For forecasting, the most useful data in a CRM is often the change in the data.
A CRM tells you where things stand right now. It usually forgets how they got there. Every time a rep updates a stage, a close date or an amount, the old value is overwritten. That history is exactly what a forecast needs.
What does a change tell you that a field does not?
A deal sitting in proposal is useful to know. Knowing it moved into proposal yesterday is more useful. The same goes for the amount. The amount matters, but the fact that it just changed can be a stronger signal. Did it go up or down? By how much? How late in the sales cycle did it change?
Those changes capture behavior. They show how a deal is actually progressing, which is the thing a forecast is trying to predict.
| What changed | What it can tell you |
|---|---|
| Stage moved forward | The deal is progressing, and how fast compared with similar deals |
| Stage moved backward | The buyer has reopened a question the rep thought was settled |
| Close date pushed | The best single sign of a slip. A deal pushed from one quarter to the next is less likely to close, even in commit |
| Amount went down | Scope or price is under pressure, often late in the cycle |
| Amount went up | Scope grew, or the number was sandbagged early |
| Nothing changed | Often the earliest warning of all: no activity, no data moving and no notes |
Why does ORM snapshot pipeline every day?
Because that is the only way to see exactly what changed and when. With a daily snapshot, every opportunity has a history: each stage it passed through, each close date it carried, each amount it was worth. That gives a machine learning model a much richer set of signals than a static CRM record.
It also makes it possible to measure things a CRM cannot answer on its own:
- how long deals of each type really spend in each stage - how often close dates move, and how far - how much a deal's value typically changes before it closes - when in the quarter the forecast became right
For forecasting, the state of the pipeline matters. The movement in the pipeline often matters more.
Why does the first snapshot of the quarter matter most?
It fixes the starting line. On the first day of a quarter, a snapshot records every deal with a close date inside that quarter. Across ORM customers, only about 20% of that pipeline usually closes in the quarter. So roughly 80% of the value sitting in the quarter on day one does not land there.
A team that never snapshots its pipeline cannot see its own version of that number. It treats every quarter's starting position as equally sound. Keep the day-one snapshot and you can take the quarter apart once it ends. You can see which starting deals closed and which slipped, and which new deals arrived and closed inside the quarter.
What should a pipeline snapshot capture?
At minimum, for every open opportunity, every time:
| Field | Why you need its history |
|---|---|
| Stage | Progression and regression over time |
| Amount | Changes in scope or price, and when they happened |
| Close date | Every push, and how far it moved |
| Owner | Rep changes that often precede a slip |
| Forecast category | How commit and best case moved through the quarter |
| Last activity date | Silence, which is the earliest signal |
How does this change how you treat stale deals?
It gives you a clear rule. ORM counts meaningful activity as a change in stage, close date or amount. For most customers, a deal with none of those in 12 months is stale and comes out of the forecast. Across ORM customers, it varies, but typically more than 10% of pipeline fails that test. Without snapshots, "stale" is a matter of opinion. With them, it is a query.
How do you start if you have no history?
Start today. Schedule a daily export of open opportunities with the fields above, or at least a weekly one, and keep every copy. A few months of snapshots will already show you how often close dates move and how long deals really sit in each stage. The data does not need to be perfect first. Consistent data carries signal even when it is messy, and data improves once people know it is being measured.
For how snapshot signals feed a forecast, see sales forecasting techniques, pipeline hygiene and forecast snapshots, which apply the same idea to the forecast number itself.
---
Pete Furseth is COO of ORM Technologies, which builds custom revenue forecast models on a company's own CRM data.Frequently Asked Questions
What is a pipeline snapshot?
A saved copy of every open opportunity at a point in time, with its stage, amount, close date and owner. Comparing snapshots shows exactly what changed and when, which a CRM normally overwrites.
Why does the change in CRM data matter for forecasting?
Because changes capture behavior. A deal in proposal is useful to know. A deal that moved into proposal yesterday is more useful. Stage moves, amount changes and close-date pushes show how a deal is really progressing.
Which CRM change is the strongest warning sign?
A rep changing the close date. A deal that slips from one quarter to the next is less likely to close, even in commit. Earlier still is the absence of change: no activity, no data moving and no notes.
How often should you snapshot the pipeline?
ORM snapshots pipeline every day, which is what makes it possible to see exactly what changed and when. Weekly snapshots are a practical minimum if you are starting from nothing.
What counts as meaningful activity on a deal?
ORM counts a change in stage, close date or amount. A deal with none of those for 12 months is treated as stale for most customers.
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