تجاوز إلى المحتوى
جرّب المرجع الدورة المحفوظات
العربية
← كل playbooks المؤسّسون ومديرو المنتج

playbook

Turn options into a decision you can defend

Lay tradeoffs side by side and get a recommendation with its reasoning and the biggest risk of each path — so you're sanity-checking a call, not starting one cold.

متوسّط ~30 min

متى تلجأ إلى هذا

You've got three plausible paths — build vs buy, this vendor vs that one, ship now vs wait a quarter — and a doc full of half-comparable notes. Decisions made from that pile tend to follow whoever argued last or loudest. This system turns the options into a structured comparison, then a recommendation with explicit reasoning and the biggest risk of each path. The goal isn't to outsource the call — it's to walk in already having stress-tested it, so you're sanity-checking a recommendation rather than starting from a blank page under pressure.

جهّز هذا أولًا

  • The options written down — a options.md with the candidates and whatever you know about each (cost, time, who'd own it). Rough is fine; Claude will flag the gaps.
  • The decision criteria that actually matter *to you* — "we optimize for shipping speed this quarter, then cost, then flexibility" — so the recommendation is weighted your way.
  • Any hard constraints — a budget ceiling, a deadline, a must-have integration — that rule options in or out before reasoning starts.

الـ workflow

  1. Have Claude restate the decision and its gaps

    Before comparing anything, make Claude restate what's actually being decided and flag what's missing. A recommendation built on a misread question or blank cells is worse than no recommendation.

    أنت تطلب 
    Here are our options in options.md. Before recommending anything, tell me: what decision are we actually making, what are the real options, and what key information is MISSING that I'd need to choose well? Don't pick yet.

    ما تحصل عليه A clean restatement plus a gap list — "you're deciding whether to build invoicing in-house or use a vendor; missing: the vendor's per-transaction fee at our volume, and how much eng time the build really costs." Fill the gaps before trusting any recommendation.

    A recommendation over missing data is just a confident guess. Make Claude name the blanks first.

  2. Lay the tradeoffs side by side

    A comparison table against *your* criteria forces every option to be judged on the same axes, instead of each being described in its own flattering terms.

    أنت تطلب 
    Now lay the options side by side as a table, scored against our criteria in priority order: shipping speed, then cost, then flexibility. One row per option, one column per criterion, plus a short 'why' in each cell. Be honest where an option is weak — don't flatten the differences.

    ما تحصل عليه A scannable comparison table — "Build: slow to ship, low ongoing cost, max flexibility; Vendor A: fast, high per-txn cost, limited customization" — with a one-line why in each cell. The differences are visible at a glance.

  3. Get a recommendation with its reasoning and each path's biggest risk

    The reasoning is the point. A recommendation you can't see the logic of is one you can't defend in the room — and the per-option risk is what you'll be asked about.

    أنت تطلب 
    Given our priorities, recommend one option. Walk me through WHY it wins on our criteria, what we're trading away by picking it, and the single biggest risk of EACH option if we chose it. Then tell me what would have to be true for your recommendation to flip.

    ما تحصل عليه A reasoned pick — "recommend Vendor A: it's the only path that ships this quarter, our top priority; we trade ~$X/yr and some flexibility. Biggest risks — Build: eng time overruns; Vendor A: lock-in; Vendor B: integration gap. This flips to Build if speed stops being priority one."

    The "what would flip this" line is the most useful sentence in the memo — it tells you exactly which assumption to pressure-test before committing.

  4. Red-team it, then write the memo

    Have Claude argue against its own recommendation before you package it. A pick that survives its own strongest counter-argument is one you can actually defend.

    أنت تطلب 
    Now argue the strongest case AGAINST your recommendation — what would a smart skeptic say? Then assemble a one-page decision memo: the decision, the options table, the recommendation with reasoning, the biggest risk of each path, and the open questions I should resolve before committing.

    ما تحصل عليه A short steelman of the opposing view, then a clean one-page memo you can circulate — decision, table, reasoned recommendation, per-path risks, and open questions. Defensible because it already faced its own best counter-argument.

اجعله ملكك

  • **Build vs buy on your own stack:** when an option is "build it ourselves," run **Read your codebase in plain English** first so the build estimate is grounded in what it'd actually touch, not a guess.
  • **Higher stakes:** for a big, irreversible call, have Claude run the comparison from two opposing personas (e.g. a cautious CFO and a growth-hungry founder) and reconcile where they disagree.
  • **Reusable memo shape:** if you make these calls often, save the structure as a /decision-memo custom command (see the *Features* tab) so every decision gets the same rigorous treatment.

انتبه إلى

  • Claude recommends; you decide. The memo is a sanity-check on a call you own — its recommendation is an input weighted by *your* priorities, never the verdict. Don't outsource the judgment to the confident paragraph.
  • Verify the numbers driving the choice. A recommendation built on a wrong cost or a hallucinated vendor fee is wrong no matter how clean the reasoning — check the figures in the table against the source, not Claude's recall.
  • If the options involve confidential terms — vendor pricing, partner contracts, financials — replace specifics with [redacted] or keep it in your sanctioned environment; a decision memo is exactly the kind of doc that gets forwarded.

ستحصل في النهاية على A one-page decision memo — options compared on your criteria, a recommendation with its reasoning, the biggest risk of each path, and what would flip the call — so you commit to a decision you've already stress-tested, not one you're making in the room.