On this page
Key takeaways
- One deals export from your CRM is enough: stage, amount, owner, created date, close date and outcome.
- Look at conversion between stages, not only total pipeline.
- Win rate and cycle time by deal size reveal where large deals stall.
- End each review with one change to how the team sells.
The data you need
Export deals (or opportunities) from your CRM. CRMLeaf, HubSpot and Salesforce can all export these fields:
| Column | Example | Notes |
|---|---|---|
deal_id | D-4821 | One row per deal |
deal_name | Acme renewal | Optional |
owner | Priya S. | Sales rep |
stage | Proposal | Current stage |
amount_usd | 18,000 | Deal value |
created_date | 2026-06-02 | When the deal was created |
close_date | 2026-08-28 | Actual or expected close |
outcome | won | won, lost or open |
source | Inbound | Optional: lead source |
If your CRM can export stage history (the date each deal entered each stage), include it. It makes stage conversion and time-in-stage exact.
Question 1: What does the pipeline look like now?
What you get: a bar chart of open pipeline by stage.
What to look for: a bulge in one early stage (plenty of deals, not moving) or a thin late stage (not enough deals to hit the number).
Follow-up: "Split this by owner" to see whose pipeline is thin.
Question 2: Where do deals drop out?
Ask: "For deals created in the last 6 months, what share reached each stage?"
What you get: stage-by-stage conversion, for example 100% at Qualified, 55% at Demo, 30% at Proposal, 18% Won.
What to look for: the largest drop between two stages. That is where the sales process leaks.
Follow-up: "Compare this by lead source." Inbound and outbound deals often leak at different stages.
Question 3: Is win rate improving?
Ask: "Win rate by quarter for the last four quarters, counting only closed deals."
What you get: a line or bar chart of won deals divided by closed (won plus lost) deals, per quarter.
What to look for: a falling win rate while pipeline grows. That usually means more deals of lower quality, not a better quarter.
Note: count only closed deals. Including open deals makes recent quarters look worse than they are.
Question 4: How long do deals take to close?
Ask: "Median days from created to won, by deal size: under $5k, $5k to $25k, over $25k."
What you get: cycle time by deal size band.
What to look for: a sharp jump for larger deals. If deals over $25k take three times as long, forecast them later and start them earlier.
Follow-up: "Which open deals over $25k have been open longer than the median?" Those are your stalled deals.
Question 5: Where does each rep win?
Ask: "Win rate and average deal size by owner for the last two quarters."
What you get: a table or chart per rep.
What to look for: reps with high win rates on large deals (their approach is worth sharing) and reps with high volume but low win rate (coaching opportunity). Use this for coaching, not ranking: small samples swing a lot.
Put it on one canvas
Arrange the five answers in order: pipeline now, stage drop-off, win rate trend, cycle time, rep view. Add one note at the top with the change you will make this month, such as "Add a technical call before Proposal for deals over $25k."
Common traps
| Trap | Effect | Fix |
|---|---|---|
| Including open deals in win rate | Recent win rate looks low | Closed deals only |
| Expected close dates treated as actual | Cycle time too short | Use won or lost deals for cycle time |
| Stages renamed mid-year | Conversion breaks | Map old names to new before upload |
| Small samples per rep | Misleading rankings | Look at two quarters or more |
| Duplicate deals | Pipeline inflated | Remove duplicates by deal ID |