Stage weighting is the most widely used forecasting method in B2B SaaS and one of the most widely blamed. The blame is usually misdirected: the method is not broken, the implementation almost always is, and it fails in two specific ways.
Failure one: subjective stages
Stage weighting can work reasonably well if there is strict entry and exit criteria for each stage.
If the business is not disciplined about those criteria, stages become subjective and left to the salesperson's discretion. And whenever you apply an objective value to a stage that is determined subjectively, you will get unexpected outcomes at the end of the quarter. You will miss.
That sentence is the whole mechanism. A 60 percent weight is a precise claim about a population of deals. It only means anything if membership in that population is decided the same way every time, by every rep. Where entry is a judgment call, the weight is being applied to a category whose contents vary by who filled it.
The failure is invisible in aggregate for a while, because optimistic and pessimistic reps offset each other. It surfaces at quarter end, when the offsetting stops being reliable.
Failure two: one set of weights for everything
The second reason is more mechanical and easier to fix.
New business, expansion and renewal should all be in different pipelines with different weights. They usually are not. If you have an enterprise motion and a commercial or SMB motion the stage weights should differ, but they do not in practice. If you sell multiple products, the weights could differ again, and again they do not.
This is not a subtle modeling preference. A renewal at the negotiation stage and a new-business deal at the negotiation stage have very different probabilities of closing, and very different amounts at risk. Averaging them produces a weight that describes neither.
| Motion | Why its weights differ |
|---|---|
| New business | Longest path, highest variance, close date driven by stage progression |
| Expansion | Existing relationship, shorter path, different failure modes |
| Renewal | Least uncertainty, amount usually known, timing driven by renewal date |
| Enterprise vs commercial | Different cycle lengths and different stage meanings |
| Multi-product | Different buying committees per product |
Why the amount compounds the error
Both failures multiply against a number that is itself usually wrong.
Weighted forecast is the stage weight applied to the opportunity amount, and most deals close for less than the value recorded in the CRM. A pipeline averaging $80,000 against closed-won deals averaging $40,000 means the weight is being applied to a figure roughly double its realizable value. See why your pipeline average deal size lies.
The result looks methodical and is wrong in two dimensions at once, which is worse than being obviously wrong, because it does not invite scrutiny.
Fixing it in the right order
Split before you tune. Separate the pipelines by motion, then by segment. This usually improves accuracy more than any amount of re-tuning a blended set, and it requires no behavior change from the field. Write real entry and exit criteria. Observable, verifiable, and the same for everyone. The test is whether two managers looking at the same deal would place it in the same stage. Re-derive the weights from your own conversion data, per pipeline, rather than carrying forward numbers nobody remembers setting. Apply a realization factor to the amount so the weight operates on a value the deal is likely to close at. Then check whether you still need weights. Once deals are grouped by behavior, a learned close curve for each group describes timing and probability together, which stage weights cannot do. See how long deals take to close by group.Stage weighting is a reasonable approximation applied to a well-defined population. Most implementations fail the second half of that sentence. For definitions see stage-weighted forecasting and unweighted pipeline.
Frequently Asked Questions
Why does stage-weighted forecasting fail?
Two reasons. Weights only hold when every stage has strict entry and exit criteria, and without that discipline the stage becomes subjective. Applying an objective value to a subjectively determined stage produces unexpected outcomes at quarter end. Second, most businesses apply the same weights to new business, expansion and renewal, and across segments that behave differently.Can stage weighting work at all?
Yes. It can work fine where entry and exit criteria are strict and consistently applied, and where separate weight sets exist for each motion and segment. The method is not broken, the usual implementation is.What is the fastest fix?
Split the weights before improving them. Separate new business, expansion and renewal into different pipelines with different weights, and separate enterprise from commercial. That change usually improves accuracy more than re-tuning a single blended set.What is the fastest improvement to a weighted forecast?
Split the pipelines before tuning the weights. Separating new business, expansion and renewal, then enterprise from commercial, usually improves accuracy more than re-tuning a blended set, and it requires no behavior change from the field.How do I know if my stage criteria are strict enough?
Two managers looking at the same deal should place it in the same stage. Where they would not, the stage is a judgment call and any weight applied to it is being applied to a category whose contents vary by who filled it.Frequently Asked Questions
Why does stage-weighted forecasting fail?
Two reasons. Weights only hold when every stage has strict entry and exit criteria, and without that discipline the stage becomes subjective. Applying an objective value to a subjectively determined stage produces unexpected outcomes at quarter end. Second, most businesses apply the same weights to new business, expansion and renewal, and across segments that behave differently.
Can stage weighting work at all?
Yes. It can work fine where entry and exit criteria are strict and consistently applied, and where separate weight sets exist for each motion and segment. The method is not broken, the usual implementation is.
What is the fastest fix?
Split the weights before improving them. Separate new business, expansion and renewal into different pipelines with different weights, and separate enterprise from commercial. That change usually improves accuracy more than re-tuning a single blended set.
What is the fastest improvement to a weighted forecast?
Split the pipelines before tuning the weights. Separating new business, expansion and renewal, then enterprise from commercial, usually improves accuracy more than re-tuning a blended set, and it requires no behavior change from the field.
How do I know if my stage criteria are strict enough?
Two managers looking at the same deal should place it in the same stage. Where they would not, the stage is a judgment call and any weight applied to it is being applied to a category whose contents vary by who filled it.
See how ORM turns these insights into action
ORM builds custom revenue forecast models for B2B SaaS companies. Not dashboards. Prescriptive analytics that tell you what to do next.
Schedule a Demo