Stages describe a process, not a commitment.
Rework your stage definitions and deals will sit in better buckets. A bucket still can't say whether the person who owns the deal believes it lands this quarter.
Forecasting
Commit to a number you can defend.
See every forecast side by side: CRM, Seller, Manager and leadership. Model the quarter without touching the live forecast, and keep every version of what was said.
Before you commit the number
Test the loss you are afraid of, while you can still act.
Move the deals you are least sure about and watch the number move with them. Demonstration data throughout.
Q3 · EMEALeadership forecast
Demonstration dataCurrent forecast£7.2mBase scenario
The gap
The CRM shows what was entered.PROOF shows what each level is forecasting, and where they disagree.
Your VP walks into the pipeline review with a number. The CRM shows a different one. Fifteen minutes of a forty-five minute meeting go into working out which is real, and you never fully settle it.
That is not a data quality problem. It is a missing object.
Q3 · EMEA · New business · Demonstration dataFour numbers. None of them wrong.
Positions differ| Position | Detail | Value | Flag |
|---|---|---|---|
| CRM position | Source field · last received 11 days ago | £2,480,000 | Stale |
| Seller Working Forecast | Authored by R. Adeyemi · unpublished change | £2,150,000 | |
| Manager Published Forecast | Authored by D. Okonjo · sealed 4 Sept | £1,920,000 | |
| Leadership Published Forecast | Authored by S. Hart · sealed 5 Sept | £1,800,000 |
Two different things
A current field and an authored forecast answer different questions.
A current value can support a roll-up. PROOF also keeps the Seller, Manager and leadership positions separate, with their authors, reasons and published history.
| Question | A current CRM field on its own | A forecast in PROOF |
|---|---|---|
| Who wrote it | The current value; history may be held separately | A named person, at a named level, at a known time |
| What it records | What was entered | What that person is prepared to commit to |
| When it was last true | A field update does not, by itself, date the underlying buyer claim | Stated: field age, receipt time and revision |
| Who may change it | Depends on configured permissions | Its author owns the position; a Manager publishes a separate one |
| What a disagreement means | Requires a review across values and context | A working moment with an owner, deals behind it and a decision to make |
| What happens to last quarter's number | Requires a dated snapshot or forecast history | Published positions remain readable exactly as they stood |
The structural gap
Clean fields help. Separate positions explain the commitment.
Better CRM hygiene matters. PROOF adds named Seller, Manager and leadership positions, with the reasons and published history behind each one.
What gets logged is what went well.
Manual logging is selective by design, so the record leans positive. Sync every mailbox and you get volume, not confirmation. Activity is easy to generate. A buyer confirming something is not.
Each forecast position has an author.
PROOF preserves separate working and published positions. A Manager's view can differ from a Seller's without overwriting it, and the difference can be traced to the deal.
One forecast workspace
See the forecasts side by side.
Keep CRM, Seller, Manager and leadership positions distinct. Use your own categories and stage weightings, then model the quarter without changing the live forecast.
DealCRMSellerManagerLeadershipWeighted ACVClose moved
NorthstarBest CaseCommitBest CaseUpside£175k2 times
HeliosCommitCommitCommitCommit£320k0 times
MeridianBest CaseBest CaseUpsideUpside£90k4 times
From one published forecast to the next
Then the movement explains itself.
Ask what changed since the last published forecast and you get the whole movement, classified, down to the contributing deals. Timing, value, category, new and removed contribution and disclosed adjustments are separate from one another.
Target, hierarchy and policy changes are held apart from deal movement, so a re-org never masquerades as pipeline slipping.
The residual is zero, or the product says so. There is no remainder line nobody can account for.
Manager Published · 4 Sept to 18 Sept · Demonstration dataEvery movement reconciles.
| Line | Deals | Amount |
|---|---|---|
| Opening published position | £1,920,000 | |
| Closed Won actuals | 3 deals | +£240,000 |
| Timing moved to Q4 | 2 deals | −£310,000 |
| Category and inclusion changes | 4 deals | −£95,000 |
| Value and allocation | 6 deals | +£107,000 |
| New contribution | 1 deal | +£83,000 |
| Disclosed adjustment withdrawn | 1 decision | −£100,000 |
| Closing published position | £1,845,000 | |
| Unexplained residual | £0 |
The forecast loop
Six questions, answered from one record.
Without translating internal terminology, opening a second system, or rebuilding the number in a spreadsheet first.
- 01What is our number?
- 02What supports it?
- 03What threatens it?
- 04What changed?
- 05What did we decide?
- 06Did we follow through?
The Forecast Review
Find out where last quarter's forecast moved away from reality.
Compare the positions your Sellers, Managers and leaders submitted with the outcome. See the direction and size of the error across teams and segments.
- What each level forecast, and what landed
- Where the error entered, by team, segment and deal size
- Whether the same bias appeared more than once
The arithmetic shows where the error entered. Explaining why requires the source, decision and timing behind each forecast.
Request a Forecast ReviewA defensible number is not produced by a forecast tool.
It is produced by three people doing different jobs on the same record.
See the number defended