تجاوز إلى المحتوى
جرّب المرجع الدورة المحفوظات
العربية
← كل playbooks المبيعات

playbook

The weekly pipeline review

Clean a messy deals export, standardize stages, flag every deal gone quiet for 30 days, total the pipeline by stage, and get a skeptical read on which deals are real.

متوسّط ~40 min

متى تلجأ إلى هذا

Your pipeline is only as good as your honesty about it — and most pipelines are full of stale deals, inconsistent stage names, and "happy-ears" forecasts that quietly slip every quarter. This is a weekly ritual: take the messy export, clean and standardize it, surface the deals that have gone quiet, total it by stage, and then ask Claude for the read you don't want to hear — which deals are actually real and which are wishful thinking. Run it every Friday and your forecast stops lying to you.

جهّز هذا أولًا

  • A fresh deals-export.csv from your CRM — deal name, stage, amount, close date, owner, and last-activity date at minimum.
  • Your canonical stage list — stages.md with the official stage names in order (e.g. Discovery → Demo → Proposal → Negotiation → Closed) — so Claude maps the mess to one standard.
  • Last week's totals (or last review's output), so you can compare and spot what moved, slipped, or stalled.

الـ workflow

  1. Clean and standardize the export — and reconcile the total

    Messy data produces a confident-but-wrong forecast. Standardize stage names and dates first, and check the cleaned total against the raw total so nothing got dropped in the cleanup.

    أنت تطلب 
    Read deals-export.csv and stages.md. Standardize every stage name to the canonical list in stages.md (map variants like "Negotiating"/"Negotiation" to one), parse all dates to YYYY-MM-DD, and flag any row with a missing amount, stage, or close date. Then confirm: how many deals and what total amount, both before and after cleanup — they should match.

    ما تحصل عليه A cleaned table plus a reconciliation line — "Before: 47 deals, $1.24M. After: 47 deals, $1.24M (3 rows flagged for missing amounts)." If the totals don't match, you know cleanup dropped something.

    Always reconcile the cleaned total against the known raw total — it's the one check that catches a silent data loss before it becomes a wrong forecast.

  2. Flag the deals that have gone quiet

    Stale deals are the silent killers of a forecast. Surface everything with no activity in 30 days so you act on it instead of carrying it as if it's live.

    أنت تطلب 
    Flag every deal with no activity in the last 30 days. Sort by amount, biggest first, and for each show: deal, owner, stage, amount, days since last touch. Then list the top 5 I should re-engage or kill this week.

    ما تحصل عليه A stale-deals table sorted by value ("Acme renewal — $80k — 41 days quiet") and a short hit-list of the five worth your attention now — so dead weight gets addressed, not ignored.

  3. Total the pipeline and weight it honestly

    Get the by-stage totals, then a weighted view — but treat the weighting as a prompt for judgment, not a forecast you trust blindly.

    أنت تطلب 
    Total the cleaned pipeline by stage: count and dollar amount per stage. Then show a weighted view using these probabilities [Discovery 10%, Demo 25%, Proposal 50%, Negotiation 75%]. Call out where the pipeline is top-heavy (lots of value stuck early) versus close to closing.

    ما تحصل عليه A by-stage table with raw and weighted totals ("Proposal: 8 deals, $420k raw, $210k weighted") and a shape read — "60% of value is still in Discovery; the quarter depends on 3 Negotiation deals."

  4. Get the skeptical read on what's actually real

    This is the step the ritual exists for. Ask Claude to play a hard-nosed VP and challenge the deals you're counting on — happy-ears doesn't survive a skeptic.

    أنت تطلب 
    Play a skeptical VP of Sales reviewing this pipeline. Which deals look like happy-ears — high amount or near close date but thin recent activity, stuck in one stage too long, or no clear next step? Name the 5 most at-risk "real" deals and what single piece of evidence would prove each is genuinely live.

    ما تحصل عليه A pointed risk list — "The $120k 'Negotiation' deal hasn't moved in 5 weeks and has no next meeting booked — that's not Negotiation, that's stalled" — plus the one proof point that would change your mind on each.

اجعله ملكك

  • **Stalled deal you want to save:** for any deal the review flags, run *The pre-call account brief* to prep a re-engagement, then *The personalized outreach engine* to write the touch.
  • **Renewals view:** filter the export to renewal deals and feed at-risk ones into *The renewal play* — the review is how you spot which renewals need the full play.
  • **Run it on a schedule:** save the four steps as a /pipeline custom command and let a scheduled agent prep the cleaned review off a fresh export every Friday morning (see the *Features* tab).

انتبه إلى

  • Reconcile the cleaned numbers against a known total every time — if the post-cleanup pipeline total doesn't match the raw total, Claude dropped or double-counted rows, and a wrong total here misleads the whole forecast. Don't trust the math without the check.
  • A deals export is confidential commercial data — amounts, names, close dates. Keep deals-export.csv in your approved workspace and never paste your pipeline into a tool you don't control; that's competitive information.
  • Claude can flag risk and challenge happy-ears, but it can't know your prospect's intent. The skeptical read is a prompt for your judgment, not a verdict — you own the forecast number you commit to, and a human decides which deals to call real.

ستحصل في النهاية على A cleaned, standardized, reconciled pipeline with stale deals flagged, honest by-stage totals, and a skeptical read on which deals are real — a forecast you can actually trust, refreshed weekly.