers, assign onboarding contacts and set response expectations. If decisions are scattered across chat, meetings, and email, establish a decision log. If cross-team planning is slow, schedule a working session for the people who need to make the decisions.
When people do meet in person, give the time a job. A training session, planning workshop, or conversation that benefits from sustained discussion may need a different schedule and attendance group from routine individual work. Coordinate days so collaborators are present together, publish the purpose and agenda, and make decisions and materials accessible afterward. Otherwise, an office day can become a series of separate video calls, with the added burden of commuting.
Use this decision rule: require attendance when the organization can identify a work activity that needs the shared setting, explain why a less restrictive option is unlikely to meet the need, and specify how it will judge whether the arrangement helped. If those conditions are not met, test a narrower intervention first. That might mean one planned overlap day for a particular team rather than a company-wide minimum, or an in-person onboarding period rather than a permanent attendance rule for every employee.
- Define each option in operational terms, not just with labels like "hybrid."
- Compare each option against the same outcomes, costs, and employee impacts.
- Schedule in-person days around shared work, and make remote participation effective.
Run a time-limited pilot with a decision rule
A pilot is not just a new rule with a later announcement. It is a planned test that gives the organization a chance to learn and adjust. Start by identifying the affected team and the specific problem. Record the baseline for a period that reflects the workflow, and note major conditions that could influence the result. If a team handles a monthly close, for example, a short trial that misses the close may not tell leaders much about its coordination needs.
Gather structured input before choosing the intervention. Ask employees and managers where work stalls, which activities require real-time interaction, what currently helps, and what support is missing. Separate operational constraints from preferences, while recognizing that both can inform the decision. Use the answers to compare interventions: clearer ownership, improved documentation, scheduled overlap, more coaching, a targeted office day, or a broader change in location expectations.
Set the pilot's length, review date, measures, and decision rule in advance. A team testing a shared weekly planning day could track whether cross-team decisions are made on time, whether dependency-related rework changes, and whether the day is useful to the people expected to attend. Also monitor relevant costs such as travel time, missed focus work, or access to meetings for remote colleagues. Agree on what would justify continuing, changing, or ending the trial. If the measures move in different directions, treat that as a result to investigate rather than selecting only the most favorable number.
At the review, compare outcomes with the baseline and consider what else changed. Ask the team whether the policy addressed the original problem and whether any new issues arose. Share a plain-language summary: what was tested, what evidence was reviewed, what remains uncertain, and what happens next. A pilot only builds confidence if leaders are prepared to change course when the results do not support the original plan.
Check who benefits, who bears the costs
A workplace policy can improve one visible measure while creating costs elsewhere. Track hiring acceptance, retention, development access, and employee feedback alongside delivery. Where appropriate and lawful, examine whether effects differ by role or employee group rather than relying only on an organization-wide average. Commute distance, disability-related needs, caregiving responsibilities, and access to a suitable workspace may shape how an arrangement works in practice. These considerations do not settle every policy question, but they can reveal a cost or barrier that a headline performance measure misses.
Look for differences in access to information and influence. Are decisions being made in office conversations that are never documented? Do remote employees have comparable access to managers, coaching, feedback, and visible assignments? Make meeting notes, decisions, and next steps available to the whole team. When an in-person session is intended for learning or planning, provide a way for employees who cannot attend to contribute and receive the same information.
Explain expectations and adjustment processes clearly. Employees should know who to contact, how a request is reviewed, what information is needed, and how sensitive information is handled. Employers should check that their policies are consistent with applicable requirements and seek appropriate advice where needed. When an attendance rule appears to affect groups differently, investigate the reason and consider whether a more targeted arrangement could meet the same operational need.
A policy's results include its side effects: who gains access, who carries the cost, and who is left out.
Make the decision - and keep it open to review
For employers, a defensible decision can be summarized in five parts: the problem, the evidence, the options considered, the expected trade-offs, and the review plan. If leaders cannot explain what the policy is meant to improve or how they will tell whether it helped, they are not ready to treat a mandate as an evidence-based solution. They may still need to act, but they should be candid about the uncertainty and set a way to learn.
For employees and candidates, verify the actual routine rather than relying on a job listing's "remote" or "hybrid" label. Ask how often this particular team meets in person, who chooses the days, how much notice is given, whether the policy is under review, and how exceptions work. Ask how remote employees participate in onboarding, feedback, and development. You can browse remote job listings and remote companies, then use interviews to clarify how the arrangement works for the role you are considering.
If you are comparing employers, see how to evaluate remote-work policies when comparing employers. For a related guide to interpreting claims about employee experience, read about remote work evidence and engagement.
The evidence does not make one location right for every job or organization. It can help employers replace assumptions with better questions, test a response against the problem, and notice costs that would otherwise be overlooked. Define the need, choose a proportionate intervention, measure the result, and review it honestly. The best return-to-office decision is not necessarily the broadest or the most flexible: it is the arrangement that fits the work and continues to earn its place.
- Employers: explain the rationale, measures, trade-offs, and review date.
- Employees and candidates: confirm team-level expectations and how they may change.
- Treat the policy as revisable when the work or the evidence changes.