HubSpot lead scoring assigns numeric values to contacts based on their attributes and behaviours – creating a composite score that indicates how likely a contact is to become a customer. Done correctly, lead scoring allows sales and marketing teams to prioritise high-value leads for immediate follow-up, route lower-scoring leads to nurture automation, and define a shared, data-grounded definition of what constitutes a Marketing Qualified Lead (MQL). HubSpot offers two distinct lead scoring approaches in 2026: traditional manual lead scoring (available on Marketing Hub Professional) and HubSpot AI lead scoring (Breeze AI-powered predictive scoring, available on Marketing Hub Professional and above). This guide covers both approaches, how to set them up, and how to build the workflows that make lead scores actionable.
It makes the score easier to trust.
That keeps the model grounded in actual buying behaviour.
It also helps when the setup is easy to tie back to real conversion signals.
The best guide is the one that makes prioritisation feel less random.
A practical explanation should help the reader see where lead scoring helps most in the workflow.
That means the guide should connect the scoring logic to real sales activity.
For many teams, the value is in turning prioritisation into a consistent process rather than a subjective one.
It should also show how the scoring model supports better follow-up and cleaner handoffs.
A good guide should explain why scoring is useful before getting into the setup details.
HubSpot lead scoring is useful because it helps teams prioritise leads based on behaviour and fit instead of treating every contact the same. That makes it easier to focus sales attention where it is most likely to matter.
It makes the score easier to trust.
That keeps the model grounded in actual buying behaviour.
It also helps when the setup is easy to tie back to real conversion signals.
The best guide is the one that makes prioritisation feel less random.
A practical explanation should help the reader see where lead scoring helps most in the workflow.
That means the guide should connect the scoring logic to real sales activity.
For many teams, the value is in turning prioritisation into a consistent process rather than a subjective one.
It should also show how the scoring model supports better follow-up and cleaner handoffs.
A good guide should explain why scoring is useful before getting into the setup details.
HubSpot lead scoring is useful because it helps teams prioritise leads based on behaviour and fit instead of treating every contact the same. That makes it easier to focus sales attention where it is most likely to matter.
It makes the score easier to trust.
That keeps the model grounded in actual buying behaviour.
It also helps when the setup is easy to tie back to real conversion signals.
The best guide is the one that makes prioritisation feel less random.
A practical explanation should help the reader see where lead scoring helps most in the workflow.
That means the guide should connect the scoring logic to real sales activity.
For many teams, the value is in turning prioritisation into a consistent process rather than a subjective one.
It should also show how the scoring model supports better follow-up and cleaner handoffs.
A good guide should explain why scoring is useful before getting into the setup details.
HubSpot lead scoring is useful because it helps teams prioritise leads based on behaviour and fit instead of treating every contact the same. That makes it easier to focus sales attention where it is most likely to matter.
Manual Lead Scoring vs HubSpot AI Lead Scoring
Before configuring lead scoring, choose the right approach for your organisation:
- Manual (rule-based) lead scoring: You define the scoring rules – assign positive or negative point values to specific property values and contact behaviours. Full control over what gets scored and why. Requires ongoing maintenance as the ICP and scoring logic evolve. Appropriate when your team has clear hypotheses about what predicts a quality lead and wants full transparency into why a contact scored high or low
- HubSpot AI Lead Scoring (Breeze Intelligence): A machine learning model trained on your historical contact conversion data – contacts who became customers vs. those who didn’t. The model identifies patterns in contact attributes and behaviours that correlate with conversion and assigns a score accordingly. Less manual maintenance, but requires sufficient historical data (typically 500+ converted contacts) to build a meaningful model. Less transparent than manual scoring – you see the score but not always the full explanation of why
Most organisations start with manual scoring (available on Marketing Hub Professional without additional Breeze credits) and move to AI scoring as they accumulate sufficient conversion data. Both can coexist – using manual scoring for known signal factors and AI scoring for holistic ranking.
Setting Up Manual HubSpot Lead Scoring
Step 1: Define Your Ideal Customer Profile (ICP) Criteria
Before creating scoring rules, define the characteristics of contacts and companies that convert to customers at the highest rates. Consult your Closed Won deal history to identify patterns:
- Firmographic signals (positive): Company Industry = Technology, Company Size = 100-1,000 employees, Annual Revenue ? $10M, Country = United States, Canada, or United Kingdom
- Demographic signals (positive): Job Title contains “VP”, “Director”, “Head of”, or “Manager” of the relevant buying function; Seniority level = Director or above
- Negative signals: Company size < 10 employees, personal email domain (gmail.com, yahoo.com, hotmail.com – indicates consumer or student rather than business buyer), Country in regions where you don’t sell
Step 2: Define Behavioural Engagement Signals
Behavioural signals indicate active interest and buying intent – they should be weighted more heavily than demographic signals because a qualified demographic profile without engagement is a cold contact, while high engagement indicates active investigation:
- High-value engagement (20-30 points): Visited pricing page, submitted a demo request form, viewed a product video, attended a live webinar, opened a sales sequence email
- Medium engagement (10-15 points): Downloaded a whitepaper or case study, subscribed to newsletter, opened a marketing email and clicked a link, visited the website 3+ times in 7 days
- Low engagement (5 points): Opened a marketing email (without clicking), visited a blog post once, followed on LinkedIn
- Negative engagement (-10 to -20 points): Email marked as spam, visited careers page only (indicates job seeker not buyer), contacted support before becoming a customer (indicates existing customer or reseller inquiry)
Step 3: Configure Scoring in HubSpot
- Navigate to Settings ? Properties ? Contact Properties
- Search for “HubSpot Score” – this is HubSpot’s default contact scoring property, a calculated property that accumulates points from your defined scoring rules
- Click Edit to open the scoring configuration
- Click “Add criteria” to add positive or negative scoring rules
- For each rule, define: the property or activity, the condition (e.g., “Industry is any of Technology, Software”), and the point value to add or subtract
- Repeat for each scoring criterion
- Save when complete – HubSpot recalculates all existing contact scores immediately based on the new rules
Step 4: Define Your MQL Threshold
Decide the HubSpot Score threshold at which a contact qualifies as a Marketing Qualified Lead (MQL) – the score at which the contact should be handed off to the sales team for follow-up. There is no universal right answer – the threshold should be calibrated based on:
- What score range represents contacts with both profile fit and genuine engagement signals
- The volume of MQLs your sales team can handle – if your threshold produces 500 MQLs per month and your SDR team can work 200, either raise the threshold or hire more SDRs
- The quality feedback from sales on the leads they receive at the current threshold – if reps consistently tell you the MQLs they receive are unqualified, raise the threshold
Start with a threshold in the 50-80 range for a score that maxes out at 100+, and adjust based on MQL-to-SQL conversion rate after 60-90 days of data.
Building the MQL Workflow
The HubSpot Score only delivers value when connected to automation that acts on the score. Build a Contact Workflow triggered by the HubSpot Score:
Trigger: Contact Property “HubSpot Score” is greater than or equal to [MQL threshold] AND Lifecycle Stage is not “Marketing Qualified Lead” (prevents re-triggering for already-qualified contacts)
Actions:
- Set Lifecycle Stage = “Marketing Qualified Lead”
- Rotate contact to owner (round-robin assign to the SDR queue)
- Send internal email notification to the assigned SDR: “New MQL assigned: [Contact First Name] [Last Name] at [Company] – Score: [HubSpot Score]. Follow up within 2 hours.”
- Create a task for the SDR: “Call or email [Contact Name] – MQL from [Lead Source]” – due in 2 hours
This workflow ensures that every contact crossing the MQL threshold is immediately assigned, the right SDR is alerted, and a time-bound task is created – preventing MQLs from sitting unworked in the lead queue.
HubSpot AI Lead Scoring (Breeze Intelligence)
HubSpot’s AI lead scoring, branded as Breeze Intelligence Buyer Intent scoring, is available on Marketing Hub Professional and above with Breeze Intelligence credits. The AI model analyses:
- Historical conversion data from your HubSpot portal – which contacts became customers and which didn’t
- Firmographic and demographic attributes of converted vs. non-converted contacts
- Engagement behaviour patterns (email opens, clicks, website visits, form submissions) of high-converting vs. low-converting contacts
The AI score is surfaced as a separate property (“Likelihood to Close” or a Breeze Intelligence scoring property) alongside the manual HubSpot Score – the two can be used together or independently for contact prioritisation.
Lead Scoring Maintenance
Lead scoring is not a one-time configuration – it requires quarterly review to remain calibrated to actual conversion patterns:
- MQL-to-SQL conversion rate: Report on what percentage of contacts who reached MQL status (were assigned to sales) were accepted as SQLs by the SDR team. If <30% are accepted, the MQL threshold is too low and needs to be raised or scoring rules adjusted
- High-score contacts who don’t convert: Identify the specific profile characteristics of contacts who score high but don’t convert – these point to scoring rules that overweight characteristics that don’t actually predict conversion in your business
- Score distribution review: Run a report of contact score distribution – if 80% of contacts score 0-10 and only 2% score above the MQL threshold, the score distribution may not provide useful differentiation. Consider adjusting point values to create better score spread
- Update for new signal types: As your website, content, and product evolve, new high-signal pages and forms emerge – add scoring rules for new pricing page variants, new product tour pages, or new webinar registration forms as they launch
What’s a good MQL lead score threshold in HubSpot?
There is no universal answer – the right MQL threshold is the one calibrated to your specific sales process and historical conversion data. As a starting point for new setups, many B2B companies use 40-60 points as an initial threshold, then adjust after 60-90 days of real data. The goal is to set a threshold where contacts passed to sales convert to pipeline at a rate you’ve agreed upon between marketing and sales – typically 20-30% for most B2B businesses.
Does HubSpot’s lead scoring support predictive AI scoring?
HubSpot Professional and Enterprise tiers include Predictive Lead Scoring, which uses machine learning to analyze patterns in your Closed Won contacts and score new contacts automatically based on their similarity to past customers. Predictive scoring works best when you have at least 200 closed deals in HubSpot for the model to learn from. If you’re below that threshold, manual rule-based scoring gives you more control and transparency until your deal history is large enough to train a reliable model.
How often should I audit and recalibrate lead scores?
Lead score models drift over time as your ICP, messaging, and product evolve. A score built 18 months ago may reward behaviors from an old campaign that no longer reflects your current buyer. Audit your model quarterly: compare the lead scores of recent Closed Won contacts against recent Closed Lost contacts, identify which scoring criteria actually differentiate the two groups, and remove or reweight criteria that no longer predict conversion. This continuous calibration keeps your MQL definition aligned with current sales reality.
Can I have different lead scores for different product lines or segments?
HubSpot’s standard lead scoring uses a single score per contact, but you can replicate multi-segment scoring using custom score properties. Create a separate calculated or manual score property for each product line or segment (e.g., “Enterprise Score” and “SMB Score”), apply different criteria to each, and build separate MQL workflows for each segment. This lets your enterprise and SMB sales teams work from scoring models calibrated to their respective buyer journeys rather than sharing one blended score.
Problem: Lead Score Thresholds Set Arbitrarily Without Data
Most HubSpot lead scoring setups assign point values based on intuition – 10 points for a page view, 20 for a form fill, 30 for a demo request – without any data linking those scores to actual sales outcomes. The result is an MQL threshold that may be too low (passing unqualified contacts to sales) or too high (burying sales-ready leads in marketing). Fix this by looking at the last 6 months of Closed Won contacts in HubSpot and calculating what their lead score was at the moment they were handed to sales. Set your MQL threshold at the score that correlates with at least a 20% pipeline conversion rate based on real data.
Problem: Positive Score Inflation From Passive Engagement
Lead scores that climb solely from email opens and page views can reach MQL status for contacts who are students, competitors, or researchers with no buying intent. A prospect reading every blog post isn’t necessarily ready to buy. Fix this by adding explicit negative scoring for disqualifying signals: -15 for a competitor email domain, -10 for a student job title, -20 for unsubscribing from sales emails, -5 for no website activity in 45 days. This decay and penalty system keeps scores representing genuine purchase intent rather than passive consumption.
Problem: Lead Score Not Visible in the Sales Rep’s Day-to-Day View
Even a well-calibrated lead score is useless if sales reps never see it. Many HubSpot setups calculate a score but never surface it prominently in the CRM view that reps actually use. Fix this by adding the Lead Score property to the default contact view, pinning it to the sidebar in HubSpot’s CRM, and including it as a column in the contacts list view sales reps use for daily prospecting. Better yet, build a HubSpot view filtered to “Lead Score is greater than [MQL threshold]” that reps check each morning as their priority call list.
The best scoring setup is the one that reflects real lead quality. If the model is built on weak signals, the results are less useful.
The best scoring setup is the one that reflects real lead quality. If the model is built on weak signals, the results are less useful.
The best scoring setup is the one that reflects real lead quality. If the model is built on weak signals, the results are less useful.