Pipeline Manager and forecast inspection
Each deal is scored 0–3 on Metrics, Economic buyer, Decision criteria, Decision process, Identified pain and Champion.
| Deal | Stage | Amount | Close | M·E·D·D·I·C | Score |
|---|---|---|---|---|---|
Northwind security program Northwind Logistics | Discovery | $48k | Q1 | 2 · 1 · 1 · 0 · 2 · 1 | 39% |
Bluepeak SOC 2 readiness Bluepeak Health | Proposal | $72k | Q1 | 3 · 2 · 2 · 2 · 3 · 2 | 78% |
Copperline vendor risk Copperline Fintech | Qualification | $30k | Q2 | 1 · 0 · 1 · 0 · 1 · 0 | 17% |
Quillstone pilot Quillstone Software | Negotiation | $55k | Q1 | 3 · 3 · 2 · 3 · 3 · 3 | 94% |
Stage conversion (sample cohort)
| Stage | Deals that reached it | % to next stage |
|---|---|---|
| Discovery | 40 | 60.0% |
| Solution fit | 24 | 58.3% |
| Proposal | 14 | 64.3% |
| Negotiation | 9 | 55.6% |
| Closed won | 5 | — |
Overall win rate from Discovery: 12.5% (5 of 40).
Deal aging (open deals, days since created)
0–30 days6
31–60 days4
61–90 days3
90+ days2
15 open sample deals. Anything 90+ days gets reviewed: requalify or close.
Metric definitions
| Metric | Definition |
|---|---|
| Stage conversion | Deals that reached the next stage ÷ deals that reached this stage, for the same cohort |
| Win rate | Closed won ÷ all deals that entered Discovery in the cohort |
| Deal age | Today minus the deal's created date, for open deals |
| Weighted pipeline | Sum of amount × stage probability for open deals |
| MEDDIC score | Sum of the six 0–3 scores ÷ 18, as a percentage |
Forecast reliability: what it can and can't tell you
Can: show where deals stall, which stages leak, and whether the open pipeline roughly covers the target.
Can't: predict a specific deal, or be trusted with a sample this small. Five wins over two quarters gives wide error bars, so a single deal swings the rates. Stage probabilities are set by hand, not learned from history.
I treat it as a prompt for questions in pipeline review, not a number to commit to.
Conversion SQL
Learning / reference-- Stage-to-stage conversion for a cohort of deals. Learning example on sample tables.
WITH reached AS (
SELECT stage, COUNT(DISTINCT deal_id) AS deals
FROM deal_stage_history
WHERE entered_at >= DATE '2026-04-01'
GROUP BY stage
),
ordered AS (
SELECT r.stage, r.deals, s.sort_order,
LEAD(r.deals) OVER (ORDER BY s.sort_order) AS next_deals
FROM reached r
JOIN stages s ON s.name = r.stage
)
SELECT stage, deals,
ROUND(100.0 * next_deals / NULLIF(deals, 0), 1) AS pct_to_next_stage
FROM ordered
ORDER BY sort_order;