Deal pipelines
HubSpot deal pipeline setup: built around the buyer journey.
TL;DR: A HubSpot deal pipeline needs three to seven stages, and every stage should name something the buyer did, not something your team did. Each transition is triggered by a property, so the board tells the truth without anyone remembering to drag a card.
Scattered deal cards jitter, then line up left to right into a clean labelled pipeline of stages, and a single live deal slides through it, relabelling to the lifecycle stage it has reached.
Why your pipeline stopped matching how you sell.
Most pipelines start honest. Someone sets HubSpot up in week one, picks stages that sound about right, and the team gets on with selling.
Twelve months later the stages describe your admin, not your buyer. There is a Proposal Sent stage holding deals that went quiet in March. There is a Negotiation stage that means four different things depending on which rep you ask. A third of the board sits in one stage, because that stage became the place deals go when nobody is sure what to do with them.
You notice it at forecast time. The weighted number says one thing. The rep who owns the deals says it feels like half that. Both are reading the same board.
The problem is not discipline. The problem is that the stages do not describe anything you can observe, so there is nothing to be disciplined about.
The rule: a stage is a buyer commitment.
One rule fixes most pipelines.
A stage names something the buyer did. Not something you did.
Proposal Sent is something you did. You can send a proposal into a void. The deal advances on your board while nothing has changed in the buyer’s world.
Proposal Reviewed is something the buyer did. They opened it, they came back, they asked about the scope. That is a commitment, and it is observable, which means you can trigger on it.
Apply the rule across the board and the fake progress goes away. Deals stop advancing because a rep felt optimistic on a Friday. They advance because something happened.
The test for each stage is short. Can you name the buyer action that puts a deal here? Can you see that action in HubSpot without asking anyone? If either answer is no, the stage is describing your internal process and it will drift.
- Proposal Sent: you can send into a void
- Follow-up Made: measures activity, not progress
- Negotiation: means whatever the rep decides
- Advances on effort, so the forecast inflates
- Proposal Reviewed: they opened it and came back
- Discovery Agreed: they signed off on the requirement
- Verbal Agreement: they said yes and named the signer
- Advances on an observable event, so it can be triggered
How many stages you actually need.
Three to seven. Below three you cannot forecast, because everything is either open or closed. Above seven the reps stop maintaining it, and a board nobody maintains is worse than no board at all.
Most teams land on five. The number matters less than the spacing. Each stage should hold deals for a roughly similar length of time. If deals sit in one stage for six weeks and pass through the next in two days, those are not two stages. That is one stage with a status field inside it.
Count your stages against your sales cycle. A four week cycle does not support seven stages. A nine month enterprise cycle probably needs more than four.
The five stages most B2B pipelines land on.
These are the five we ship most often. Take them as a starting shape, not a template to copy.
- Qualified. The buyer has confirmed a problem worth solving and a rough budget.
- Discovery complete. The buyer has agreed on what success looks like.
- Proposal reviewed. The buyer has read it and come back with a question or an objection.
- Verbal agreement. The buyer has said yes and named who signs.
- Closed won. Signed, with Closed lost alongside as the exit.
Each one in full below.
Qualified
The buyer has confirmed a problem worth solving and a rough budget. Before this point it is a lead, not a deal. Deals created earlier than this are the single biggest cause of a bloated, meaningless pipeline.
Discovery complete
You have run the call, you understand the requirement, and the buyer has agreed on what success looks like. The buyer action is the agreement, not the call.
Proposal reviewed
The buyer has been through the proposal and come back with a question, an objection or a request to change something. Silence is not this stage.
Verbal agreement
The buyer has said yes and named who signs. This stage exists to separate the deals waiting on paperwork from the deals still being decided. Skip it and your forecast blurs the two.
Closed won
Signed. Not verbally agreed, not about to be signed. Signed.
Closed lost sits alongside as the exit, with a required lost reason so the board teaches you something after the fact.
- 1QualifiedThe buyer has confirmed a problem worth solving and a rough budget. Before this point it is a lead, not a deal.
- 2Discovery completeYou understand the requirement and the buyer has agreed on what success looks like. The commitment is the agreement, not the call.
- 3Proposal reviewedThe buyer has been through it and come back with a question, an objection or a change request. Silence is not this stage.
- 4Verbal agreementThey said yes and named who signs. Separates deals waiting on paperwork from deals still being decided.
- 5Closed wonSigned. Not verbally agreed, not about to be signed. Closed lost sits alongside as the exit, with a required lost reason.
Setting the pipeline up in HubSpot.
The build itself takes under an hour. Getting the stages right is the work; this part is mechanics.
Go to Settings, then Objects, then Deals, then the Pipelines tab. Create the pipeline before you touch stages, and name it after the motion it describes, not the team that runs it. New Business and Renewals are motions. Dave’s Pipeline is not.
Add your stages in order. Each one gets a name taken from the buyer action, a deal probability, and a set of required properties. Drag them into sequence once, then leave the order alone. Reordering stages later rewrites history on every open deal.
Set the two closed stages last. HubSpot needs one stage marked closed won and one marked closed lost, and it treats them differently from the rest in every report you will build.
Then set stage automation. This is the part most portals skip, and skipping it is why the board drifts. For each transition, decide the property that moves the deal, and build one workflow that writes it. One workflow per transition. Two workflows writing the same stage is how deals start bouncing between stages overnight.
- 1Create the pipelineSettings, then Objects, then Deals, then the Pipelines tab. Name it after the sales motion, not the team. New Business and Renewals are motions.
- 2Add stages in orderEach gets a buyer-action name, a probability, and its required properties. Set the sequence once; reordering later rewrites history on every open deal.
- 3Mark the closed stagesOne closed won, one closed lost. HubSpot treats both differently from the rest in every report you will build on this pipeline.
- 4Build one workflow per transitionDecide the property that moves each deal and build a single workflow that writes it. Two workflows on one stage makes deals bounce overnight.
Deal probability and what it does to your forecast.
Every stage carries a probability. HubSpot multiplies it by the deal amount to produce the weighted forecast leadership reads.
The default probabilities are evenly spaced guesses. Yours should come from your own history: of the deals that reached this stage last year, what share closed? That is your probability. If you do not have enough history yet, start conservative and revisit each quarter.
Two things go wrong here. The first is optimistic probabilities, which inflate the forecast until leadership stops believing any of it. The second is leaving every stage at the default, which produces a forecast that is arithmetically fine and tells you nothing about your business.
Set them from data, review them quarterly, and the weighted number becomes a number people act on.
Stage probability is not forecast category
These two get used as if they were the same control. They are not, and the difference matters when the forecast is wrong.
Stage probability belongs to the stage. Every deal sitting in that stage inherits the same percentage, and that percentage is what produces the weighted amount.
Forecast category belongs to the deal. A rep sets it by hand, choosing between Pipeline, Best case, Commit, Closed won and Omitted, to say what they believe about that specific deal regardless of the stage it is in.
The practical consequence: moving a deal to Commit does not change its weighted amount, and dragging a deal to a later stage does change it. If leadership is reading a weighted forecast while reps are managing forecast categories, the two are describing different things and the meeting goes in circles.
- One percentage per stage, inherited by every deal in it
- Multiplied by the amount to produce the weighted forecast
- Set from your own close-rate history, reviewed quarterly
- Changes when the deal moves stage
- Pipeline, Best case, Commit, Closed won, Omitted
- Set by hand by the rep, as a judgment on that one deal
- Does not change the weighted amount
- Changes when the rep's confidence changes
Required properties: gating each stage.
A required property is a field the rep must fill before the deal can advance. It is the strongest tool in the pipeline and the most commonly ignored.
Gate the fields you will need later, at the moment the rep actually knows them. Close date and amount belong at Qualified. The decision maker and the competitor belong at Discovery complete. The lost reason belongs at Closed lost, always.
Keep it to two or three fields per stage. Gate more than that and reps route around the board, logging deals late or parking them in whatever stage has the fewest required fields. A gate people avoid is worse than no gate.
The payoff arrives about two quarters later, when someone asks why win rates dropped in one segment and the answer is already sitting in your data.
If your property layer is already showing this kind of decay, a systems audit is the right first step.
When you need a second pipeline.
Less often than people think. Most requests for a second pipeline are really requests for a filter.
You need a separate pipeline when the stages genuinely differ. New business and renewals are different motions with different buyer commitments, so they earn separate pipelines. Two regions selling the same thing through the same steps do not; that is one pipeline with a region property and two dashboard filters.
The cost of a second pipeline is real. Reports that span pipelines get harder, and every stage automation has to be built and maintained twice. Split when the stages differ, filter when only the audience differs.
- New business and renewals: different buyer commitments
- Different stage counts or different required properties
- A transition in one motion has no equivalent in the other
- Accept the cost: cross-pipeline reports and duplicated automation
- Two regions running the same steps
- Deal size bands inside one motion
- Per-rep or per-team views of the same board
- One pipeline, one property, and dashboard filters
Moving deals between pipelines.
You can move a deal by opening the record, clicking the deal stage field, and choosing a stage in the other pipeline. For several deals at once, select them from the deals index and use the bulk edit action.
Three things to know before you do it in bulk.
The deal keeps its history, but stage timestamps from the old pipeline do not map onto the new stages. Time in stage reporting for those deals will be wrong for a while.
Required properties from the destination pipeline are not enforced on a move. Deals can arrive missing fields that every other deal in that pipeline has, which quietly breaks reports that filter on them.
Stage automation fires on arrival. If the destination stage triggers a notification or a task, moving fifty deals sends fifty notifications. Pause the workflows first.
What changes after the rebuild.
Three things, usually within a quarter.
The forecast starts matching the reps. Not because anyone got better at estimating, but because stages now describe observable buyer commitments, so the weighted number is built from facts rather than feel.
Stalled deals become visible. When each stage means one thing, a deal sitting in it for triple the normal time stands out. Before the rebuild, those deals hid inside stages that meant everything and nothing.
Pipeline reviews get shorter. The meeting stops being a debate about what each stage means and becomes a conversation about the deals themselves.
Internal links.
We ship a stage definition sheet and a workflow runbook with every pipeline build.
Rebuild your deal pipeline
If your forecast and your reps disagree, the stages are usually the reason. Book a discovery call. We run a free thirty-minute review of your current board and tell you which stages describe your admin rather than your buyer, and what the rebuilt sequence would look like.
Book a free consultCommon questions.
What are the 5 stages of a sales pipeline?
The five stages most B2B pipelines land on are Qualified, Discovery complete, Proposal reviewed, Verbal agreement, and Closed won, with Closed lost sitting alongside as the exit. Each one is named after a buyer commitment rather than an internal task, so a deal only advances when something observable has happened on the buyer's side.
How many deal stages should a HubSpot pipeline have?
Between three and seven. Below three you cannot forecast, because every deal is either open or closed. Above seven the team stops maintaining the board. Most teams land on five. Spacing matters more than the count: each stage should hold deals for a roughly similar length of time, and the number should match your sales cycle.
How do I build a deal pipeline in HubSpot?
Go to Settings, then Objects, then Deals, then the Pipelines tab. Create the pipeline and name it after the sales motion it describes. Add stages in order, giving each one a name taken from a buyer action, a deal probability, and two or three required properties. Mark one closed won stage and one closed lost stage. Then build one workflow per stage transition so the property, not a person, moves the deal.
How do I move deals between pipelines in HubSpot?
Open the deal record, click the deal stage field, and choose a stage in the other pipeline. To move several at once, select them from the deals index and use bulk edit. Before moving in bulk, pause the destination pipeline's stage automation, because arriving deals fire it. Note that stage timestamps from the old pipeline do not map onto the new stages, and destination required properties are not enforced on a move.
What is the difference between deal stage probability and forecast category in HubSpot?
Stage probability belongs to the stage: one percentage per stage, inherited by every deal in it, multiplied by the deal amount to produce the weighted forecast. Forecast category belongs to the deal: a rep sets it by hand, choosing between Pipeline, Best case, Commit, Closed won and Omitted, to express a judgment about that specific deal. Moving a deal to Commit does not change its weighted amount; moving it to a later stage does.