# Pre-Mortems: Plan for Success Before You Start! Newsy: Dev Concepts Weekly Published: 2025-12-10T14:00:00.940Z Source: https://newsyai.com/edition/47634172-2225-4c0f-8fa2-734495d1943d Archive: https://newsyai.com/newsy/249754d5-9f20-4825-9aeb-daeec7c365c9.md --- # The Pros Know the Fundamentals Better! Daily Edition — December 10, 2025 > “Design for how things fail, not just how they work.” Three high‑impact topics (about 90 seconds each) to compound your growth. Plain English, vivid metaphors, practical moves. --- ## 1) Pre‑Mortems: Plan the Crash Before Takeoff Most teams do postmortems after the damage. Pros run a pre‑mortem before the project starts: imagine it’s six weeks from now and the effort failed—why? Then design the countermeasures now. Why this compounds - **Sharper focus:** You surface the few risks that can actually sink you, not the hundred that just slow you down. - **Better estimates:** Naming the unknowns turns fog into tasks, which tightens timelines. - **Calmer execution:** You won’t be improvising seatbelts at 30,000 feet. Metaphor - Think of a fire drill. You don’t learn how to exit a building when it’s full of smoke—you practice when it’s bright and quiet. How to run it (45 minutes) 1. Set the scene: “It failed.” Everyone writes 3–5 reasons in silence. 2. Cluster reasons: tech, people, timeline, dependencies. 3. Pick the top 5 risks by impact x likelihood. 4. Assign a countermeasure per risk with an owner and an “early smoke” signal. 5. Log them in a one‑pager that lives with the project plan. Try this today - For your next feature, run a 30‑minute pre‑mortem and create a “Top 5 Risks + Owners” card. Review it at each stand‑up like a flight checklist. --- ## 2) Backpressure: Don’t Outrun Your Brakes Systems fail when they accept work faster than they can finish it. Backpressure is the habit of saying “not so fast” upstream when you’re near your limits. It prevents pileups, protects downstream services, and turns chaos into steady flow. Why this compounds - **Fewer meltdowns:** You trade catastrophes for graceful slowdowns. - **Predictable latency:** A little rejection now avoids a lot of timeout later. - **Happier users:** Honest “please wait” beats silent spinning and lost work. Metaphor - Picture a busy restaurant. Without a host limiting seating, the kitchen drowns and everyone eats late. The host is backpressure. Design moves - Set a **queue size** limit; when full, shed or defer requests. - Use **leaky buckets/tokens** to pace work—small, steady sips instead of gulps. - Return **clear signals**: quick “try again” or “we’ll notify you,” not vague slowness. - Prefer **lossy overflow for non‑critical work** (e.g., analytics) and **reservations** for critical paths (e.g., payments). - Expose **capacity hints** to callers so they naturally smooth their send rate. Try this today - Add a hard cap to one hot queue and a friendly “busy—retry after N seconds” response. Track how many requests you save from timing out. --- ## 3) SLOs and Error Budgets: Make Reliability a Product Choice Availability isn’t a sticker—it's a promise you must keep. A Service Level Objective (SLO) defines the reliability users should expect (e.g., 99.9% success over 30 days). The **error budget** is the allowed failure within that promise. It turns reliability into a clear tradeoff with speed. Why this compounds - **Aligned priorities:** When you “spend” the budget, everyone understands why new features pause. - **Less arguing, more deciding:** Numbers replace vibes. - **User‑centered:** You build the right amount of reliability—no more, no less. Metaphor - Think of a household budget. If you’ve spent the month’s dining‑out money, you cook at home. Blow the error budget? You “cook” reliability work until you’re back on plan. How to set SLOs - Start with the **critical user journey** (e.g., “checkout succeeds within 2s, 99.9% of the time”). - Measure from the **user’s point of view** (end‑to‑end), not just server health. - Choose a **review window** (14–30 days) and track weekly. - Define **guardrails**: if the budget is >50% spent mid‑window, freeze risky launches and fix the top failure mode. Try this today - Pick one flow, set a lightweight SLO, and put the error budget as a number on your team dashboard. Use it to decide this week’s tradeoffs. --- ### Quick Recap - Run pre‑mortems to turn surprise into a plan. - Build backpressure so your system brakes as well as it accelerates. - Use SLOs and error budgets to make reliability an explicit, shared choice. Keep compounding: small, boring habits—applied consistently—turn into elite performance.