Decisions Are the Bottleneck
Software engineering is fundamentally a decision-making discipline. Every day, teams face choices: monolith vs. microservices, REST vs. gRPC, build vs. buy, Kubernetes vs. serverless. The code itself is the output of decisions. When decisions stall, everything downstream stalls — sprints slip, engineers context-switch to other work, and momentum dies.
Consider a concrete scenario. Your team wants to migrate from Jenkins to GitHub Actions. This decision touches CI/CD pipelines (your team), security policies (the security team), cost budgets (finance/management), and developer experience (every engineering team). Without a framework, here is what typically happens:
- You write a proposal document and share it in Slack.
- Three people respond with questions. Seven people ignore it.
- You schedule a meeting. Four of the ten stakeholders attend.
- The attendees raise concerns you hadn't considered. The meeting ends with "let's think about it more."
- Two weeks pass. You schedule another meeting. Different people attend. The same concerns are raised again.
- After six weeks, the proposal dies — not because it was bad, but because no one could figure out how to say "yes."
This pattern is so common it has a name in organizational theory: decision debt. Like technical debt, it compounds. Unresolved decisions create ambiguity, which creates duplicate work, which creates frustration, which creates attrition.
Consensus frameworks solve this by answering three questions upfront: Who has input? Who decides? When is the deadline? With those three answers, even contentious decisions can resolve in days instead of months.
Key Point: In platform engineering and DevOps, decisions frequently cross team boundaries — infrastructure, security, development, operations. The ability to drive cross-cutting decisions is what distinguishes a senior engineer from a staff engineer. Interviewers at the staff+ level explicitly probe for this skill. They want to hear how you drove a decision, not just what you decided.