Lead Scoring: Fit, Behavior, and Handoff Thresholds

Updated September 20, 2026· 8 min read

Lead scoring is a shared ranking of prospects by customer fit and recent behavior, used to decide who is ready for sales and who should stay in nurture. A useful score gives marketing and sales a consistent, reviewable handoff rule.

Pair scoring with audience segmentation, event tracking, and conversion rate reporting. Reporting and analysis is where those scores get compared with accepted leads, opportunities, and closed revenue.

Fit scoring versus behavior scoring

Keep the two scales separate so a poor-fit contact cannot look sales-ready just because they clicked a lot.

Scale	What it measures	Typical inputs	Failure mode if used alone
Fit	How well the person or account matches the customers you can serve	Title, company size, industry, geography, serviceability	Routes busy people who will never buy
Behavior	How recently they showed purchase research	Pricing visits, demo requests, product use, email clicks	Inflates students, competitors, and one-off browsers
Combined total	Fit plus behavior after decay and penalties	Both scales on one record	Hides which half of the score is doing the work

Explicit scoring uses fields the contact or CRM already stores. Implicit scoring uses tracked actions. A predictive model, such as HubSpot’s Likelihood to close property or Salesforce Einstein Lead Scoring, estimates conversion likelihood from historical records. Treat predictive output as a rank, not as an auditable point list, unless the vendor documents the inputs you can inspect.

Auditable sample model

Here is a worked 30-day B2B services example. The point values illustrate the math; each business should calibrate its own weights against sales outcomes.

Fit (maximum 50)

Signal	Points	Why it is on the card
Job title matches the buyer persona	+20	Decision role
Company size in the served range	+15	Account fit
Industry in the ICP list	+15	Offer match
Student, competitor, or personal-email role	-20	Poor fit

Behavior (maximum 50, last 30 days)

Signal	Points	Why it is on the card
Requested a demo	+30	Direct sales ask
Visited the pricing page	+15	Purchase research
Downloaded gated product content	+8	Early research
Clicked a sales email	+5	Recent attention
No qualifying activity for 30 days	-10	Decay

Score each positive behavior once during the 30-day window and cap the positive behavior subtotal at 50. Keep contactability separate: an unsubscribe suppresses email regardless of the lead’s score.

Worked contact:

• Fit: title +20, size +15, industry +15 = 50
• Behavior: pricing +15, email click +5 = 20
• Total = 50 + 20 = 70

Example: 50 fit points plus 20 behavior points equal a total of 70. The MQL rule also requires minimum fit and behavior scores.

Thresholds you can audit

Write the handoff as a rule, then measure what happens at each stage.

Stage	Rule in this example	What to count
Nurture	Fit below 20, behavior below 10, or total below 40	Contacts still owned by marketing
Marketing qualified lead (MQL)	Fit 20 or more, behavior 10 or more, and total 40 or more	MQLs created in the window
Sales accepted lead	Sales accepts the MQL within 5 business days	Accept rate = accepted / MQL
Opportunity	CRM opportunity created from an accepted lead	Opportunity rate = opportunities / accepted
Close	Won opportunity	Close rate = wins / opportunities
Sales rejected	Sales reviews the MQL and rejects its fit or intent	Rejection rate = rejected / reviewed MQLs
Stalled	MQL receives no disposition inside the service-level window	Stalled rate = undispositioned / MQLs

A falling accept rate is a diagnostic, not a verdict. Check the threshold, source mix, duplicate records, routing delay, and sales capacity before changing weights. If accepted leads rarely become opportunities, review both the score and the sales-ready definition.

Decay belongs on the behavior scale. Old clicks should not keep a contact above the MQL line after the research window has closed. Sales should be able to see, on the record, which points produced the current total.

Lead-scoring tools

These products can store or calculate scores. The team still defines the handoff rule and the outcome used to recalibrate it.

Product	What it is for	Official starting point
HubSpot lead scoring	Custom fit, engagement, or combined scores, with optional High/Medium/Low or A1 to C3 thresholds	Understand the lead scoring tool and build lead scores
HubSpot Likelihood to close	Separate predictive probability that a contact becomes a customer within 90 days	Predictive lead scoring properties
Salesforce Einstein Lead Scoring	Predictive rank after you choose a conversion milestone and fields	Enable Einstein Lead Scoring
Salesforce Account Engagement (formerly Pardot)	Numeric score for activity and letter grade for explicit fit	Lead scoring in Account Engagement and lead grading; product home: Marketing automation
Adobe Marketo Engage	Separate demographic and behavior score fields, updated with Change Score flow steps	Build person scoring models and Change Score

Confirm live edition limits in the account. A score property is only useful if sales can see it and can reject the handoff.

How to launch and recalibrate

1. Write the sales-ready definition with sales in the room.
2. Score fit and behavior on separate fields, even if you also store a total.
3. Start with a short list of high-signal rules. Extra rules make the score harder to audit.
4. Set the MQL threshold with minimum fit, minimum behavior, and total-score conditions.
5. Decay behavior on a stated clock. Handle unsubscribes and other contact restrictions as suppression rules, independent of the score.
6. Review the model frequently during launch, then move to a stable cadence. Compare scores with acceptance, rejection, stalled handoffs, opportunities, and closes before changing weights.

Lead scoring should predict a handoff decision that sales can validate. Separate fit from behavior, record why points were assigned, and review closed outcomes so the model does not reward activity that never becomes qualified demand.

Video explainer

Video: How to Score and Prioritize Leads Like a Pro with HubSpot by HubSpot Academy, published September 5, 2025.

Frequently Asked Questions

What is lead scoring in marketing?

Lead scoring assigns points for customer fit and for observed behavior, then uses agreed thresholds to decide who is ready for sales follow-up and who stays in nurture.

How do you create a lead scoring model?

Write the handoff rule first. Score fit and behavior on separate scales, set an MQL threshold, log why points were given, decay stale activity, and review accepted leads, opportunities, closes, and false positives with sales.

What is an example of lead scoring?

In a simple B2B model, a matching title, company size, and industry can total 50 fit points, while a recent pricing visit and email click add 20 behavior points. If MQL requires at least 20 fit points, 10 behavior points, and 40 total points, that contact is eligible for sales review.

Why is lead scoring important?

It gives marketing and sales a consistent way to prioritize follow-up, explain why a lead was routed, and improve the rule using accepted leads, opportunities, closed business, and sales feedback.

Sources

• HubSpot Knowledge Base, Understand the lead scoring tool, accessed September 20, 2026.
• HubSpot Knowledge Base, Build lead scores to qualify contacts, companies, and deals, accessed September 20, 2026.
• HubSpot Knowledge Base, Determine likelihood to close with predictive lead scoring, accessed September 20, 2026.
• Salesforce Help, Enable Einstein Lead Scoring, accessed September 20, 2026.
• Salesforce Trailhead, Lead scoring in Account Engagement, accessed September 20, 2026.
• Salesforce Trailhead, Lead grading in Account Engagement, accessed September 20, 2026.
• Adobe Experience League, Build person scoring models for Marketo Engage, accessed September 20, 2026.
• Adobe Experience League, Change Score, accessed September 20, 2026.