SyncGTM now works inside Claude and ChatGPT. Learn more
SyncGTM logoSyncGTM logo
  • Platform

    • Find Leads
    • Enrich Data
    • Intent Signals
    • Integrations
    • GTM MCP
    • Enrichment API

    Person Data

    • Work Email
    • Mobile Number
    • Personal Email
    • Experience
    • Verify Emails
    View all →

    Company

    • Techstack
    • Website Traffic
    • Revenue
    • Google Ads
    • LinkedIn Posts
    View all →

    Person signals

    • Job Change
    • Promotions
    • Leadership Change
    • LinkedIn Posts
    • WhatsApp Validation
    View all →

    Account signals

    • Headcount Growth
    • Job Openings
    • Techstack Change
    • Funding
    • Product Launched
    View all →
    • Sales
    • RevOps
    • Agency
    • Startups
    • GTM Engineers
    • CRM Enrichment
    • Account Research
    • Inbound Automation
    • Outbound Automation
    • Monitor Signals
  • Fractional GTMe
  • Pricing
  • Learn

    • Blog
    • Docs
    • Templates
    • Reports
    • Case Studies
    • Webinars
    • Video Tutorials
    • Product Tour

    Resource

    • Affiliate
    • Free Tools
    • Credit Estimator
    • Compare Coverage
    • Agency Directory
    • Product Updates
    • Roadmap
    • Join community

    AI

    • GTM agents
    • GTM skills
  • Login
  • vsClay logoClay
  • Platform

    • Find Leads
    • Enrich Data
    • Intent Signals
    • Integrations
    • GTM MCP
    • Enrichment API

    Person Data

    • Work Email
    • Mobile Number
    • Personal Email
    • Experience
    • Verify Emails
    • View all →

    Company

    • Techstack
    • Website Traffic
    • Revenue
    • Google Ads
    • LinkedIn Posts
    • View all →

    Person signals

    • Job Change
    • Promotions
    • Leadership Change
    • LinkedIn Posts
    • WhatsApp Validation
    • View all →

    Account signals

    • Headcount Growth
    • Job Openings
    • Techstack Change
    • Funding
    • Product Launched
    • View all →
    • Sales
    • RevOps
    • Agency
    • Startups
    • GTM Engineers
    • CRM Enrichment
    • Account Research
    • Inbound Automation
    • Outbound Automation
    • Monitor Signals
  • Fractional GTMe
  • Pricing
    • Blog
    • Docs
    • Templates
    • Reports
    • Case Studies
    • Webinars
    • Video Tutorials
    • Product Tour

    Resource

    • Affiliate
    • Free Tools
    • Credit Estimator
    • Compare Coverage
    • Agency Directory
    • Product Updates
    • Roadmap
    • Join community

    AI

    • GTM agents
    • GTM skills
  • Login
  • vsClay logoClay
SyncGTM logo
support@syncgtm.com
Built withfrom&

Product

  • Lead Finder
  • Data enrichment
  • Intent Signals
  • Integrations
  • GTM MCP
  • Enrichment API
  • Enrichment CLIsoon

Use Cases

  • CRM Enrichment
  • Account Research
  • Inbound Automation
  • Outbound Automation
  • Monitor Signals

Person Data

  • Work Email
  • Mobile Number
  • Personal Email
  • Experience
  • Verify Emails
  • View all →

Company Data

  • Techstack
  • Website Traffic
  • Job Listing
  • Google Ads
  • LinkedIn Posts
  • View all →

Person Signals

  • Job Change
  • Promotions
  • Leadership Change
  • LinkedIn Posts
  • WhatsApp Validation
  • View all →

Account Signals

  • Headcount Growth
  • Job Openings
  • Techstack Change
  • Funding
  • Product Launched
  • View all →

Solutions

  • Sales
  • Agency
  • RevOps
  • Startups
  • GTM Engineers
  • Fractional GTMe

Company

  • Contact
  • Affiliate
  • Sign Up
  • Login
  • Terms
  • Affiliate Terms
  • Privacy

Resources

  • Pricing
  • Credit Estimator
  • Templates
  • Blog
  • Case Studies
  • Webinars
  • Video Tutorials
  • Slack Community
  • Docs
  • Agency Directory
  • Free Tools
  • Product Updates
  • Roadmap

Copyright ©2026 SyncGTM

Featured on SaaSBison
Lead scoring models: fit and engagement scorecards, rule-based vs predictive scoring, HubSpot and Salesforce setup

In this Blog

  • TL;DR
  • What Is a Lead Scoring Model?
  • 6 Types of Lead Scoring Models
  • Fit vs Engagement Scoring
  • Rule-Based vs Predictive Lead Scoring
  • 3 Lead Scoring Model Examples
  • Lead Scoring Template You Can Copy
  • How to Build a Lead Scoring Model in 7 Steps
  • How to Build a Lead Score in HubSpot
  • How to Score Leads in Salesforce
  • The Enrichment Data Lead Scoring Depends On
  • Common Lead Scoring Mistakes
  • Feed Your Lead Score With SyncGTM
  • Conclusion
  • FAQ

Lead Scoring Models: Types, Examples and Templates (2026)

Amy Johnson

By Amy Johnson · November 16, 2026 · 15 min read

Key Takeaway

A lead scoring model ranks leads by two questions: do they fit your ICP, and are they showing buying intent? Rule-based models assign points by hand and work from day one; predictive models learn from past conversions but need enough history. Most B2B teams should run separate fit and engagement scores, add negative rules and decay, and back-test thresholds against real deals. Every fit score is only as good as the firmographic and contact data behind it, so enrich before you score.

Lead scoring models rank every lead by how likely it is to become a customer, so sales calls the right people first. A good model answers two questions at once: does this lead fit who we sell to, and is it showing intent right now?

This guide is for RevOps, marketing ops and sales leaders who own lead routing. It covers the six model types, rule-based vs predictive scoring, three example scorecards, a template you can copy, setup in HubSpot and Salesforce, and the data your score depends on.

TL;DR

  • A lead scoring model assigns points or a probability to each lead based on who they are (fit) and what they do (engagement).
  • Rule-based models use hand-set points and work with little data. Predictive models learn from past conversions and need enough history to train.
  • Keep fit and engagement separate, then combine them in a matrix. HubSpot labels combined scores A1 to C3.
  • Add negative scoring and decay so students, competitors and stale leads drop out of the queue.
  • Back-test before launch. Score last quarter's leads and check that high scores actually became opportunities.
  • Enrich first. Blank title, industry or company size fields make good leads look bad.

What Is a Lead Scoring Model?

A lead scoring model is a set of rules or a statistical model that turns lead data into a single priority number or grade. Sales uses that number to decide who to call, and marketing uses it to decide who stays in nurture.

Every model scores two kinds of data:

  • Explicit data — what the lead or a data provider tells you: job title, seniority, company size, industry, location, tech stack.
  • Implicit data — what the lead does: page visits, form fills, email clicks, demo requests, product usage.

Salesforce's Account Engagement makes the split explicit. Its grading system rates fit from explicit data with a letter grade, while its scoring system adds points for implicit activity.

Lead scoring sits on top of qualification, it doesn't replace it. If you're still defining what a qualified lead looks like, start with our guide on how to qualify a B2B lead.

6 Types of Lead Scoring Models

Most guides list a dozen “models” that are really just criteria. In practice there are six model types, and most teams combine three or four of them.

1. Fit (demographic and firmographic) scoring

Fit scoring rates how closely a lead matches your ideal customer profile. It uses person traits (title, seniority, department) and company traits (industry, employee count, revenue, location, tech stack).

Fit changes slowly, so it is the stable half of any model. A VP of Sales at a 200-person SaaS company fits today and will still fit next month.

2. Engagement (behavioral) scoring

Engagement scoring rates buying intent from actions. High-intent actions such as demo requests or pricing page visits earn more points than low-intent ones such as a blog visit or newsletter open.

Engagement goes stale fast. That is why it needs decay, covered below.

3. Negative scoring

Negative scoring subtracts points for signals that a lead won't buy. Common rules: personal email domain, student or job-seeker titles, competitor domains, careers page visits, unsubscribes and long inactivity.

Salesforce's own Trailhead scoring example subtracts points for viewing a jobs page, because job seekers are not buyers.

4. Combined (matrix) scoring

Combined scoring keeps fit and engagement as separate scores, then reads them together. HubSpot's lead scoring tool labels combined scores from A1 to C3, where the letter is fit and the number is engagement.

So a C1 lead is highly engaged but a poor fit, and an A3 lead is a great fit that hasn't engaged yet. Those two need very different follow-up.

5. Predictive scoring

Predictive scoring uses machine learning on your historical leads to estimate how likely each new lead is to convert. You don't set the weights; the model finds which fields and behaviors separated past wins from losses.

Examples include Salesforce Einstein Lead Scoring and HubSpot's AI-generated scores. More on both below.

6. Account-level scoring

Account scoring rolls individual contact scores up to the company. It matters when several people from one account engage, which is normal in B2B buying committees.

HubSpot company scores can aggregate associated contacts by sum, average, minimum or maximum, per its build lead scores guide. For mapping the people inside those accounts, see our B2B sales qualification guide.

Fit vs Engagement Scoring

Fit tells you whether a lead is worth selling to. Engagement tells you whether now is the time.

 Fit scoreEngagement score
Question it answersShould we sell to them?Are they interested now?
Data typeExplicit: title, industry, size, tech stackImplicit: visits, form fills, clicks, usage
Where data comes fromForms, CRM fields, enrichment providersWebsite tracking, email tool, product analytics
ChangesSlowly (job or company changes)Daily
Needs decay?NoYes
Main failure modeBlank or outdated fieldsRewarding low-intent noise

Never add fit and engagement into one number without keeping the parts visible. A student who downloads ten ebooks can outscore a CFO who visited pricing once, and a single blended score hides that.

Rule-Based vs Predictive Lead Scoring

Rule-based scoring is where almost every team should start. Predictive scoring becomes worth it once you have enough conversion history and clean data.

 Rule-basedPredictive
How weights are setBy you, based on sales input and past dealsBy a model trained on past leads
Data neededWorks from day oneNeeds a history of converted and unconverted leads
ExplainabilityHigh: every point has a reasonLower: shows top factors, not full logic
MaintenanceManual quarterly reviewRetrains automatically
Best forNew teams, new segments, low lead volumeHigh lead volume with years of CRM history

Predictive models have a minimum data bar. Salesforce's Sales Einstein implementation guide (Spring '23 edition) says Einstein Lead Scoring needs at least 1,000 leads created in the last 200 days, with at least 120 converted, and falls back to a global model built on anonymized data when an org has less.

Predictive models also learn your data gaps. If half your converted leads have a blank industry field, the model can't learn that industry matters.

3 Lead Scoring Model Examples

These scorecards are illustrative starting points, not benchmarks. Replace every point value after you back-test against your own closed-won deals.

Example 1: Sales-led B2B SaaS (inbound + outbound)

Fit carries most of the weight because reps work a defined ICP. Fit is capped at 60 and engagement at 40, for a 100-point total.

CriterionTypePoints
Title is VP, Head of or C-level in Sales, RevOps or MarketingFit+20
Title is Manager or Director in the same functionsFit+12
Company has 50–1,000 employeesFit+15
Industry is software, SaaS or IT servicesFit+10
Uses HubSpot or Salesforce as CRMFit+10
Requested a demoEngagement+30
Visited pricing page (last 30 days)Engagement+15
Replied to an outbound emailEngagement+15
Attended a webinarEngagement+8
Personal email domain (gmail.com, yahoo.com)Negative−15
Visited careers pageNegative−10

Routing rule: fit ≥ 35 and engagement ≥ 25 goes to an SDR within the hour. Fit ≥ 35 with low engagement goes to an outbound sequence instead of nurture.

Example 2: Product-led growth (free trial or freemium)

Product usage replaces most marketing engagement. The score flags product-qualified leads that sales should contact.

CriterionTypePoints
Signed up with a work emailFit+10
Company has 20+ employeesFit+15
Completed the key activation actionUsage+25
Invited a teammateUsage+20
Hit a plan limitUsage+20
Connected an integrationUsage+10
No login in 14 daysNegative−20

Routing rule: 60+ triggers a sales-assist email from an account executive. Below 30 stays in automated onboarding.

Example 3: Enterprise grade + score (Account Engagement style)

Enterprise teams often use a letter grade for fit and a points score for engagement, as Salesforce Account Engagement does. Per Trailhead, grades run from F to A+, every prospect starts at D, and each criterion moves the grade by one-third, two-thirds or a full letter.

Grade criterionAdjustment
Company has 1,000+ employees+1 letter
Industry is financial services or healthcare+2/3 letter
Located in North America or Western Europe+1/3 letter
Title is Director or above+1 letter
Department is IT, Security or Operations+2/3 letter

Routing rule: grade B or better plus a score above your back-tested threshold creates a Salesforce task for the account owner. High score with a grade of C or lower goes to nurture.

Lead Scoring Template You Can Copy

Copy this template into a spreadsheet before you touch your CRM. Fill the weight column only after step 2 of the build process below.

GroupCriterionCRM field or eventPointsEvidence
Fit: personSeniorityJob title / seniority__% of won deals with this seniority
Fit: personDepartmentDepartment__Won vs lost split
Fit: companyEmployee rangeNumber of employees__Win rate by size band
Fit: companyIndustryIndustry__Win rate by industry
Fit: companyTech stackTechnologies used__Tools common to won accounts
Fit: companyGrowth signalHiring, funding, headcount growth__Signal present at deal creation
Engagement: high intentDemo or contact requestForm submission__Lead-to-opportunity rate
Engagement: mid intentPricing or comparison pagePage view__Lead-to-opportunity rate
Engagement: low intentContent download, email clickForm / email event__Lead-to-opportunity rate
NegativeBad-fit signalsEmail domain, title keywords, careers page−__Never converted

Then map the two scores into a routing matrix:

 High engagement (1)Medium engagement (2)Low engagement (3)
High fit (A)A1: route to sales nowA2: SDR follow-up this weekA3: outbound sequence
Medium fit (B)B1: SDR follow-up this weekB2: targeted nurtureB3: nurture
Low fit (C)C1: check data, then nurtureC2: nurtureC3: no action

Pay attention to C1. Highly engaged “low fit” leads are often good leads with blank or wrong fields, so check the record before you discard it.

How to Build a Lead Scoring Model in 7 Steps

  1. Pull your closed-won and closed-lost deals. Export the last two to four quarters with the contacts and companies attached. This is your training set, even for a rule-based model.
  2. Enrich the records. Fill title, seniority, company size, industry and tech stack wherever they are blank. You can't find patterns in fields that are empty.
  3. Compare won vs lost. For each trait and action, check how often it shows up in won deals versus lost ones. Traits that appear in both at the same rate get zero points.
  4. Assign points in proportion to the gap. The bigger the difference between won and lost, the more points. Cap fit and engagement groups separately.
  5. Add negative rules and decay. Subtract for bad-fit signals and let engagement points expire.
  6. Back-test and set thresholds. Score last quarter's leads with the new model. Set the MQL line where conversion to opportunity clearly jumps.
  7. Agree the SLA with sales, then review quarterly. Write down what each score band triggers and how fast sales responds. Re-run steps 3–6 every quarter.

For context on what share of qualified leads typically turn into customers, see how many qualified leads convert into sales in B2B.

How to Build a Lead Score in HubSpot

HubSpot's current lead scoring tool replaced the old single HubSpot Score property with separate fit, engagement and combined scores. It is available on Marketing Hub or Sales Hub Professional and Enterprise; contact scores need Marketing Hub, deal scores need Sales Hub, and AI-generated scores need Marketing Hub Enterprise.

The steps below follow HubSpot's build lead scores article:

  1. Go to Marketing > Lead Scoring and click Create score.
  2. Pick the object (Contacts, Companies or Deals) and the score type: Fit, Engagement or Combined.
  3. Add a property group for fit criteria or an event group for engagement criteria. Set a group point maximum so one group can't dominate.
  4. Add rules with add or subtract points. Event rules can be filtered by timeframe and frequency, such as “visited pricing 2+ times in 30 days.”
  5. Turn on Decay scores for engagement groups. HubSpot supports decay every 1, 3, 6 or 12 months.
  6. Choose which records to score: all, or specific segments.
  7. Under Score thresholds, set the ranges for A–C fit and 1–3 engagement.
  8. Use Test a contact and Preview distribution, then Review and turn on.

Each score writes to CRM properties you can use in workflows, lists and lead rotation. If fit rules reference fields such as industry or employee count, check how many contacts actually have those fields filled before you turn the score on.

How to Score Leads in Salesforce

Salesforce gives you three routes, depending on your licenses.

Option 1: Formula field scoring (any edition with custom fields)

Create a number formula field on Lead that adds points with CASE and IF logic. Here is a simple fit score:

CASE(Industry, "Software", 10, "Technology", 10, 0)
+ IF(NumberOfEmployees >= 50 && NumberOfEmployees <= 1000, 15, 0)
+ IF(CONTAINS(Title, "VP") || CONTAINS(Title, "Head"), 20, 0)
+ IF(CONTAINS(Email, "@gmail.com"), -15, 0)

Formula scores are transparent and free, but they only see fields on the Lead record. Engagement usually needs a separate field updated by Flow or your marketing tool.

Option 2: Account Engagement scoring and grading

Account Engagement (formerly Pardot) runs points-based scoring for engagement and letter grading for fit. Per Trailhead, every account ships with base scoring rules you can edit, and automation rules can adjust scores further.

Grading uses a profile of criteria, and the default profile measures company size, industry, location, job title and department. Route leads on both grade and score, not score alone.

Option 3: Einstein predictive scoring

Einstein Lead Scoring analyzes historical field data on the Lead object to score how likely each lead is to convert. Per Trailhead, it is available with Sales Cloud Einstein or High Velocity Sales licenses, not only Account Engagement.

Account Engagement also offers Einstein Behavior Scoring, which gives each prospect a 0–100 engagement score that typically updates every 4 hours. Check the data requirements in the implementation guide before you plan around it.

The Enrichment Data Lead Scoring Depends On

Every fit rule is a question about a field. If the field is blank, the lead gets zero points, and a perfect-fit buyer looks like a C.

Inbound forms rarely collect more than name, email and company, and CRM data goes stale as people change jobs. That makes enrichment the step most scoring projects skip and most scoring failures trace back to.

DataScoring useWhere it comes from
FirmographicsIndustry, employee count, revenue, location fit rulesCompany enrichment from the email domain
Person dataTitle, seniority and department rulesLinkedIn profile enrichment
Technographics“Uses tool X” rules and competitor flagsTech stack detection
Growth and timing signalsHiring, recent funding, headcount growth, leadership changesSignal enrichments
Contact validityNegative points for invalid or personal emailsEmail verification
Intent dataEngagement points for research on third-party sitesIntent data providers

No single provider fills every field, which is why waterfall enrichment checks several providers in sequence. For tech stack rules, see our guide to technographics, and for third-party research signals, our roundup of buying intent data tools.

Common Lead Scoring Mistakes

  • Guessing the points. Weights set in a meeting reflect opinions. Weights set from won vs lost data reflect buyers.
  • One blended score. It hides whether a lead is high-fit or just busy. Keep fit and engagement visible.
  • No decay. A webinar from last year still inflating a score today sends sales after cold leads.
  • Rewarding volume over intent. Ten blog visits shouldn't beat one demo request. Cap low-intent groups.
  • Scoring blank fields. If most records lack company size, your size rule does nothing. Enrich before you launch.
  • No sales feedback loop. If reps reject a third of MQLs, find out which rules sent them and fix those.
  • Set and forget. Your ICP, product and market move. Review the model every quarter.

Feed Your Lead Score With SyncGTM

SyncGTM doesn't replace HubSpot or Salesforce scoring. It fills the fields those scores read, from 50+ data providers, and writes them back.

  1. Pull leads from your CRM. Use the HubSpot lead source or the Salesforce lead source to load contacts or leads into a table. Duplicate rows are removed before anything syncs back.
  2. Fill fit fields. Enrich Organization adds industry, size, location, revenue range and funding. Company TechStack adds the tools each company runs.
  3. Add timing signals. Headcount Growth, Is Hiring, Raised Funding Recently and Leadership Change give your model growth and timing criteria to score.
  4. Clean contact data. Verify Email (0.2 credits) flags invalid addresses for negative scoring, and Find Work Email (1 credit) replaces personal emails with work emails.
  5. Write back and act. Sync enriched fields to HubSpot or Salesforce so the score recalculates. Push high-fit, low-engagement leads to Instantly, Smartlead or HeyReach for outbound, with first lines drafted by the Write Email with AI column.

SyncGTM also connects to Attio, Pipedrive and Close. You can start free and check credit costs on the pricing page.

Conclusion

The best lead scoring models are simple to explain: who fits, who is engaged, and what each combination triggers. Start rule-based, keep fit and engagement separate, and add negative rules and decay from day one.

Then earn the right to go predictive with clean history. Back-test every threshold against real deals, enrich the fields your rules depend on, and review the model with sales every quarter.


Frequently Asked Questions

Scores stuck low because fields are blank?

Enrich job titles, company size, industry and tech stack on every lead before your CRM scores it.

Share Article


Ask AI

Give Your Lead Score the Data It Needs

SyncGTM pulls leads from HubSpot or Salesforce, fills firmographic, technographic and contact fields from 50+ data providers, and writes them back so your scoring rules have something to score. Start free, no credit card required.