Pipeline to plan
Most plans are set top-down and reconciled to the pipeline afterwards. This one runs the other way: revenue month by month, built from what is contracted, what customers are due to renew, the deals already in the pipeline — and the pipeline sales has not created yet. Three cases run side by side, each on its own drivers.
Two things change when a machine does the arithmetic. The forecast rolls at every month-end instead of being rebuilt once a year, so the horizon never shortens. And the pipeline is weighted deal by deal, on its own history, rather than at fixed stage probabilities that let stale deals prop up the plan. Close the first six months and see which view the actuals agree with.
Three cases · full year
Realistic case · Start of the year
Versus target
+€923k
above the €22m target
Secured
49%
recognised or under contract
Not yet created
€1.1m
of the year, from July on
Monthly revenue by source · Realistic
VerdictStage averages flatter the year
On the realistic case the year lands €923k above target, but the conservative case misses it.
Recognised and contracted revenue covers 49% of the year; from July, part of every month rests on pipeline nobody has created yet — €1.1m of the year, more than the headroom. Scoring each deal would take €1.6m off the year: stage averages give every Commit deal the same odds, stale ones included.
Drivers · Realistic
Each case keeps its ownScales every win probability. 1.00× is history: 27% of new pipeline won.
Creation to close. Open deals close what is left of it; new pipeline lands a full cycle out.
Annual contract value created each month, across every region.
Share of contracts that reach the end of their term and do not renew.
The band spans the three cases, conservative to aggressive; the bold tick is realistic. Two of the three cases clear it.
Open pipeline · Start of the year
Realistic win rates| Stage | Open | Stage avg | Scored |
|---|---|---|---|
| Qualified5 deals | €1.9m | 20% | 13% |
| Proposal4 deals | €1.8m | 40% | 34% |
| Negotiation8 deals | €3.1m | 60% | 59% |
| Commit10 deals · €560k overdue | €4.4m | 90% | 50% |
| Weighted | €11.1m | €6.9m | €4.9m |
Stage averages give every deal in a stage the same odds, including Commit deals already past their close date. Deal scores come from a model trained on closed deals: it reads each deal on its own signals and has learned how rarely overdue deals land.
The year by source · Realistic
Start of the year| Source | Year | Share |
|---|---|---|
| Order book | €11.30m | 49% |
| Renewals | €5.14m | 22% |
| Open pipeline | €5.36m | 23% |
| Not yet created | €1.13m | 5% |
| Full year | €22.92m | 100% |
Month by monthEvery month on the chart as a schedule, in €m
| Month | Order book | Renewals | Open pipeline | Not yet created | Realistic | Range |
|---|---|---|---|---|---|---|
| Jan | 1.43 | — | — | — | 1.43 | 1.43–1.43 |
| Feb | 1.31 | 0.11 | 0.08 | — | 1.50 | 1.48–1.51 |
| Mar | 1.23 | 0.18 | 0.29 | — | 1.69 | 1.61–1.73 |
| Apr | 1.18 | 0.23 | 0.46 | — | 1.86 | 1.70–1.92 |
| May | 1.00 | 0.38 | 0.53 | — | 1.92 | 1.75–1.99 |
| Jun | 0.98 | 0.40 | 0.56 | — | 1.94 | 1.80–2.10 |
| Jul | 0.87 | 0.49 | 0.57 | 0.05 | 1.99 | 1.80–2.17 |
| Aug | 0.81 | 0.54 | 0.57 | 0.11 | 2.03 | 1.80–2.23 |
| Sep | 0.78 | 0.57 | 0.57 | 0.16 | 2.08 | 1.83–2.30 |
| Oct | 0.72 | 0.62 | 0.57 | 0.22 | 2.12 | 1.85–2.36 |
| Nov | 0.52 | 0.79 | 0.57 | 0.27 | 2.15 | 1.84–2.42 |
| Dec | 0.46 | 0.84 | 0.57 | 0.32 | 2.19 | 1.87–2.48 |
| Jan +1 | 0.46 | 0.84 | 0.57 | 0.38 | 2.24 | 1.91–2.55 |
| Feb +1 | 0.41 | 0.87 | 0.56 | 0.43 | 2.27 | 1.91–2.61 |
| Mar +1 | 0.39 | 0.88 | 0.54 | 0.49 | 2.29 | 1.91–2.66 |
| Apr +1 | 0.30 | 0.96 | 0.52 | 0.54 | 2.31 | 1.91–2.70 |
| May +1 | 0.30 | 0.94 | 0.51 | 0.59 | 2.35 | 1.92–2.76 |
| Jun +1 | 0.24 | 1.00 | 0.50 | 0.65 | 2.39 | 1.93–2.82 |
Notes on the model
What it shows
- How much of the plan rests on revenue nobody has signed. The hatched wedge is pipeline that does not exist yet, and it grows the further out you look.
- A rolling forecast re-bases on actuals at every close: the cases converge and the horizon keeps its eighteen months. A budget set in January only gets shorter.
- Stage averages overstate the pipeline when stale deals sit in late stages. Deal scores show it at the start of the year; the actuals confirm it by the summer.
How it works
- Revenue is recognised monthly. Signed contracts run to the end of their term, then renew at one minus churn, term after term.
- Open deals are weighted by their chance of closing and land after the unexpired part of the sales cycle, give or take a month. New pipeline is won at the historic rate, a full cycle after it is created.
- Behind the forecast sits a seeded simulation of the business — contracts, renewals, deals won, lost and stalled — which supplies the actuals as each month closes.
- Deal scores stand in for a trained classifier: they read each deal’s quality imperfectly and have learned how rarely overdue deals land.
What it leaves out
- Every figure is illustrative, for a mid-sized business-to-business company. Contracts, deals and the model’s errors are all simulated.
- No price rises, expansion or seasonality, and every new deal is a twelve-month contract that starts the month after it closes.
- The cases differ only in their drivers. Real downside cases also move discounting, timing of collections and cash — none of which this model touches.
Next experiment · 04
The variance writes back
Budget versus actuals in, board-ready commentary out. Tune materiality and audience.
Draft the commentaryUsing this on a real decision?
Talk it through
The defaults are illustrative. The inputs that decide the answer are the ones you only know from inside the business.
Get in touch