Do Less Better: The Discipline Behind Bezos
- Alex Khachaturian
- Nov 8
- 5 min read
Promise: When you build less, but build it right, you trade noise for clarity and speed for endurance. That’s how to actually move the needle.
TL;DR
Most teams don’t need more ideas, they need fewer, better commitments.
Jeff Bezos optimizes for what won’t change, makes a small number of high-quality decisions, and uses “disagree and commit” to move fast.
Your job: define the “main thing,” pick controllable inputs, say “no” more often, and do less better.
Key Takeaways
Anchor on invariants: Build around customer needs that won’t change.
Fewer, better decisions: Senior leaders are paid to make a small number of high-quality decisions.
Inputs over outputs: Measure the levers you control weekly; outputs follow.
Visible tradeoffs: Every “yes” must come with a public “Not Doing List.”
Velocity by mechanism: Use “disagree and commit” when consensus stalls.
Quick Links
Next Read
When the leader finally says “no”
Leaders get inundated with ideas. The paradox: speed comes from subtraction, not addition.
Bezos reframed the classic strategy question, not “what will change?” but “what won’t change?” You can organize an entire operating model around stable truths (e.g., customers will always want lower prices, faster delivery, and broader selection). That’s where compounding lives.
He also insists leaders are paid to make a few high-quality decisions and keep decision velocity high, often by invoking “disagree and commit.” You don’t need universal buy-in; you need a clear owner and a date to move.
If your company’s drowning in ideas but starving for execution, this is your reset button.
Quick Win: Lock your ONE thing in 30 minutes
Set a 30-minute timer. Bring your core team.
Name the invariant (3–5 yrs): Write 1–3 sentences that won’t change for your customers (e.g., “fastest response in our category”).
Choose controllable inputs: Pick 3 weekly levers that deliver that promise (e.g., time to first reply, time to schedule, time to resolution). Review weekly.
Publish a Not Doing List: Kill or park anything that doesn’t strengthen those inputs.
Single-thread ownership: Assign one DRI per input; remove competing priorities.
Decide with speed: If you lack consensus, disagree & commit by a set date.
Recommended Books
Playbook: Do Less Better in 90 Days
1) Decide what won’t change (Week 1)
Run a 2-hour offsite. Identify three enduring customer promises (e.g., “lowest friction,” “fastest response,” “greatest clarity”). Publish the one-page Enduring Promises Doc (EPD).
2) Translate promises into inputs (Week 1)
For each promise, pick three input KPIs your team directly controls (e.g., time to first reply, time to schedule, first-contact resolution rate). Add weekly owners and review rhythm.
3) Make fewer, better decisions (Weeks 1–2)
Create a decision SOP: name the decider, time-box debate, document options/risks, then disagree & commit if needed. Measure Decision Velocity (request → decision).
4) Single-thread the mission-critical (Weeks 2–3)
For each input KPI, assign one DRI with calendar protection. Limit each DRI to ≤2 mission-critical projects at once.
5) Publish your Not Doing List (Week 2)
List the projects you are not doing this quarter. Tie each “yes” to a visible “no.” Update monthly.
6) Day-1 cadence (6-week sprints; repeat)
Adopt a Day-1 operating loop: plan (week 1), build (weeks 2–5), demo/retro (week 6). Keep it customer-obsessed and long-term oriented.
7) The 6-page memo for big bets (as needed)
For high-impact moves, write a narrative memo: problem, customer truth, 3 inputs, single-threaded owner, risks, kill criteria, and decision date.
8) Protect focus on the calendar (daily)
Leaders schedule 3 decision blocks/day; DRIs get uninterrupted deep-work blocks. The goal isn’t more decisions, it’s better ones.
Troubleshooting (Symptom → Cause → Fix)
Symptom: We keep starting, not finishing.
Cause: Too many priorities, split ownership.
Fix: Cap WIP; assign one DRI per input; publish the Not Doing List.
Symptom: Endless debate; no decision.
Cause: Consensus-seeking culture.
Fix: Time-box, name the decider, disagree & commit.
Symptom: We chase metrics we can’t move.
Cause: Output obsession.
Fix: Replace with controllable input metrics (e.g., response cycle-times).
Symptom: Fire drills every week.
Cause: No enduring promises.
Fix: Create the EPD (Enduring Promises Doc) and align roadmaps to it.
Symptom: Leaders are exhausted; choices are sloppy.
Cause: Too many decisions per day.
Fix: Limit to a few high-quality decisions at your best energy time.
Symptom: Strategy du jour.
Cause: No Day-1 operating model.
Fix: Install a 6-week Day-1 cadence; keep the customer obsession visible.
FAQ
Isn’t “do less better” risky in fast markets?
Sustained speed comes from clarity and repetition, not from adding projects. Bezos focuses on inputs and rapid course correction—being slow is costlier than being wrong.
How do we pick the “right” inputs?
Work backward from the promise until you hit a lever your team directly controls (e.g., reply time). If you can’t change it weekly, it’s not an input.
What if senior leaders disagree?
Debate is healthy; stalling isn’t. Name the decider and disagree & commit to preserve velocity.
How do I protect the main thing from “urgent” noise?
Public Not Doing List + calendar locks + DRI authority. Every new yes requires a visible no.
Is Day-1 just for startups?
No. Day-1 is a durable operating model: customer obsession, long-term orientation, and decisive execution.
Field Checklist
Three Enduring Promises published and visible.
Three input KPIs per promise; weekly review.
Single DRI per input; calendar protected.
Not Doing List updated monthly.
Decision SOP with disagree & commit.
Day-1 cadence running (6-week loop).
6-page memo template for big bets (with kill criteria, decision date).
Results & ROI
Leading indicators (inputs)
Time to First Response (goal: <10 minutes)
Decision Velocity (request → decision; goal: <72 hours)
WIP per DRI (goal: ≤2 mission-critical)
% Calendar Protected for DRIs (goal: 50–60%)
Lagging indicators (outputs)
CSAT/NPS trend, Repeat purchase/renewal rate, Cycle-time to resolution, Gross margin on service, Employee eNPS
Example 12-week lift
Decision Velocity: 11.4 days → 2.9 days
Time to First Response: 54 min → 8 min
WIP per DRI: 6.2 → 2.1
On-time commitments: 61% → 88%
This mirrors Bezos’s principle: improve controllable inputs and outputs follow.
Wrap-Up
Pick one enduring promise, one input, one owner. Kill two nice-to-haves. Announce the decision, disagree & commit, and move. The room gets quiet when the leader finally says “no.” That’s when momentum starts.
Comments