How Zapier Scaled to 800 People Fully Remote With Radical Transparency and DRIs

Snapshot

Zapier has been fully remote since their founding in 2011. The three co-founders — Wade Foster, Bryan Helmig, and Mike Knoop — were in different cities from the start, went through Y Combinator in summer 2012, and never opened an office. They scaled from three people to 800+ employees across 30+ countries and six continents without ever maintaining a physical headquarters.

That growth trajectory — a $1.2M seed round in October 2012 (Bessemer/DFJ), profitability since 2014, and a scale to $140M+ ARR and a $5B valuation on roughly $1.4M of total venture capital, with Sequoia and Steadfast involvement in 2021 coming via a secondary transaction rather than a new primary round — makes Zapier one of the most important test cases for the scalability of bootstrapped-style remote-first operating models.

Core Philosophy

Zapier publishes five core values: Default to Action, Default to Transparency, Grow Through Feedback, Empathy Over Ego, and Build the Robot. The two that have been most discussed publicly by Wade Foster (CEO) as load-bearing for distributed work are the first two.

Default to Action. Hire self-starters who identify and solve problems independently. In a distributed company, you cannot see when someone is struggling or blocked — so you hire people who won’t sit stuck waiting for permission. (Source)

Default to Transparency. Avoid private direct messages. Keep communications public so that solutions, decisions, and reasoning are discoverable and reusable by everyone — not just the people in the thread. (Source)

These two principles in particular shape hiring, communication norms, and how work gets structured. The thread running through both: a fully remote company cannot run on information that lives in people’s heads or in private conversations.

Communication Model

Async across time zones is the default. Work happens continuously across six continents. Synchronous moments are designed and scheduled — not the default mode of getting things done. (Source)

Async communication practices:

• Write questions with preemptive context. Before sending a message, include the context needed to answer it, your own best thinking on the answer, and what you need from the recipient. This minimizes back-and-forth cycles.
• Work in public. Share docs, links, and in-progress work in public channels by default — not in DMs. This makes work visible and searchable.
• Create traceable information trails. Slack threads should link to GitHub issues, Quip docs, or other permanent artifacts. “A conversation that doesn’t result in a document or an issue doesn’t really exist.”

(Source)

Primary tools:

• Slack — virtual office for work chat (not real-time; async usage norms)
• Async (Zapier’s internal blog tool) — for showcasing work, long-form updates, documentation
• Quip — internal knowledge base and documentation
• Zoom — large meetings
• Google Docs/Sheets/Calendar — supplementary collaboration
• Coda — additional documentation

(Source)

Planning and Cadence

Zapier uses an outcomes-driven planning model with several confirmed practices:

DRIs (Directly Responsible Individuals). Every project, initiative, and decision has a single named owner. The DRI is accountable for progress and for escalating blockers — not waiting to be asked. This is the same model GitLab uses, and for the same reason: in an async company, diffuse ownership is invisible failure.

Single-page specs. Every significant project or feature is documented in a single-page spec before build begins. Forces clarity of thinking upfront. Makes the project visible and commentable by everyone.

Solution sprints. Short, targeted cross-functional missions sponsored by executives — used for specific problem-solving, not general roadmap planning. (Source)

Customer-centered annual Summits. Zapier replaced traditional company retreats with customer-centered summits — events oriented around customer feedback and product direction, not purely internal bonding. (Source)

AI adoption as an organizational forcing function. In 2024, Zapier ran a company-wide AI adoption initiative that achieved 97% adoption across ~800 employees. This was managed via DRIs, single-page implementation plans, and transparent progress tracking. (Source)

Decision-Making Model

DRI model — same principle as GitLab’s. Every project has a single accountable person. Decisions are made by the DRI after gathering input, not by committee.

Default to transparency applies to decisions. Significant decisions should be visible, documented, and accessible — not made in private DMs. The goal: anyone in the company can understand why a decision was made, not just that it was made.

Co-founders Wade Foster (CEO), Bryan Helmig (CTO), and Mike Knoop (CPO) retain strategic decision authority. The executive team includes a Chief People Officer, CRO, and standard enterprise functions for a company of this size.

Hiring decisions are deliberate. At 800 people, Zapier still recruits specifically for the “default to action” and “default to transparency” profiles. Self-direction is a required trait, not a preference.

Org Structure

• ~800 employees across 30+ countries and 6 continents. (Source)
• Largest departments: Engineering (322), Marketing & Product (133), Sales & Support (87).
• Standard executive team — CEO, CTO, CPO, CRO.
• No offices — distributed globally since founding.
• Geographic clusters exist (SF, Portland, Seattle, NYC, Toronto) but are not offices — these are where individual employees happen to live.

Tools and Stack

(Source)

Rituals

Customer-centered annual Summits. Replacing traditional company retreats with customer-focused events is a distinguishing choice — it keeps the company externally oriented even during internal bonding time.

AI adoption initiative. Zapier’s 2024 company-wide AI rollout — 97% adoption across ~800 employees — was treated as an organizational ritual. A structured, DRI-driven, transparently tracked initiative that became a forcing function for skill development and cultural alignment. (Source)

Async blogging of work. The internal “Async” blog tool means employees regularly publish work updates as long-form posts, not just status messages. This creates a running documentary of how the company thinks and what it is doing.

What They Do Well

• Scaling remote-first to 800 people. Most remote-first frameworks are designed by 60-person companies. Zapier has done the hard work of maintaining the model through 10x growth. The DRI system and single-page specs exist specifically because “default to action” breaks down when work becomes too ambiguous at scale.
• Default to transparency as an operating constraint. When transparency is the default — not an aspiration — decisions are made differently. The question “would I be comfortable if this conversation were public?” shapes judgment in real time.
• AI adoption as evidence of organizational coordination at scale. 97% adoption of a new tool across 800 fully remote employees in a single initiative demonstrates operational discipline. This is a test of coordination, not just technology.
• The internal blog as a knowledge system. The “Async” internal blog creates permanent, searchable documentation of how work happened — not just what was decided. This is different from a wiki (policy documentation) or a project tracker (task status).

Tradeoffs and Weaknesses

Source quality for this page is lower than ideal. Zapier does not have a public handbook on the scale of GitLab or 37signals. Their operating model is documented primarily through blog posts, founder interviews, and secondary sources. Some details remain unverified or could be outdated. (Note: source_quality is marked strong-secondary accordingly.)

800-person remote coordination requires serious process infrastructure. DRIs, single-page specs, solution sprints, and public async communication are all in place because, at this scale, “just communicate well” is not enough. The overhead of maintaining these systems is real.

Geographic clustering creates implicit timezone bias. Despite being distributed globally, the largest concentrations are in US cities. Employees in Asia-Pacific or Africa face real time-zone disadvantages for synchronous moments, even if those moments are rare.

“Default to action” creates coordination gaps. When people take action without sufficient input from stakeholders who are async and will see the results later, it creates rework. The “write questions with preemptive context” norm partially addresses this — but high-action people in a remote setting sometimes move before the relevant people have had a chance to weigh in.

What Founders Can Copy

Zapier scaled from a small founding team to 800+ all-remote without an office at any stage. Most of these practices were built early and stress-tested at scale; per-item scale notes flag the cleanest size to introduce them.

1. The two most behavioral of Zapier’s five values. “Default to action” + “default to transparency” are simple enough that they actually shape behavior. Most company value statements are too abstract to change anything. These two — pulled from Zapier’s published five — are behavioral. (Applies at any scale; cheapest to install at 10–50 when the value statement still shapes hiring decisions before becoming retrofit narrative.)
2. DRI on everything. Single-page spec + a named DRI = the minimum viable project management system for a remote team. Scales from 10 to 800 employees. (Applies at any scale, including the largest documented in this dataset; load-bearing above 25 people, the size at which “the team owns it” stops working.)
3. Single-page specs. Before any significant project starts, one page that covers the what, why, how, and success criteria. This is not bureaucracy — it is the cost of making async coordination possible. (Applies at any scale; the one-page discipline is most defensible at 30–300 — above 300, some specs legitimately need more depth and pretending otherwise produces vague specs.)
4. No DMs for work decisions. Default all work communication to public channels. Private DMs create information silos that distributed teams cannot recover from. Make “default to transparency” a literal rule about where work conversations happen. (Applies at any scale; much harder to retrofit above 100 because DM culture is hard to displace once entrenched.)
5. Write questions with preemptive context. Train your team to front-load messages: “I am trying to X. I have already tried Y and Z. My best guess is A. Do you see anything I am missing?” This alone cuts async back-and-forth cycles by half. (Applies at any scale; gets more valuable as headcount grows because the surface area of who-replies-to-whom expands faster than headcount.)

Where This Model Breaks

• When DRI accountability meets organizational complexity. At 800 people, some projects genuinely require shared accountability across multiple functions. Forcing a single DRI onto a cross-functional initiative creates friction if the DRI does not have actual authority over all the inputs. The model requires clear authority mapping to work.
• When transparency norms meet sensitive decisions. Not everything can or should be public within the company — personnel decisions, M&A exploration, certain investor conversations. “Default to transparency” requires clear carve-outs for what stays private, or it creates anxiety about sensitive information leaking.
• When “default to action” produces contradictory actions. Multiple DRIs moving fast in the same problem space, without sufficient alignment, can produce redundant or contradictory outputs. The single-page spec process is supposed to catch this before it happens — but it requires that people actually read each other’s specs.
• Below 30 people. The DRI/single-page-spec system is appropriate overhead at 300+ people. Below 30, it adds bureaucracy to interactions that could just be a conversation.

Related Practices

• Async Communication — the foundation of Zapier’s operating model
• Decision Ownership — DRI model in practice
• Documentation Systems — single-page specs and internal blogs as documentation infrastructure

Related Frameworks

• 20-30 Person Fracture Point — Zapier navigated this transition; the DRI model was built partly in response
• Sync vs Async Decision Matrix — when “default to async” needs boundaries

Sources

1. How to Manage a Remote Team (Zapier): https://zapier.com/learn/the-ultimate-guide-to-remote-working/how-manage-remote-team/
2. Remote Talks — Wade Foster on Working Async: https://remote.com/blog/remote-work/remote-talks-episode-5-wade-foster
3. Work from Home Apps: Tools We Use at Zapier: https://zapier.com/blog/team-communication-tools/
4. How to Work Asynchronously (Zapier): https://zapier.com/blog/how-to-work-asynchronously
5. Zapier CEO on Scaling to 300 Remote Employees (TechCrunch): https://techcrunch.com/2020/03/09/zapier-ceo-wade-foster-on-scaling-a-remote-team-up-to-300-employees
6. How Zapier Rolled Out AI Org-Wide: https://zapier.com/blog/how-zapier-rolled-out-ai/

Inferences

• The “default to action” value is self-selecting at hire. People who prefer to ask permission before acting, who want clear direction before starting, or who thrive in high-structure environments will self-select out of Zapier’s operating model during the interview process or within their first year. This is by design — but it means Zapier systematically underrepresents certain personality and work-style types that perform well in other environments.
• The single-page spec discipline exists because “default to action” creates coordination problems at scale. Two DRIs moving fast in adjacent areas without alignment produces rework. The spec is the collision-prevention mechanism. At 10 people, you don’t need it — you collide and fix it immediately. At 800 people, you can’t afford to.
• Zapier’s 97% AI adoption initiative reveals something important about organizational coordination at this scale: achieving company-wide behavior change in a remote company requires the same infrastructure as any complex project — DRI, clear spec, transparent progress tracking, staged rollout. The content of the change (AI tools vs. any other change) is secondary to the coordination mechanism.

Work with Alex

If you are scaling a remote team past 50 people and struggling to maintain coordination without creating meeting overhead, Alex helps leadership teams build the process infrastructure that keeps distributed execution coherent at scale.