HubSpot Deal Stages

Deal stages in HubSpot define the steps a deal moves through from first sales conversation to closed. Each stage has a name, a probability, and exit criteria that determine when a deal should advance. Well-designed stages give you accurate forecasting, clear pipeline visibility, and consistent sales process across reps.

The principle: deal stages should reflect your actual sales process, not a textbook funnel. If your reps go from discovery to proposal in one step, don't force a "demo" stage between them. Stages that don't match reality get ignored or gamed.

Default vs Custom Stages

HubSpot defaults (don't use these as-is)

Why to customize

The defaults are designed for a generic sales process that doesn't exist. Your sales process has specific steps, specific qualification criteria, and specific milestones that the defaults don't capture. Always customize.

Designing Custom Stages

Stage design framework

Stage design rules

• 5-7 active stages maximum. (Plus Closed Won and Closed Lost.) Fewer than 5 and you can't see where deals stall. More than 7 and reps can't remember the differences. 5-7 is the sweet spot
• Every stage must have clear entry criteria. "What must be true for a deal to be in this stage?" If you can't define it in one sentence, the stage is too vague
• Stages should be verifiable, not assumed. "Decision Maker Bought-In" is assumed. "Verbal Yes Received" is verifiable. Design stages around observable milestones, not hoped-for mental states
• Probability should reflect your actual win rate from that stage. Don't guess. Pull 6 months of data. What % of deals that entered Stage 3 eventually closed? That's the probability. Update annually
• Include a "Closed Lost" reason. HubSpot supports close reasons as a property. Require a reason for every Closed Lost: competitor, no budget, timing, no decision, churned champion. This data drives improvement

Stage naming rules

• Use outcome-based names. "Qualification Complete" not "Qualifying." The stage name should indicate what's been achieved, not what's in progress
• Avoid jargon. "Proposal Sent" is clear to everyone. "SAL" or "Stage 3" is only clear to people who memorized the stages. Use plain language
• Don't use numbers alone. "Stage 1, Stage 2, Stage 3" tells reps nothing about what the stage means. Name them descriptively

Probabilities

Setting probabilities correctly

Pull data for the last 6-12 months:

For each stage:
  Count deals that ENTERED this stage
  Count deals that eventually CLOSED WON

Stage probability = Closed Won from this stage / Entered this stage

Example:
  Discovery: 200 deals entered, 30 closed won → 15%
  Qualification: 120 entered, 28 closed won → 23%
  Evaluation: 80 entered, 24 closed won → 30%
  Business Case: 50 entered, 22 closed won → 44%
  Proposal: 35 entered, 20 closed won → 57%
  Negotiation: 25 entered, 18 closed won → 72%

Probability rules

• Calculate from your data, not from industry benchmarks. Your sales process, your ACV, and your ICP produce unique win rates. Benchmarks are starting points for new companies only
• Update annually. Win rates shift as your product, market, and team evolve. Stale probabilities produce inaccurate forecasts. Review and recalculate every 12 months
• Probabilities should increase monotonically. Each stage should have a higher probability than the previous one. If Stage 3 has a lower probability than Stage 2, the stage definitions are wrong
• Don't use round numbers. 10%, 20%, 40%, 60% look made up. 12%, 23%, 37%, 52% look calculated. Even if the difference is small, data-backed numbers signal rigor

Required Properties per Stage

What to require at each stage

Property rules

• Require deal amount at Discovery. Even a rough estimate. You can't forecast without it. Update as the deal develops
• Require close date at Discovery. Reps will estimate. That's fine. The estimate gets refined as the deal progresses. Without it, pipeline reports are useless
• Require close reason at Closed Lost. Every lost deal should have a reason. This data feeds competitive intelligence, product development, and sales enablement. Don't make it optional
• Don't over-require. If you require 15 fields to move from Discovery to Qualification, reps will leave deals in Discovery forever to avoid the data entry. 2-4 required fields per stage transition is the maximum

Pipeline Configuration in HubSpot

Setup steps

1. Settings → Objects → Deals → Pipelines
2. Edit the default pipeline or create a new one
3. Add/rename stages with your custom names
4. Set probability for each stage
5. Set "Deal properties required" for each stage transition
6. Configure automation:
   - Auto-set close date when deal enters Proposal
   - Auto-task when deal enters Negotiation
   - Auto-notify manager when deal enters Closed Lost

Multiple pipelines

Multiple pipeline rules

• Don't create a new pipeline for minor differences. If the stages are 80% the same, use one pipeline with a property to distinguish. Multiple pipelines multiply reporting complexity
• Maximum 3 pipelines. More than 3 and pipeline reporting becomes a full-time job. Simplify before adding
• Reports must work across pipelines. Ensure your forecasting reports can aggregate across pipelines when needed

Measurement

Pre-Setup Checklist

• [ ] 5-7 active stages defined (plus Closed Won and Closed Lost)
• [ ] Each stage has clear entry criteria (verifiable, not assumed)
• [ ] Probabilities calculated from historical data (not guessed)
• [ ] Required properties set per stage (2-4 per transition, not more)
• [ ] Deal amount required at Discovery (even an estimate)
• [ ] Close date required at Discovery (even an estimate)
• [ ] Close reason required at Closed Lost (dropdown with 5-7 options)
• [ ] Stage names are descriptive and outcome-based
• [ ] Automation configured for key transitions (notifications, tasks)
• [ ] Team trained on stage definitions and when to advance deals
• [ ] Pipeline report and forecast built using the new stages

Anti-Pattern Check

• Using HubSpot's default stages. "Appointment Scheduled" and "Qualified to Buy" don't match any real sales process. Customize stages to reflect your actual steps. The defaults are a starting point, not a solution
• 10 deal stages. Reps can't remember the difference between Stage 6 and Stage 7. Deals sit in wrong stages. Data is unreliable. Cut to 5-7 stages with clear differences between each
• Probabilities are guessed at 10%, 20%, 40%, 60%, 80%. These look like defaults because they are. Calculate from your own data. Your Evaluation stage might close at 33%, not 40%. Accuracy matters for forecasting
• No required properties. Deals move through stages with no amount, no close date, no contact. Your pipeline is numbers without data behind them. Require the minimum viable fields at each stage
• No Closed Lost reason. 40% of pipeline closes as lost. Nobody knows why. You can't improve what you don't measure. Require a close reason dropdown for every lost deal
• Reps leave deals in early stages to avoid data entry. You required 8 fields to move from Discovery to Qualification. Reps leave deals in Discovery and work them through to Closed Won in one jump. Pipeline visibility is zero. Reduce required fields to 2-4 per transition
• Stage names are internal codes. "SAL," "SQL-A," "Stage 3B." Nobody outside sales ops knows what these mean. Use descriptive names that any stakeholder can understand