A HubSpot deal pipeline is only as useful as the discipline behind it. The stages can be named correctly, the probabilities can be set thoughtfully, and the required properties can be configured carefully - and the pipeline still produces an unreliable forecast if sales reps are not updating deal stages to reflect real buyer commitment, close dates are not moving when deals slip, and stagnant opportunities sit at 60% probability for six weeks without anyone challenging them.
This guide covers both sides of the problem: the configuration (how to build a pipeline that reflects the actual buying process rather than the internal selling process) and the governance (how to keep the pipeline clean, the forecast credible, and the sales team working from accurate data). It is written for sales ops, RevOps, and marketing leaders responsible for HubSpot pipeline management in B2B organisations.
For the HubSpot implementation context that precedes pipeline design, see our HubSpot onboarding guide 2026. For how lead scoring connects to deal creation and MQL routing, see our HubSpot lead scoring guide 2026.
What Deals and Pipelines Are in HubSpot
In HubSpot CRM, a deal is a record that represents a specific sales opportunity - a potential transaction between your company and a prospect or customer. Deals sit alongside contacts and companies in the CRM data model: a deal is associated with one or more contacts (the people involved in the purchase decision) and usually with a company (the account).
A pipeline is the sequence of stages a deal moves through from initial qualification to close. HubSpot creates a single default pipeline - called 'Sales Pipeline' - when a new portal is set up. This default pipeline serves most simple use cases but should be reviewed and adapted to reflect the actual buying process before the sales team begins entering live deals.
The key architectural point: HubSpot's deal object stores the deal's current stage, the deal value, the expected close date, and any custom properties defined by the organisation. The combination of stage probability and deal value produces the weighted pipeline figure that drives forecast accuracy. Every configuration decision in this guide connects to that number. (Source: HubSpot Knowledge Base - Deal pipelines and deal stages: knowledge.hubspot.com/deals/create-deal-pipelines-and-deal-stages)
The Core Principle: Buyer Commitment, Not Seller Activity
The most common pipeline design mistake in B2B CRM: stages are named after seller activities ('Demo Sent', 'Proposal Created', 'Follow-Up Scheduled') rather than buyer commitments ('Discovery Complete', 'Solution Evaluated', 'Verbal Commitment Received'). A pipeline built around seller activities tracks what the rep did, not whether the buyer is progressing toward a decision.
The difference matters for forecasting. A deal at 'Proposal Sent' in a seller-activity pipeline tells you that a proposal document was emailed. It tells you nothing about whether the prospect read it, whether a decision-maker is engaged, or whether the deal is real. A deal at 'Verbal Commitment' in a buyer-commitment pipeline tells you the prospect has explicitly said yes - which is a fundamentally different probability signal.
The design principle for each pipeline stage: it should only be possible to advance a deal to a new stage after the buyer has done something, said something, or agreed to something. If the sales rep can advance a deal without any buyer action - by sending an email, completing an internal task, or simply deciding the deal is further along - the pipeline has no information value.
Creating and Customising Pipelines in HubSpot
Accessing Pipeline Settings
Pipeline configuration in HubSpot is found in Settings > Objects > Deals > Pipelines. From this screen, you can create new pipelines, edit existing stages, set stage probabilities, configure required properties per stage, and manage pipeline access by user or team.
Before creating stages, confirm:
- Which teams or reps will use this pipeline - pipeline access can be restricted by team
- Whether this is a new-business, renewal, or expansion pipeline - each has different stage logic (covered in the Multi-Pipeline section below)
- What the average sales cycle length is - this will inform stage sequencing and stagnation thresholds
Creating a Stage
To add a stage in HubSpot's pipeline settings: click Add Stage, enter the stage name, set the probability, and configure required properties. Stages can be reordered via drag and drop. The first stage and the two closed stages (Won and Lost) cannot be deleted, but their names and properties can be edited.
Stage naming conventions that improve usability: use consistent capitalization (title case throughout), use noun phrases that describe the buyer's state rather than verb phrases that describe seller actions (Verbal Commitment rather than Verbal Committed), and keep names under 30 characters so they display cleanly in the pipeline view and reports.
Pipeline Stage Template: A Buyer-Commitment Framework
The following stage template reflects a standard B2B SaaS or professional services deal - 8 active stages plus 2 closed stages. Adapt stage names, probabilities, and required properties to match the actual buying process at your organisation. The logic of each stage matters more than the specific names.
Two important design notes on this template:
Discovery Scheduled vs Discovery Complete are separate stages. Many pipelines combine these into a single 'Meeting Booked' stage. Keeping them separate is worth the additional stage count: it reveals how many meetings are booked but not held, which is a leading indicator of pipeline health. A high ratio of 'Scheduled' to 'Complete' suggests reps are booking meetings that prospects are not attending - a qualification problem.
Verbal Commitment is a distinct stage before Contract Out. In B2B sales with a procurement process, there is often a gap between 'the prospect said yes' and 'the contract is out for signature' - legal review, budget approval, vendor management forms. Capturing the verbal commitment as a distinct stage gives the revenue team confidence that the deal is real while the contracting mechanics are in progress.
Setting and Calibrating Stage Probabilities
Every deal stage in HubSpot has a probability - the percentage likelihood the deal closes, used to calculate weighted pipeline (deal value × probability = weighted value). HubSpot provides default probability values for its default stages; custom stages start at the probability you set at creation.
The critical distinction: probabilities should reflect historical win rates from that stage, not aspirational conversion rates. A deal at 'Proposal Sent' should carry a probability equal to the percentage of past deals that reached 'Proposal Sent' and eventually closed won. If that figure is 35%, the stage probability should be 35% - not 60% because the rep feels good about the deal.
To calibrate stage probabilities from historical data:
- Pull a deals report showing all closed deals (won and lost) in the last 12 months.
- For each stage, calculate: (deals that were at this stage and closed won) ÷ (all deals that reached this stage).
- Use this conversion rate as the stage probability - it reflects the real funnel, not an optimistic one.
- Recalibrate annually or after a significant change in sales motion, product, or market conditions.
A useful sanity check: if your Proposal Sent stage has a 70% probability but your actual win rate from that stage is 30%, your weighted pipeline overstates expected revenue by more than 2×. Leadership decisions made from that forecast will be systematically overconfident.
Required Properties: The Hygiene Layer
Required properties in HubSpot allow you to enforce that specific deal fields are populated before a deal can advance to a stage. This is the most underutilised pipeline configuration feature - and the one with the highest impact on forecast quality.
(Source: HubSpot Knowledge Base - Required properties for CRM records: knowledge.hubspot.com/properties/require-properties-for-crm-records)
The logic: each pipeline stage should have a short list of required properties that confirm the deal is as far along as the stage name suggests. A deal cannot enter 'Discovery Complete' without a populated Timeline field, a Budget Indication field, and a Next Step field - because without those fields, there is no evidence the discovery conversation produced the information that 'complete' implies.
Required property configuration in HubSpot: go to Settings > Objects > Deals > Pipelines, select a pipeline, hover over a stage, and click the settings icon. Add required properties from the deal property list. These can include contact and company properties pulled from associated records.
Recommended required properties per stage category:
- At prospect / early stages: Contact owner, lead source, ICP fit indicator
- At discovery and evaluation stages: BANT fields (budget indication, authority confirmed, need, timeline), decision-maker contact name, next step with date
- At proposal and commitment stages: Deal value, expected close date, decision process documented, legal or procurement contact
- At closed lost: Lost reason (from a structured dropdown), competitor (if applicable)
Multi-Pipeline Strategy: When One Is Not Enough
HubSpot supports multiple deal pipelines in the same portal. The right number of pipelines is the minimum needed to meaningfully track distinct sales motions - not the maximum the system permits. Every additional pipeline creates additional reporting complexity and requires additional training for the sales team.
The test for whether a new pipeline is warranted: does the sales motion have a meaningfully different stage sequence, different average cycle length, or different win-rate benchmarks from the existing pipeline? If the answer is yes, a separate pipeline improves reporting accuracy. If the answer is 'it's a slightly different type of deal,' use custom properties within the existing pipeline instead - deal type, channel source, and product line can segment data within a single pipeline without adding a second one.
Forecasting: From Weighted Pipeline to Credible Prediction
HubSpot's Forecast Tool
HubSpot's native Forecast tool (available in Sales Hub Professional and Enterprise) provides a structured view of pipeline by forecast category, weighted value, and rep-level commit. The tool pulls data from the deal pipeline in real time and allows sales managers to adjust rep forecasts upward or downward based on manager judgment. (Source: HubSpot Knowledge Base - Use the forecast tool: knowledge.hubspot.com/sales/use-the-forecast-tool)
Forecast Categories
Every deal in HubSpot can be assigned a Forecast Category - a field that reflects the rep's confidence level independently of the stage probability. The category system allows reps and managers to express nuance that stage probabilities cannot: a deal at 'Proposal Sent' (60% stage probability) might be 'Commit' because the rep knows the decision is effectively made, or 'Pipeline' because the key stakeholder just went on leave.
AI Deal Scoring in 2026
HubSpot's AI-powered deal scoring (available in Sales Hub Enterprise and some Professional configurations) analyses deal properties, engagement history, and activity patterns to produce a predicted likelihood-to-close score. This score differs from the stage probability in an important way: it is contact- and activity-specific, not stage-generic. Two deals at the same stage can have very different AI scores based on email response rates, meeting attendance, and stakeholder engagement.
Using AI deal scores in practice:
- Surface at-risk deals early. A deal with a high stage probability but a low AI score signals that stage progression is outpacing actual engagement. This is a common pattern for deals that reps advance to look good in pipeline reviews but where buyer activity does not support the stage.
- Identify deals ready to push. A deal with a low stage probability but a high AI score may be progressing faster than the stage suggests - high stakeholder engagement, rapid email responses, multiple contacts involved. Flag these for proactive rep outreach.
- Do not replace human judgment. AI scores are inputs to the forecast conversation, not outputs. A score of 23% does not mean a deal is dead; it means recent engagement patterns are weak and warrant a rep conversation about the deal status.
Automating Pipeline Hygiene
Automation is the practical mechanism for maintaining a clean pipeline at scale. Manual hygiene depends on manager discipline in pipeline reviews - which is inconsistent. Automated hygiene fires reliably every time a trigger condition is met. See our marketing automation workflow guide 2026 for workflow design principles that apply to internal sales automations as well as marketing sequences.
The four automation workflows every HubSpot deal pipeline should have:
Stagnation alert. Trigger: deal has not had an activity logged in 14 days AND is not in a Closed stage. Action: create an internal task assigned to the deal owner ('Deal stagnant - last activity 14+ days ago. Update stage or log next step.') This workflow surfaces deals that are being ignored before they distort the forecast.
Close date slip alert. Trigger: deal expected close date is in the past AND deal stage is not Closed. Action: create a task for the deal owner and optionally notify the sales manager. This workflow prevents the silent accumulation of slipped deals that haunt monthly forecasts.
Lost reason enforcement. Trigger: deal moved to Closed Lost. Action: if Lost Reason property is empty, re-open the deal and create a task: 'Please select a lost reason before closing this deal.' This workflow is a fallback for cases where required properties are bypassed.
Discovery follow-up. Trigger: deal enters Discovery Complete stage AND no meeting is scheduled for the next 14 days. Action: create a task for the deal owner: 'Next step not scheduled - book solution presentation within 14 days.' This workflow prevents discovery from becoming a dead end rather than a funnel accelerator.
Pipeline Hygiene Checklist
Deal Pipeline Best Practices for 2026
- Review the pipeline weekly as a team, not just individually. A weekly pipeline review meeting where reps present their top deals - with required properties populated and stage progression justified - is more effective at maintaining hygiene than any automation. The automation catches what slips through; the review builds the habit.
- Audit stage probabilities against actual win rates annually. Stage probabilities set at implementation drift from reality as the sales motion, market, or product changes. Pull a trailing 12-month win rate by stage and update probabilities accordingly. A 10-percentage-point discrepancy between probability and actual win rate is a material forecast error.
- Limit deal properties to what is genuinely used. A deal record with 40 custom properties where 30 are always empty provides no value and makes the interface harder for reps to use. Review deal properties quarterly and archive those with less than 50% population rate across active deals.
- Standardise deal naming conventions. A deal named 'John Smith' in one rep's pipeline and 'Acme Ltd - Enterprise' in another's makes reporting and deduplication harder. Establish a convention (Company Name - Deal Type - Year, for example) and enforce it through training and periodic audits.
- Use deal tags or custom properties for segmentation, not separate pipelines. The instinct to create a new pipeline for every deal variation (SMB vs enterprise, product A vs product B) leads to a fragmented system that is hard to report across. Use a Deal Type or Product Line property to segment within a pipeline unless the stage sequence is genuinely different.
- Connect deal creation to marketing attribution. Every deal should have an associated contact with an original source tracked in HubSpot. This connection enables marketing to report on pipeline influenced - the total deal value in the pipeline that has a marketing touchpoint. Without it, marketing attribution data is unavailable at the deal level.
Work with Belt Creative
Belt Creative implements HubSpot for B2B teams including Sales Hub pipeline design, deal stage configuration, required property setup, forecast tool training, and the automation workflows that maintain pipeline hygiene. If you are setting up a new HubSpot pipeline or auditing an existing one that is producing unreliable forecasts, we can help.
See our work for examples of HubSpot implementations, or get in touch to discuss your pipeline and forecasting requirements.
A Note on Sources
HubSpot Knowledge Base - Deal pipelines and deal stages: knowledge.hubspot.com/deals/create-deal-pipelines-and-deal-stages. HubSpot Knowledge Base - Required properties: knowledge.hubspot.com/properties/require-properties-for-crm-records. HubSpot Knowledge Base - Forecast tool: knowledge.hubspot.com/sales/use-the-forecast-tool. All HubSpot feature and configuration guidance verified against official documentation, mid-2026. HubSpot's product evolves - verify current feature availability at knowledge.hubspot.com.
Frequently Asked Questions
How Many Stages Should a HubSpot Deal Pipeline Have?
Between 6 and 10 active stages for most B2B sales processes. Fewer than 6 and you lose the resolution needed to identify where deals are stalling and why. More than 10 and reps find stage advancement burdensome - deals get skipped forward or the pipeline becomes a checkbox exercise rather than a meaningful representation of the sale. If your process genuinely requires more than 10 stages, examine whether some stages can be captured via required properties rather than separate stages.
How Do I Set Stage Probabilities Correctly?
Calculate the actual win rate from each stage across the last 12 months of closed deals: (deals that reached this stage and closed won) ÷ (all deals that reached this stage). Use that conversion rate as the stage probability. If you do not have historical data, start with conservative estimates and update after 6 months of data. Never set probabilities based on how the team feels about deals at that stage - use the data.
Should I Use Multiple Deal Pipelines?
Only if the sales motions are genuinely different - different stage sequences, different average cycle lengths, or different win-rate benchmarks. The most common legitimate use cases for multiple pipelines are separating new business from renewals, or separating a high-velocity SMB motion from an enterprise motion. If the difference is primarily in deal value, company size, or product type but the stage logic is the same, use a custom property to segment rather than a new pipeline.
What Is the Difference Between Stage Probability and AI Deal Score?
Stage probability is a static value set by pipeline configuration - it applies equally to every deal at that stage, regardless of the specific deal's engagement level. AI deal score (Sales Hub Enterprise) is calculated dynamically per deal based on contact activity, email engagement, meeting history, and deal properties. A deal at 'Proposal Sent' (60% stage probability) might have an AI score of 18% if the prospect has gone quiet, or 82% if stakeholder engagement has been high. Use both: stage probability for weighted pipeline totals; AI score for identifying specific deals that need attention.
How Do I Handle Deals That Go Dark After Reaching a Late Stage?
Configure a stagnation alert automation that fires at 14 days without logged activity and creates a task for the deal owner. At 30 days without activity at Proposal Sent or later, create a manager-level task and flag the deal in the weekly pipeline review. At 45 days without activity, enforce a decision: close-lost with a reason, or update the close date with documented evidence the deal is still alive. A deal at 80% probability that has had no activity in 45 days is almost certainly not an 80% deal - pipeline hygiene requires naming that reality.
Can I Create Custom Deal Properties and Use Them in Pipeline Reporting?
Yes. Custom deal properties are created in Settings > Properties > Deal Properties. They can be made required for specific pipeline stages, included in deal list views and reports, and used as filters in HubSpot's reporting tools. Common useful custom deal properties for B2B: ICP fit score (numeric), primary use case (dropdown), competitor in evaluation (dropdown), legal/procurement contact name, and ARR or MRR (if different from the standard deal amount field).
Sources: knowledge.hubspot.com/deals; knowledge.hubspot.com/properties; knowledge.hubspot.com/sales. All configuration and feature information verified mid-2026.