Why pipeline design matters
Everything else in Pipedrive sits on top of the pipeline. Stage probabilities feed the weighted forecast. Conversion reports in Insights measure movement between stages. Rotting settings count days per stage. If the stages do not describe reality, none of those numbers mean anything, and the team learns to ignore them within a month.
The default pipeline that ships with a new account is a placeholder, not a recommendation. Teams that keep it end up with deals parked in a stage called Negotiations Started for half a year. The fix costs a day of thinking. The damage from skipping it costs quarters.
There is a second, quieter reason. A pipeline that mirrors how your buyers actually decide gives reps a script for the next step. A pipeline that mirrors your org chart gives them nothing.
The design decisions that matter
Decision one: what counts as a stage. A good stage is a verifiable buyer event, something that either happened or did not. Discovery call held. Proposal sent. Verbal yes received. A bad stage is a seller mood, like Interested or Hot. If two reps can disagree about whether a deal belongs in a stage, the stage is broken.
Decision two: where the pipeline starts. The first stage should require qualification. Everything before that belongs in the Leads inbox, which Pipedrive built precisely so your pipeline is not a dumping ground.
Decision three: one pipeline or several. Split only when the process differs. New business, renewals and partner deals follow different steps, so they earn separate pipelines. Regions and products usually do not. A field filters better than a pipeline, and every extra pipeline splits your reporting.
Decision four: the exit criterion per stage. Write one sentence per stage that says what must be true before a deal moves on. This becomes the backbone for probabilities and rotting settings later.
A worked example
Here is a pipeline we deploy for B2B service companies with a thirty to ninety day cycle. Five stages, each with a verifiable exit.
| Stage | Exit criterion |
|---|---|
| Qualified | Budget, need and timeline confirmed in a real conversation. |
| Discovery held | Discovery meeting done, pain and decision process documented on the deal. |
| Proposal sent | Written proposal delivered and a review meeting booked. |
| In negotiation | Buyer has responded to the proposal with changes or conditions. |
| Verbal agreement | Buyer said yes. Only paperwork remains. |
Note what is missing. There is no Contacted stage, because contact is a lead activity. There is no On Hold stage, because a hold is a close date change, not a process step. And there is no Closing stage, because Verbal agreement already says it with a testable claim.
Common mistakes
The most common mistake is stages that describe seller activity. Calling, Following up and Chasing are things a rep does, not places a deal sits. They produce pipelines where deals bounce backwards, which wrecks conversion reporting.
The second is pipeline sprawl. We audit accounts with nine pipelines for a twelve-person team. Each one felt reasonable at creation. Together they make the Insights reports nearly unreadable, because every report must be filtered and summed across pipelines by hand.
The third is the parking stage. Nurture, Long term and On ice collect deals that should be lost. Mark them lost with a reason, and let a won-back automation or a quarterly review revive them. A lost deal with a clear reason is data. A parked deal is noise.
Maintenance
Review the pipeline twice a year, not more. Pull the deal conversion report in Insights and look for two smells. A stage where almost every deal converts is a stage nobody uses as a checkpoint, so consider merging it. A stage where deals sit three times longer than anywhere else usually hides two real steps, so consider splitting it.
Resist stage changes outside those reviews. Every change resets the team's muscle memory and muddies historical comparisons. This is also why pipeline design is step two of our six step method, right after process mapping and before anything technical. Get it right once, then leave it alone.
Questions
How many stages should a Pipedrive pipeline have?
Five to seven for most B2B teams. Fewer than four hides where deals stall. More than eight means some stages describe activities instead of buyer commitment, and reps will skip them.
When do we need a second pipeline?
When the selling process itself differs, not the product or the team. New business and renewals deserve separate pipelines. Two products sold through the same steps belong in one pipeline with a product field.
Can we change stages later without losing history?
You can rename and reorder stages freely. Deleting a stage forces you to move its deals first, and historical conversion reports get harder to read. Design carefully now, then change rarely.
Should won and lost be stages?
No. Won and lost are deal statuses in Pipedrive, not stages. Every open stage should describe a step on the way to a decision. The decision itself closes the deal.
Where do unqualified prospects go if not in the pipeline?
Into the Leads inbox. Pipedrive keeps leads separate from deals for exactly this reason. A pipeline full of maybes destroys your conversion numbers and buries the deals that deserve attention.