Pipeline volume is not pipeline value. A board showing 4x coverage can still miss the quarter if the underlying deals are the wrong shape. Quality scoring forces a structured answer to the question every revenue leader already asks informally: is this deal real?
Step 1: Define the Dimensions of Quality
A useful pipeline quality score combines four independent dimensions. Score each deal on all four before weighting them into a composite.
ICP fit. Does the company match your ideal customer profile on the firmographic signals that correlate with closed-won deals in your segment? Common inputs include industry, employee count, tech stack, and whether a known pain category applies. Fit is binary or tiered, not a sliding scale. Engagement depth. Have the right people at the prospect actually engaged? Champion activity (opens, meetings attended, replies) matters less than whether the economic buyer has been in a conversation. A deal where only the champion is active is a different quality tier than one where finance or the C-suite has participated. Stage completeness. Each stage in your sales process should have a required set of outcomes before a deal advances. If a deal is listed as "Proposal Sent" but no business case has been documented or no mutual close plan exists, the stage label is wrong. Score completeness as the ratio of required stage milestones that are actually confirmed in CRM. Economic buyer access. Has a revenue leader, CFO, or budget holder been on a call? For deals above a certain contract value, no economic buyer contact is a structural risk. Score this as a binary signal applied as a multiplier.Step 2: Build the Composite Score
A simple weighting model works better than a complex one you won't maintain. Assign weights that reflect your business. A common starting point:
| Dimension | Weight |
|---|---|
| ICP Fit | 20% |
| Engagement Depth | 30% |
| Stage Completeness | 30% |
| Economic Buyer Access | 20% |
Step 3: Set Cut Scores for Each Forecast Category
Your forecast categories need defensible thresholds tied to quality scores.
Commit deals should require a quality score above your agreed threshold and must have economic buyer access confirmed. Best-case deals can carry lower scores but should still show engagement signals. Pipeline deals can remain below threshold, but require a plan and an explicit owner.
Deals that fall below the floor score on two consecutive reviews should be moved to a separate bucket. Label it clearly. The point is not to delete the opportunity but to stop counting it as forecast coverage until something changes.
Step 4: Score at Review Cadence, Not Just Stage Gates
Quality scores decay. A deal does not maintain its score by sitting still. Build the scoring framework into your pipeline review so scores update before every weekly or biweekly call. Reviews are most useful when reps walk in with current scores visible, not when scoring happens as an afterthought during the call.
Step 5: Use the Score to Drive Specific Actions
A quality score only earns its place if it changes behavior. For each low-scoring dimension, define the action that would move the score.
Low ICP fit: this deal may not belong in your pipeline at all. Qualify out or deprioritize resource allocation.
Low engagement depth: the rep needs to bring a new stakeholder into the conversation within a defined window.
Low stage completeness: identify which milestone is missing and block advancement until it is documented.
No economic buyer access: make EB access a defined next step with a deadline.
Common Mistakes
Treating score as a ranking, not a filter. Scores exist to identify which deals do not meet the bar for forecast inclusion, not to rank deals against each other. Letting reps self-score without verification. Scoring should be validated against CRM data and call records, not self-reported. Otherwise you are scoring intent, not reality. Using the same model across all segments. Enterprise deals and mid-market deals have different quality signals. A single universal model will systematically mis-score one segment. Ignoring score trend. A deal moving from 75 to 55 over three reviews is a different risk than a deal that has stayed at 55. Trend is as important as current score.Frequently Asked Questions
What makes a pipeline quality score different from a win rate?
Win rate is a laggard metric that averages historical outcomes. A pipeline quality score is a forward-looking signal applied to each open deal before it closes. It tells you, right now, whether a specific deal deserves to stay in the forecast.How often should we re-score pipeline?
Score at every deal review and at each stage transition. Static scores decay fast. A deal that was high-quality two weeks ago may have gone cold if the champion left or the economic buyer never engaged.What is a reasonable cut score for removing deals from the forecast?
There is no universal number. Set your threshold by back-testing: look at deals that closed versus deals that slipped in the prior two quarters and find the score that best separated the two groups. The right cut score depends on your model weights and your business, not an industry average.Can quality scoring work if our CRM data is incomplete?
Partially. Score what you can, and treat missing required fields as a negative signal. If you cannot confirm economic buyer access for a deal over a certain size, the absence of that data is itself a quality problem.For a structured look at how quality fits into broader pipeline health, see the guides on pipeline quality score and predictive deal scoring.
Frequently Asked Questions
What makes a pipeline quality score different from a win rate?
Win rate is a laggard metric that averages historical outcomes. A pipeline quality score is a forward-looking signal applied to each open deal before it closes. It tells you, right now, whether a specific deal deserves to stay in the forecast.
How often should we re-score pipeline?
Score at every deal review and at each stage transition. Static scores decay fast. A deal that was high-quality two weeks ago may have gone cold if the champion left or the economic buyer never engaged.
What is a reasonable cut score for removing deals from the forecast?
There is no universal number. Set your threshold by back-testing: look at deals that closed versus deals that slipped in the prior two quarters and find the score that best separated the two groups. The right cut score depends on your model weights and your business, not an industry average.
Can quality scoring work if our CRM data is incomplete?
Partially. Score what you can, and treat missing required fields as a negative signal. If you cannot confirm economic buyer access for a deal over a certain size, the absence of that data is itself a quality problem.
See how ORM turns these insights into action
ORM builds custom revenue forecast models for B2B SaaS companies. Not dashboards. Prescriptive analytics that tell you what to do next.
Schedule a Demo