More Context, Fewer Hands
Give a small trusted team enough shared context to act from end to end.
- Difficulty
- Advanced
- Time to result
- ~ongoing to results
- Steps
- 5
- Confidence
- 94%
StackBlitz's operating model is to concentrate context in fewer people rather than divide work among many narrow roles. A small, long-tenured core develops trust and enough shared understanding to act without repeatedly asking for approval. Engineers can hear a customer problem, diagnose it, change the interface, and ship the fix from front to back. During Bolt's hypergrowth, the company reinforced the model with a daily whole-company video call at 8 a.m. Pacific, reviewing the business end to end. Simons accepts that this is an expensive use of time; the intended return is near-zero communication fidelity loss while conditions change daily. The structure is deliberately contingent. As the team grows or the crisis passes, it should split into smaller syncs rather than preserving the ritual forever.
Origin
StackBlitz carried this model from its bootstrapped roots. Five to seven core people stayed for more than five years, accumulated deep context around WebContainer, and then handled Bolt's sudden demand with a total company of roughly fifteen to twenty people.
Core principles
- 01Shared context lets people make better independent decisions.
- 02High trust removes approval layers.
- 03End-to-end ownership beats narrow functional handoffs.
- 04Long-tenured teams can sustain longer technical bets.
- 05During extreme change, communication fidelity can matter more than meeting cost.
How to run it
- 1
Hire for motive and ego
Choose people motivated by building excellent things, not by titles or personal territory. Favor builders who can check their ego at the door and work across boundaries.
Pro tip Recruit from the product community when possible; StackBlitz found many hires among people already using its technology.
Watch out A brilliant specialist who hoards ownership can lower the agency of the entire group.
- 2
Load shared context
Make product, customer, operational, and technical information visible to the whole core team. The goal is for each person to understand how their decisions affect the rest of the company.
Watch out Context without clear priorities can become noise; share what changes decisions.
- 3
Grant front-to-back agency
Let a person move from observing a customer issue to designing, implementing, and landing the fix. Use trust and context in place of standing approval committees.
Pro tip Define the outcome and constraints, then let the owner choose the route.
- 4
Raise sync frequency when change spikes
During a launch, crisis, or hypergrowth period, temporarily gather the whole team frequently enough to audit the operation end to end. StackBlitz used a daily one-hour company call after Bolt launched.
Pro tip Use the call to expose cross-functional changes and decisions, not to recite individual status updates.
Watch out The meeting cost is justified only while information changes fast enough to require it.
- 5
Split before context becomes congestion
As the company grows, move from one shared room to smaller, well-connected syncs. Preserve decision-relevant context without forcing every person into every discussion.
Watch out A ritual that once prevented fidelity loss can later create it by overwhelming participants.
In the wild
During Bolt's launch surge, engineers could join a call with someone experiencing a problem, diagnose it, change the interface, and land the fix without involving a sequence of managers or committees. Long tenure and shared context meant the engineer understood both the technical system and the product consequences.
→ A company of roughly fifteen to twenty people could respond to demand from millions of registered users without conventional layers of coordination.
The week after Bolt launched, almost the entire company began meeting on Zoom every day at 8 a.m. Pacific for about an hour. They reviewed the operation front to back because traffic, customer issues, pricing, and infrastructure were changing too quickly for lossy handoffs.
→ Every person retained enough current context to make more independent decisions outside the meeting.
Common mistakes
Equating autonomy with isolation
People can act independently only after they share enough context and trust. Removing approvals without supplying context produces inconsistent decisions.
Hiring narrow title-seekers
A small team needs people willing to cross role boundaries. Optimizing for title or individual brilliance undermines collective front-to-back ownership.
Keeping the all-company call forever
Simons explicitly expects the daily full-team structure to splinter as StackBlitz grows. The cadence is a response to extreme change, not a permanent doctrine.
Is it for you?
Best for
Small, high-trust teams operating in technically complex or unusually fast-moving environments.
Not ideal for
Large organizations where every employee cannot usefully absorb all company context or join one decision forum.
From the episode
Inside Bolt: From near-death to ~$40m ARR in 5 months—one of the fastest-growing products in history
Eric Simons (founder and CEO of StackBlitz)