How To Scale? Ask Kilby
Kilby for entrepreneurs and executives..
User Prompt:
My team is moving too slowly. Every decision gets stuck in meetings and approvals. How do I speed things up without losing quality?
Kilby’s Response:
This is a classic scaling problem — and it’s almost never about the people. It’s about decision architecture.
The sharpest framework here is Bezos’s Type 1 vs. Type 2 distinction, and it’s worth internalizing because it gives you a shared vocabulary to break the logjam.
The core diagnosis: Your team is almost certainly applying heavyweight process to lightweight decisions. Every choice — reversible or not — gets routed through the same meetings, the same approvals, the same deliberation cycle. The result is you move at the speed of your slowest approver on decisions that could have been made and reversed in a day.
The fix is structural, not motivational:
Classify every decision as Type 1 or Type 2. Type 1 = irreversible, high-consequence (entering a new market, killing a product, major architecture choices). Type 2 = reversible, low blast radius (feature tweaks, pricing experiments, process changes, most day-to-day calls). When in doubt, default to Type 2 — the reversible assumption is almost always safer.
Route them differently. Type 1 decisions get careful deliberation, broad input, your direct involvement. Type 2 decisions get made by the person closest to the information, same day, with minimal approval overhead. No meeting required. No committee. If it’s wrong, they walk it back.
Set the 70% threshold. Most decisions should be made with ~70% of the information you wish you had. Waiting for 90% certainty means you’re almost certainly too slow. The remaining 30% gets revealed by execution, and course-correcting is nearly always cheaper than waiting.
Escalate misalignment fast, not slow. When two people or teams disagree, don’t let it simmer through three rounds of meetings and Slack threads. Escalate immediately, get a decision, move. Slow resolution of disagreement is one of the most insidious sources of drag.
Stop punishing good-process / bad-outcome decisions. If someone made a fast, reversible call with reasonable information and it turned out wrong — that’s fine. That’s the system working. If your team learns that the only safe move is to wait until the answer is obvious, you’ve trained them to be slow.
The meta-principle: If you’re good at course-correcting, the cost of being wrong drops dramatically. That changes the math. Being wrong becomes cheap; being slow becomes expensive.
Concrete next step: Take the last 10 decisions that got stuck in your approval flow. Classify each one — Type 1 or Type 2. I’d bet 7-8 of them were Type 2 and never needed a meeting. Use that as the evidence to reset expectations with your team: “If it’s reversible, make the call. If you’re wrong, we fix it. The only thing I don’t want is slow.”
That’s the unlock. Not more process — less process, applied more precisely.
Learn more: https://agentkilby.com/
Kilby speaks