Remote-First Networking: Staying Visible Without an Office

Distributed work removed the accidental encounter — the corridor, the lift, the person you ended up beside at lunch — and nothing automatically replaced it. The best evidence on what that did to professional networks comes from a firm-wide natural experiment at Microsoft, and the finding is specific: networks did not shrink so much as harden, with fewer connections bridging between parts of the organisation.

What the Microsoft data showed

Longqi Yang and colleagues published “The effects of remote work on collaboration among information workers” in Nature Human Behaviour in 2022 (volume 6, pages 43–54). They analysed communications data — email, calendars, instant messages, and video and audio calls — for 61,182 US Microsoft employees over the first six months of 2020, using the shift to firm-wide remote work as the shock and comparing employees who were already remote against those who moved.

Three findings matter for anyone thinking about their own network:

• The collaboration network became more siloed. Connections concentrated inside existing groups, and there were fewer bridges between disparate parts of the organisation. In the vocabulary of the weak-ties literature, the edges that reach somewhere new are exactly the ones that were lost.
• It became more static. People kept collaborating with the same people. New ties formed less readily.
• Communication shifted from synchronous to asynchronous. Less real-time contact, more messages and email.

The authors conclude that together these effects may make it harder for employees to acquire and share new information across the network. This is not a claim about productivity on current tasks; it is a claim about the structure of who knows whom, which is what careers run on.

Two supporting results

Natalia Emanuel, Emma Harrington and Amanda Pallais studied software engineers at a Fortune 500 firm between 2019 and 2024 in NBER working paper 31880, using office closures in 2020 and return-to-office mandates in 2022–23 as two shocks to physical proximity. Sitting near teammates increased the volume of coding feedback engineers received by 18.3 per cent. The distributional result is the important one: junior engineers gained skills from proximity, while senior engineers wrote less code when co-located. Proximity was a transfer, not a free lunch — and the people paying the cost of remote work were the ones with the least to lose in output and the most to lose in development.

Melanie Brucks and Jonathan Levav’s “Virtual communication curbs creative idea generation” (Nature, 2022, volume 605, pages 108–112) tested a narrower mechanism with 151 pairs in the lab and field experiments across five countries. Video calls produced fewer creative ideas than in-person meetings, and eye-tracking suggested why: a screen narrows visual and cognitive focus. Notably, videoconferencing groups were as good as or better than in-person groups at selecting which ideas to pursue. That is a usable division of labour — generate in person, decide on the call.

The Canadian shape of the problem

Remote work has receded but not reverted. Statistics Canada reported that in May 2025, 17.4 per cent of employed people worked primarily from home, down from 18.7 per cent a year earlier — a fourth consecutive annual decline. About 10.0 per cent of workers had hybrid arrangements, splitting hours between home and another location, and 82.6 per cent of workers were commuters, up 1.3 percentage points on May 2024.

The implication of a hybrid majority is that the office no longer generates encounters reliably either. If half your team is in on Tuesdays and you are in on Thursdays, you have the commute cost of co-location and the network structure of remote work. Attendance without coordination is the worst of both.

What to do about it

The tactics below are craft. They are chosen to attack the specific failure the research identifies — missing bridges to other groups — rather than to make you generally more sociable.

Asynchronous visibility

In a distributed organisation, written output is the substitute for being seen. The rule that follows from the siloing finding is not “post more” but “write where people outside your group can find it.”

• Put decisions and their reasoning in a searchable place — a wiki page, a design doc, a public channel — rather than in a call or a direct message. The value is that someone in another team finds it six months later and now knows who you are.
• Answer questions in the widest channel where the question was legitimately asked. A helpful answer in a shared channel is a bridge; the same answer in a DM is not.
• Write short internal summaries after projects, including the parts that went badly. These circulate further than status updates because they contain information nobody else has.
• Keep your internal profile current. In a remote firm, the directory entry does the work a desk location used to.

Communities of practice

A community of practice — a group organised around a craft rather than a reporting line — is the most direct replacement for the bridges that disappeared. Inside a company that is the guild or chapter spanning teams; outside it, the professional association, the standards working group, the open-source project, the regional industry body. Attendance is not the point; contributing is, because that is what makes people remember your name.

Virtual coffees that people actually accept

The reason most “quick virtual coffee” requests are declined is that they ask for unbounded time to discuss nothing in particular. Four things raise the acceptance rate, in rough order of effect:

1. Name a specific question. “How did you handle the bilingual requirements on that rollout?” is answerable. “I’d love to hear about your career” is a chore.
2. Name a short duration and hold to it. Twenty minutes, and end at twenty minutes even if it is going well. People accept a second call from someone who respected the first one.
3. Say why them. A falsifiable reason — a post they wrote, a project you both touched, a mutual contact.
4. Offer something. Even small: what you found when you looked into the same problem.

Given the Brucks and Levav result, use these calls for the things video does well — a specific question, a decision, a judgement about which of two options to pursue — and do not expect them to generate the loose, wandering exchange that produces new ideas. That is not what the medium is good at.

Travel budget as a network investment

If proximity increases feedback by roughly a fifth and video calls suppress idea generation, then a few days a year in a room with people is not a perk; it is the mechanism that produces the bridges the Microsoft data says are missing. The argument for a conference or an offsite is stronger when it is made in those terms, and the spending is more effective when planned that way.

Purpose	Best channel	Why
Generating options and new ideas	In person	Video narrows cognitive focus and reduced idea generation in Brucks and Levav’s experiments
Choosing between known options	Video call	Videoconferencing groups selected ideas as well as or better than in-person groups
Building a tie with someone outside your group	In person, then maintained asynchronously	Bridging ties are what firm-wide remote work eroded; they form best face to face
Being findable by people who do not know you	Written, in a searchable shared space	Nobody bumps into you; they have to search for you
Developing junior colleagues	Co-located time, deliberately scheduled	Proximity raised feedback volume 18.3 per cent, with gains concentrated among junior engineers

If you manage people

Two things follow from the evidence rather than from taste. First, the cost of distributed work falls unevenly: junior staff lose development, senior staff lose focus time. A blanket policy makes one of those groups pay for the other. Second, “everyone in on the same days” is doing real work in the data — synchronised attendance is what converts office days into encounters, and unsynchronised attendance produces the commute without the benefit.

The practical version

The measured harm of remote work to networks is not loneliness; it is the loss of ties that cross group boundaries, in a network that also stopped changing. Attack that directly. Write in shared, searchable places so people outside your team can find you without bumping into you. Join one community organised around your craft rather than your employer, and contribute to it rather than lurking. Ask for twenty minutes with a specific question rather than an open-ended coffee. Use video for decisions and rooms for generating ideas. And if you are hybrid, coordinate your in-office days with the people you actually need to run into — an office day nobody else attends buys you nothing at all.