Subscribe or follow on X for updates when new posts go live.
In many organizations, particularly high-performing or fast-scaling ones, there is a dangerous undercurrent that runs beneath everyday work: the unspoken belief that every decision carries existential risk. This mentality, while often rooted in good intentions and a desire for quality, leads to analysis paralysis, bloated review processes, and a culture of fear. Over time, it throttles speed, stifles innovation, and erodes trust in individual contributors.
It doesn’t have to be this way.
A mental model popularized by Jeff Bezos offers a practical and liberating way to operate at high speed without sacrificing judgment. It’s the distinction between Type 1 and Type 2 decisions. It's a framework that, when adopted with discipline, enables companies to move fast where it’s safe and slow where it matters.
Let’s walk through this model, why most companies get it wrong, and how engineering teams can reclaim autonomy and speed by applying it.
The core idea is simple:
Most people nod in agreement when this is explained. But in practice, very few organizations act accordingly.
Instead, the default behavior is to treat all decisions like they are Type 1. New feature idea? Better run it through a stakeholder meeting. Minor API change? Needs a cross-functional review. Marketing test? Let’s schedule a workshop and define OKRs before we spend $50.
This approach is cautious, and it’s often disguised as "due diligence." But what it really signals is a lack of trust, either in the decision-making process or in the people who are supposed to own it.
When you treat every decision like a potential company-killer, here’s what happens:
I’ve seen this firsthand on teams where a culture of caution crept in over time. The company wasn’t dysfunctional. In fact, on paper, it looked great. But inside, nothing shipped unless it was polished, reviewed by three levels of management, and survived multiple cycles of internal debate.
Ironically, this obsession with safety made the company less safe. Delayed features hurt user retention. Missed experiments meant missed growth opportunities. Top engineers left for environments where they could move faster and own outcomes.
This is the counterweight. The antidote. The liberating truth.
Most decisions are reversible. You can ship that UI tweak and roll it back tomorrow. You can A/B test a new onboarding step and revert it if it underperforms. You can push that API change behind a feature flag and kill it if it breaks something.
Treating every decision like Type 1 is not only inefficient, it’s incorrect. The vast majority of decisions we make in product and engineering are Type 2, and we should act like it.
Here are a few examples of Type 2 decisions:
None of these changes are going to sink the company. They might not even be noticed. And yet, in many organizations, each would trigger a formal process designed for a database migration or a security policy update.
That’s backwards.
There are decisions that deserve the full treatment. You want to slow down when:
Examples:
These are decisions where the downside is massive. Get them wrong, and you’re talking about customer churn, legal exposure, or millions of dollars in rework. These deserve multiple layers of review, senior sign-off, and deliberate execution.
Start by educating your team on the difference between Type 1 and Type 2 decisions. A short workshop or lunch-and-learn can be enough. This model is intuitive once people hear it.
Define what counts as a Type 1 decision for your team or org. Keep it simple, but explicit. If security, compliance, or infrastructure is involved, it’s probably Type 1. Otherwise, default to Type 2.
Empower individual engineers, designers, and PMs to make Type 2 decisions without waiting for approval. Give them a fast lane. Let them ship. Trust the team to own their decisions and learn from them.
Reversibility is only comforting if you actually can reverse. Make sure every change has a clear rollback plan, even if it’s just a simple re-deploy or feature flag toggle.
Make speed a virtue. Reward people for making thoughtful, reversible decisions quickly. Highlight wins that came from testing an idea, not over-analyzing it.
At one company I worked with, we had an internal culture of treating everything like a legal contract. A product manager wanted to test a new onboarding question; just one additional step asking users about their experience level to tailor their dashboard.
Instead of shipping and testing it, the idea sat in limbo for weeks. It went through multiple design reviews, stakeholder alignments, and an eventual roadmap re-prioritization. By the time it launched, we’d missed a key marketing campaign, and engagement uplift was never measured.
In contrast, a single engineer on another team pushed a performance improvement to a key API endpoint without asking for permission. It sped up the page load by 40 percent. No reviews, no meetings. Just a quick PR, a clear rollback path, and a measurable outcome.
Guess which change had more business impact.
The belief that every decision could destroy the company is corrosive. It creates bottlenecks, saps morale, and undermines the trust needed to move fast and build great products.
Instead, adopt the Type 1 / Type 2 framework as a cultural norm. Treat irreversible, high-impact decisions with the care they deserve. But everything else? Recognize it for what it is — a reversible bet, an experiment, a learning opportunity.
Let your team walk through the two-way doors. Often. Confidently. And without waiting for permission.
The real risk isn’t moving too fast. It’s forgetting that you can always turn around.