See how Alex worksWork with Alex
All practices

Practice — Onboarding

Remote Onboarding: How Distributed Companies Ramp New Hires Without an Office

Remote onboarding is the structured process of integrating new employees into a distributed company without in-person orientation. The examples combine documented expectations, named support contacts, early participation and staged onboarding; the duration of the ramp varies by role and company.

By Alex Karasyov — Operating consultant for product & engineering teams

Last reviewed November 1, 2026

What It Is

Remote onboarding is the process of bringing a new employee into a distributed company — equipping them with the tools, context, relationships, and clarity they need to do effective work — without the informal scaffolding of a shared physical office.

[Inference — application framing.] A distributed newcomer has fewer opportunities for informal observation in a shared office. The examples below make guidance, relationships and expectations explicit; they do not show that remote work removes every informal learning opportunity.

Why It Works

[Inference — these company examples illustrate onboarding designs; they do not measure an optimal number of contacts or universal ramp duration.]

37signals names several support routes. Its guide assigns an Ops buddy, a cross-team buddy and a manager, alongside a welcome project. It says supported real project work begins in the first week and most newcomers feel up to speed in about three months. (Guide)

GitLab separates orientation from heavy contribution pressure. Its handbook recommends at least two full weeks of onboarding, with team-specific training from week three, while explicitly welcoming participation in team work during the first two weeks. (Handbook)

Automattic publishes support-based orientation with an exception. The general careers page describes two weeks of WordPress.com support plus a yearly week. Its more specific first-day guide tells VIP hires to wait for lead instructions for a different support rotation. Treat that as a policy example, not a universal rule or proof that it changes how engineers code. (General guidance, first-day guide)

Where It Fails

[Inference — risks to assess rather than ranked or measured cross-company failure rates.]

Autonomy without support. A checklist needs named routes for questions and adjustment. Self-direction is not proof that a newcomer should navigate everything alone.

Isolation. GitLab discusses loneliness as a possible remote-work drawback; the sources do not rank it as the most common onboarding failure or prove coffee chats solve it. (Drawbacks)

One ramp expectation for every role. 37signals describes about three months to feel up to speed; GitLab’s first orientation period is not the whole ramp. Neither establishes a universal productivity deadline. (37signals, GitLab)

Reading without context. The examples use staged tasks and supported work. That is a design choice to evaluate, not proof that every role must contribute in week one.

Best Examples

37signals / Basecamp: welcome project and named support. Equipment is ordered for the role before joining, with Omarchy/Mac scope specified by the current guide. On day one, the newcomer finds a tailored project and meets the manager to set expectations and recurring 1:1s. Ops and cross-team buddies provide additional help. (Guide)

Automattic: trial work and support orientation. General guidance describes a paid two-to-six-week contract trial and support orientation; the first-day guide tells VIP hires to wait for lead instructions for a different support rotation. The reviewed pages do not establish a universal first-year meetup requirement or that every onboarding record lives in P2. (General guidance, first-day guide)

GitLab: issue-based onboarding. People Operations opens an issue at least four days before the employee’s start date; the guide also uses four business days in its manager instructions. Tasks combine general and role-specific work, named support channels and a pre-start call every two weeks. Participation in team work is welcome during the first two weeks without pressure for heavy contribution. (Process)

Research Context

A 2026 study of fully remote Caster employees found benefits from experienced teammates. It describes six months of initial training and excludes employees’ first six months from its analytical sample. It therefore does not test first-90-day onboarding or the effect of assigning three buddies. The analysis uses 2022–2023 records from one business-services department, so transfer to other roles remains uncertain. (Study)

AI-assisted performance is not proof of learning. A study published October 26, 2026 reports that consultants given ChatGPT access and training performed better across three data-science task types, but the researchers found no evidence of better post-experiment technical answers without it than in the control group. The reported analysis covers 487 consultants. This was not a remote-onboarding study. [Inference — application guidance.] If onboarding uses AI assistance, assess task output and the person’s understanding separately. (Published study)

Additional Company Examples

Todoist (formerly Doist) publishes a first-90-days onboarding guide with weekly mentor and team-head meetings alongside written tasks. (Source)

Buffer sends role-specific Q&A documents before interviews. This is candidate preparation, not evidence about post-hire onboarding outcomes. (July 2026 account)

Implementation Guide

[Inference — implementation guidance from the cited examples, not validated outcome guarantees.]

  1. Design a staged plan. Explain the role, tools, expectations and where to seek help; adjust it to the newcomer.
  2. Name support responsibilities. Separate technical setup, cultural orientation and work direction where useful. Three people is one design, not a proven optimum.
  3. Give tasks a home. A personal project or issue can connect reading, access and early work.
  4. Choose supported early work. 37signals describes work in week one; Todoist stages first contributions later. Match the plan to the role rather than declare one schedule mandatory.
  5. Review progress together. Use checkpoints to clarify expectations; neither a 90-day date nor completed reading proves full productivity.
  6. Provide relationship-building options. Combine written conversation and calls as useful, accounting for time zones and individual needs.

(37signals, Todoist, GitLab)

Common Mistakes

[Inference — design risks, not measured failure rates.]

  • Automation without accessible people: make it clear whom to contact when a task or expectation is unclear.
  • Reading with no application: connect guidance to supported tasks when the role permits.
  • A nominal buddy assignment: agree how and when the buddy will check in, then adjust based on the newcomer’s needs.
  • No chance to revisit expectations: schedule conversations about progress and support without treating one universal 90-day check as an empirical requirement.

Sources

Automattic — Welcome to Your First Day (VIP hires follow a different support rotation): https://automattic.com/fieldguide/welcome-to-your-first-day/

  1. Getting Started — 37signals Employee Handbook: https://books.37signals.com/3/the-37signals-employee-handbook/7/getting-started
  2. How We Work — Automattic: https://automattic.com/how-we-work/
  3. Complete Guide to Remote Onboarding — GitLab: https://handbook.gitlab.com/handbook/company/culture/all-remote/onboarding
  4. GitLab Onboarding Process: https://handbook.gitlab.com/handbook/people-group/general-onboarding/
  5. All-Remote Drawbacks — GitLab: https://handbook.gitlab.com/handbook/company/culture/all-remote/drawbacks

Ishikura — Experienced teammates (PLOS One, March 2026): https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0342730

Todoist onboarding: https://handbook.todoist.net/s/6ca6f66e-2d02-4e05-b629-b1c772f5ff23

Buffer candidate Q&A (July 2026): https://buffer.com/resources/buffer-q-a-doc/

Wiles et al. — Generative AI and the temporary upskilling of knowledge workers (Nature Human Behaviour, October 26, 2026): https://www.nature.com/articles/s41562-026-02566-6

Inferences

  • Support-based orientation can expose newcomers to customer questions, but Automattic’s VIP instructions mean the standard WordPress.com rotation should not be universalized. The evidence does not measure empathy or coding-quality effects.
  • A named buddy and agreed check-ins may make support easier to find. The sources do not establish that unstructured buddy relationships always fail or identify an optimal number of contacts.

Work with Alex

If your onboarding process is “read the wiki and figure it out,” and you’re losing new hires in their first 90 days to confusion and disconnection, Alex helps distributed teams build the structured onboarding systems that actually work.

onboardingculturepracticeremote

Last reviewed November 1, 2026