Two teams open the same CRM, pull a figure for pipeline generation, and arrive at different numbers. Both are reading the data correctly. The report is assembled from fields that arrive with vendor defaults, and it counts a unit created by an event that nobody has defined in writing. Before pipeline generation can be improved, both teams have to agree on what the number in front of them is measuring, and most of that work sits in field definitions rather than in campaigns.
Demand, a lead and an opportunity are three separate objects
Demand is interest that exists whether or not you captured it: someone has the problem, has budget pressure, and is already reading about it. That interest is what demand generation work acts on, and it is real before any of it reaches a form.
A lead has a name, an address and permission of some kind attached: a person or an account you are able to contact.
An opportunity is a record with a value, an expected close date and a stage. Pipeline is the set of those records. Each word names a different object, so a number about one of them cannot answer a question about another. This piece is about the third.
The number on the dashboard is a balance, not a flow
“Pipeline generated last quarter” asks about a flow: what entered during a window, and what it was worth on the way in. “Pipeline” on most dashboards is a balance: the open pipeline, meaning every opportunity that is neither won nor lost as of this morning.
The two move for different reasons. Open pipeline rises when deals sit in place and nothing closes, and it drops the week a large opportunity is won, which is the opposite of bad news. Created pipeline is untouched by both events. Each figure was set the day its opportunity was created and does not move again unless somebody edits the record.
The figure that reaches us as pipeline generated is usually the open pipeline.
The substitution is not a rounding problem. A team can report a rising number in a quarter where it created almost nothing, and a falling one in its strongest quarter of created pipeline. To see whether generation is working, filter the report on the created date rather than on what happens to be open today.
Where the probability in a weighted pipeline figure comes from
A pipeline figure is not a raw sum. It is the output of settings that shipped with the software, and both major systems arrive with a considered view of what a deal in progress is worth.
Data table
| Default deal stage | Probability |
|---|---|
| Appointment scheduled | 20% |
| Qualified to buy | 40% |
| Presentation scheduled | 60% |
| Decision maker bought-in | 80% |
| Contract sent | 90% |
| Closed won | 100% |
| Closed lost | 0% |
HubSpot’s default deal pipeline ships with seven stages, each carrying a probability: Appointment scheduled 20%, Qualified to buy 40%, Presentation scheduled 60%, Decision maker bought-in 80%, Contract sent 90%, Closed won 100% and Closed lost 0%. The same HubSpot documentation records that a custom probability can be entered instead.
Those probabilities then do arithmetic. HubSpot documents its weighted amount as “the Amount multiplied by the Deal probability,” and its forecast category as “the likelihood of the deal closing, categorized as Not forecasted, Pipeline, Best case, Commit, or Closed won.” A weighted pipeline figure is therefore the product of two inputs: an amount somebody entered on the record, and a percentage that came with the product.
Salesforce runs the same logic from the other end, starting from the stage. Its documentation states that “If the stage of an opportunity changes, the probability and forecast category also change,” and, separately, that “Users can edit the Forecast Category field on opportunities that they own, overriding the default stage-to-category mapping.” The override is there on purpose: the design leaves room for knowledge the stage does not encode.
Neither vendor is doing anything questionable here. A default has to ship or nobody can use the product on day one. The consequence is what matters: a pipeline number presented to a board rests on a probability chosen before anyone had looked at your business, and on a field a person is entitled to change by hand. That makes it a claim about the future, not a reading taken from an instrument.
A stage entry criterion has to be an event both teams can check
All of that assumes the unit exists. Created pipeline counts opportunities, so somebody has to decide when a record becomes one. That decision is the stage entry criterion, and it only functions if both sides of the handoff can verify it.
An observable criterion describes an event that either happened or did not: a meeting with a named attendee, a budget owner identified by name, a written requirement received. Either team can open the record and see it. Either team can refuse the entry and say exactly why.
Judgment words cannot carry that weight. “Qualified,” “engaged” and “sales-ready” describe a state of mind belonging to whoever wrote them down. Two people can read the same account and disagree without either being careless, because there is nothing to check. That is how one account gets counted twice, once by each team, or dropped by both.
We ask to see that criterion in writing. It is rarely anywhere both teams can read it.
The test is short: can somebody who was not in the meeting decide, from the record alone, whether the criterion was met? If nobody outside the room can answer that, the criterion has not been written yet.
The sourcing field, and the person who fills it in
Once you know when an opportunity is created, the second argument starts: who created it. Sourcing is rarely derived. It is recorded in a sourcing field on the record, set once by a person or by an automation somebody configured, at or near creation.
Marketing-sourced, sales-sourced and partner-sourced read like statements of history. What they hold is what whoever touched the record believed at that moment. With no written definition behind the field, it records the belief instead of the history, and the end of the quarter becomes an argument about intent.
Sourcing fields get filled in under one of two rules, and the same events produce different numbers under each:
- First touch. The sourcing field records the channel that first identified the account, whatever happened afterwards.
- Creating event. The field records what caused the opportunity to exist, most often the meeting that produced it.
Choose one and write it down. The choice bites hardest where the work is shared, for instance when a sales development team, or a B2B lead generation agency, books meetings into accounts marketing has been working. Under first touch, that meeting is marketing-sourced. Under creating event, it is not. The number breaks only when both rules run at once, one team on each.
What becomes possible once both rules are in writing
Settle it with two sentences: what event puts an opportunity into the pipeline, and what the sourcing field records. Write those and three things open up that were closed before.
Created pipeline becomes recomputable. Both rules are properties of the record, so past quarters can be counted again under today’s definitions, and you get a trend instead of a run of differently defined numbers.
Disagreement now has somewhere to go. “That should not have counted” stops being a comment on a colleague’s judgment and turns into a question about a field on a record, which anyone can open.
Coverage stops being a borrowed multiple. Once created pipeline is defined and stable, a team can work out, from its own win rate and under its own definitions, how much open pipeline it needs. What that coverage ratio implies for spending is a separate question, and it belongs with the rest of your B2B marketing budget planning.
Nobody needs a reporting project for this. The fields already exist in the CRM; the two sentences do not, and they have to come from the teams who will live with them.
FAQ: pipeline generation
What is pipeline generation?
Pipeline generation is the work of creating new opportunities: records that carry a value, an expected close date and a stage. Measured properly, it is a flow, counted by created date across a window. That is a different figure from the open pipeline on most dashboards, which is a balance as of today.
Is pipeline generation the same as lead generation?
No: lead generation produces people and accounts you are able to contact, while pipeline generation counts the opportunities created from them, which is why a strong month for one can sit next to a flat month for the other.
What counts as an opportunity?
Whatever your stage entry criterion says. Unwritten, it changes with whoever is asked. A workable criterion names an observable event, such as a meeting with a named attendee or a documented requirement received, so either team can open the record and see whether it was met.
How do you measure pipeline generation?
Count created pipeline: the number and value of opportunities whose created date falls inside the window, judged against one stage entry criterion. Keep it separate from open pipeline, which is a balance and shifts when deals age or close. If you also report a weighted figure, state which probabilities produced it.
Who owns pipeline generation, marketing or sales?
Both teams touch it, so it needs a written rule more than an owner. The two definitions that remove most of the argument are the stage entry criterion, which decides when an opportunity exists, and the sourcing field, which decides who is credited. Agree on both, in writing.
Why do marketing and sales report different pipeline numbers?
Often because they are answering two different questions. One number is created pipeline across a window, a flow. The other is open pipeline as of today, a balance. The rest of the gap comes from fields carrying defaults or edits: HubSpot documents its weighted amount as the Amount multiplied by the Deal probability, and Salesforce documents that owners can, by design, edit the forecast category on opportunities they own.
COSEOM Team
COSEOM is a multilingual B2B growth agency. Since 2008 the COSEOM team has run international SEO, generative engine optimization, paid media and demand generation programs for B2B software and technology companies, in English, German, Spanish and Portuguese. This account publishes our research, playbooks and product guides. More about COSEOM