Following HubSpot lead scoring best practices starts with a simple question: what makes one prospect more valuable to sales than another? The answer should come from your customer data, sales process, and buying patterns. It should not come from an arbitrary points system where every email click, page visit, and form submission pushes someone closer to becoming an MQL.
HubSpot changed the way lead scoring works when it retired its legacy scoring properties on August 31, 2025. In 2026, Marketing Hub Professional and Enterprise users work with the current Lead Scoring tool, which supports separate fit and engagement scores, combined scoring, positive and negative points, thresholds, event frequency, score decay, inclusion and exclusion lists, and record-level testing.
Those features give marketing teams more control, but the quality of the model still depends on the decisions behind it. A well-built score should help sales distinguish a strong-fit buyer showing active intent from a poor-fit contact who happens to consume a lot of content.
In this detailed blog post, we’ll explain how HubSpot lead scoring works in 2026, which fit and engagement criteria to use, how to set scoring thresholds, where negative scoring and score decay fit, and how to build a model that supports better qualification and sales follow-up.
TL;DR
A useful HubSpot lead scoring model should answer two questions. Is this company or contact a good fit for us? And are they showing enough buying interest to warrant sales attention?
HubSpot now lets teams answer those questions with fit scores, engagement scores, or a combined score.
The strongest models keep the criteria focused, place more weight on actions close to a buying decision, reduce the influence of old activity, and use sales outcomes to set and revise qualification thresholds. HubSpot can then use score properties in workflows, segments, views, and reports.
Key Highlights
- Separate fit from engagement before building an overall qualification model.
- Base fit criteria on customers and opportunities that already match your target market.
- Give pricing, consultation, demo, and product-intent activity more weight than general content engagement.
- Use negative points and exclusions to keep poor-fit records away from sales.
- Apply score decay where old engagement should lose value over time.
- Connect qualification thresholds to ownership, lifecycle stages, notifications, and follow-up.
- Review score performance against opportunities and closed business rather than MQL volume alone.
How HubSpot Lead Scoring Works
HubSpot’s current scoring system provides marketing and sales teams several ways to qualify contacts, companies, and, with the appropriate subscription, deals. Instead of forcing every characteristic and action into one number, you can separate the reasons a lead looks promising.
1. Fit scores measure whether the prospect belongs in your target market
Fit scoring uses record properties. For a B2B company, that may include industry, company size, annual revenue, location, job function, seniority, customer segment, or other information that describes the type of organization you want to sell to.
Suppose your company sells to US healthcare businesses with 100 to 1,000 employees. A company that matches the industry, location, and employee count should receive a stronger fit score than a three-person marketing agency outside your service area.
HubSpot fit scores use property groups rather than engagement events, so the model stays focused on who the prospect is rather than what they have done.
2. Engagement scores measure what the prospect is doing
Engagement scoring looks at activities rather than profile attributes.
A prospect may visit important pages, submit a form, book a meeting, click an email, attend a webinar, or interact several times within a short period. HubSpot lets you score those events and apply criteria around frequency and time frame.
This gives marketing teams more control than assigning the same value every time someone performs an action.
For example, one visit to a pricing page during the past six months may indicate light research. Three pricing-page visits this week suggest something different. Frequency and recency help you separate those two situations.
Read More: HubSpot CRM Lead Management
3. Combined scores keep fit and engagement visible
Combined scores bring the two dimensions together without hiding either one.
HubSpot creates a total score along with separate fit and engagement properties. It can also categorize combined scores from A1 through C3. The letter represents fit, with A indicating high fit and C indicating low fit. The number represents engagement, with 1 indicating high engagement and 3 indicating low engagement.
That produces useful distinctions.
- A1: An A1 prospect is a strong fit and highly engaged. Sales may need to respond quickly.
- A3: An A3 prospect is a strong fit but has shown little recent engagement. Marketing may need to continue nurturing the account.
- C1: A C1 prospect engages heavily but does not match the target customer profile. Sending that contact to sales simply because they accumulated activity points would waste selling time.
This is one of the most useful changes in the current scoring model because it prevents engagement from masking poor fit.
10 HubSpot Lead Scoring Best Practices for B2B Teams
A strong lead scoring model does not need dozens of rules. It needs criteria that help the business make better qualification decisions. Here are the 10 best practices for HubSpot lead scoring for B2B teams, which include:
1. Define Your Ideal Customer Profile Before Assigning Points
Do not start inside the scoring tool. Start with customers and opportunities.
Review the companies that buy successfully from you and stay long enough to become valuable customers. Look for patterns in industry, employee count, annual revenue, geography, company type, job roles involved in the purchase, and products purchased.
Then compare those records with opportunities that stalled or leads sales rejected.
This gives you a factual starting point for fit scoring.
For example, you may discover that companies with 200 to 1,000 employees generate your strongest opportunities, while companies below 50 employees rarely progress beyond the first sales conversation. That difference deserves more attention in a fit score than whether somebody has downloaded three ebooks.
The same principle applies to job roles. If directors and vice presidents regularly participate in closed deals while interns and students rarely influence purchases, your scoring model should reflect that pattern.
You do not need to score every available property. Choose attributes that have a useful relationship with qualification or revenue.
2. Separate Fit From Engagement
One of the most common problems with older scoring models was that they combined every criterion into a single running total. That can produce misleading results.
Imagine two contacts.
The first is a vice president at a target account. They visited your website twice but have not submitted a sales form.
The second is a university student who subscribed to your newsletter, downloaded four guides, and attended two webinars.
A simple additive model may give the student the higher score because they generated more activity. Sales would still prefer to speak with the vice president.
HubSpot’s fit and engagement scoring provides marketing ops teams a better structure for this problem.
| Lead | Fit | Engagement | Recommended Treatment |
| Target VP viewing pricing and services | High | High | Prioritize for sales |
| Target VP reading one article | High | Low | Continue nurture |
| Student downloading several guides | Low | High | Keep out of sales queue |
| Poor-fit inactive contact | Low | Low | Low priority or exclude |
Keep the two dimensions separate during model design. You can combine them later for routing and reporting, but sales should still be able to understand why a record qualified.
3. Score Buying Intent More Heavily Than General Engagement
Not every website action deserves the same value.
A contact reading a broad educational article is engaging with your brand. A contact returning to your pricing page, reviewing implementation services, and booking a consultation is showing stronger commercial intent.
Your score should recognize the difference.
High-intent actions may include demo requests, consultation bookings, pricing-page visits, product comparison pages, implementation pages, case studies close to the prospect’s use case, or direct sales interactions.
Lower-intent actions may include general blog visits, newsletter clicks, early-stage guides, or broad webinar registrations.
This does not mean low-intent content has no value. Those interactions can help you understand interest and support nurture programs. They simply should not carry enough weight to push a weak lead into a sales queue.
HubSpot also lets you group related events and cap the points a group can contribute.
That is useful when you want to prevent repeated low-value activity from overwhelming the score. 20 blog visits should not necessarily be worth more than a consultation request.
Read More: How to Set Up a Lead-to-Close Funnel in HubSpot
4. Give Email Opens and Content Downloads Modest Weight
Email activity has historically received too much influence in lead scoring.
An open shows limited intent. Privacy features and automated email processing can also make open activity less dependable than actions such as a click, form submission, or meeting request.
Content downloads need similar restraint.
A prospect downloading a guide may be researching a problem. That does not automatically mean they are evaluating vendors.
Instead of assigning large point increases for these activities, use them as supporting evidence.
For example, a guide download could add a small amount to engagement. If that same contact then visits your pricing page several times, views a customer case study, and submits a contact form, the combined pattern becomes much more meaningful.
Scoring should reward progression toward a buying conversation, not raw activity volume.
5. Use Negative Scoring to Control Qualification
Positive scoring answers, “What makes this lead more interesting?”
Negative scoring answers, “What should reduce our confidence in this lead?”
You need both.
HubSpot allows scoring rules to add or subtract points. A contact can therefore lose points when they meet criteria that make them less useful to sales.
Suppose a company fits your target size and industry but operates in a country you do not support. That location should reduce the score enough to prevent the record from qualifying.
The same principle may apply to competitors, job seekers, students, unsupported company types, personal email domains, or contacts that sales has already disqualified.
You can also use inclusion and exclusion lists to determine which records should enter a scoring model at all.
This distinction is useful.
If a contact is a poor fit but could become relevant later, negative points may be enough.
If the record should never enter sales qualification, such as a competitor or internal employee, exclusion may be cleaner than maintaining a very low score.
6. Use Score Decay for Time-Sensitive Engagement
Intent changes. Someone who visited your pricing page yesterday deserves more attention than someone who visited six months ago and has shown no activity since.
HubSpot now supports native score decay for engagement and combined score event groups. Teams can reduce the value of an event over intervals of 1, 3, 6, or 12 months.
Use decay where time affects the meaning of the activity.
An email click may lose relevance fairly quickly. A webinar attendance may remain useful for longer. A consultation request may trigger direct follow-up before decay becomes important.
Fit criteria work differently. A company’s industry or employee count does not suddenly become less relevant because three months passed. Those properties should change when the underlying company information changes, not because a timer expired.
This is why fit and engagement should remain separate during model design. Recency affects behavioral interest much more than it affects basic customer fit.
Read more: The Best HubSpot Lead Nurturing Features
7. Set Your MQL Threshold From Sales Outcomes
A score of 50 is not inherently better than 40 or 70.
The number only becomes useful when it separates contacts that sales should work from contacts that need more time.
Start by reviewing historical leads that became sales-qualified leads, opportunities, and customers. Look at the fit and engagement characteristics they had around the time marketing handed them to sales.
Then compare them with leads sales rejected.
You may find that high-fit contacts convert well even with moderate engagement, while lower-fit contacts need substantially stronger buying behavior before sales considers them worth pursuing.
Use those patterns to create the initial threshold.
Once the model is active, keep checking whether leads above the threshold produce stronger sales outcomes than those below it. If sales rejects a large percentage of high-scoring leads, something in the model needs attention. The threshold may be too low, fit criteria may be too loose, or low-value engagement may carry too much weight.
The scoring number should support the handoff decision, not dictate it without context.
8. Connect the Score to a Sales or Marketing Action
A score has little value if it only appears as another CRM property.
Decide what should happen when a record crosses an important threshold.
HubSpot score properties can feed workflows, segments, saved views, and reports. For example, a workflow can assign an owner when an unworked contact passes a score threshold or notify a record owner when a high-value score changes.
A practical handoff often follows this sequence:
Score threshold → qualification check → lifecycle update → ownership → sales notification → follow-up
The specific process will depend on your team.
For one company, an A1 contact may move immediately to an SDR queue. Another company may require an account match and a valid business email before changing lifecycle stage. Enterprise teams may route high-scoring contacts through the account owner rather than round-robin assignment.
Document those decisions before you turn the scoring model on. Sales should know what a qualified score means, why the lead entered their queue, and how quickly the company expects them to respond.
9. Test the Model Before You Activate It
Do not build the score and immediately apply it to the full database. HubSpot lets you test records and preview score distribution before turning a model on.
Use that feature with records you already understand.
Test several strong customers. Test recent opportunities. Test leads sales rejected. Test a competitor. Test a highly engaged contact outside your target market. Test a target account that has shown little engagement.
Then inspect the results.
If known customers receive weak fit scores, your criteria may not reflect your customer base.
If competitors or students appear in your highest score ranges, your negative criteria or exclusions need more work.
If almost every contact lands in the same band, the score is not creating enough separation.
Testing five similar contacts is not enough. Use records that represent the different outcomes your sales and marketing teams see every week.
10. Revisit Scoring as Your Business Changes
Lead scoring should change when your market, product, or sales process changes.
If your company moves upmarket, employee-count and revenue criteria may need revision. If you launch a new product for a different buyer, the old engagement model may no longer work. If sales expands into a new region, a location that previously received negative points may become a strong fit.
HubSpot re-evaluates records when a score changes. Changes can therefore affect workflows, views, segments, and reports that use the score property.
That makes change control important.
Before making a major adjustment, check where the score is used. Review qualification workflows, lifecycle rules, routing, dashboards, and sales views. Then test the revised model before publishing it.
Quarterly review works well for many B2B teams, but the right schedule depends on sales volume and how frequently the business changes.
What Should You Include in a HubSpot Lead Scoring Model?
A practical scoring model usually combines four types of information: fit, engagement, disqualification, and timing.
You do not need a long rulebook. You need enough information to distinguish good prospects from active but unsuitable contacts.
1. Fit criteria
Fit should reflect the characteristics of customers your company can serve successfully.
For a B2B organization, company-level information often carries more weight than individual engagement. Industry, employee count, revenue, geography, account type, and technology environment may all help determine fit.
Contact-level information adds the buying-role layer. A VP of Revenue Operations may have more influence over a CRM purchase than a junior content writer at the same company.
The exact properties depend on your sales model. Use criteria that help explain why an account belongs in your target market.
2. Engagement criteria
Engagement should show movement toward a buying conversation.
Page visits, forms, meetings, email clicks, webinars, product activity, sales conversations, and repeat visits can all contribute, but their point values should reflect their relative importance.
HubSpot also supports time-frame and frequency rules for engagement criteria.
That lets you score patterns instead of isolated events.
A contact visiting an implementation page three times in seven days may deserve a different score from somebody who visited it once nine months ago.
3. Disqualification criteria
Some characteristics should reduce the score or prevent a record from entering the model.
That may include unsupported locations, competitors, students, personal-use inquiries, companies below a minimum size, or sales-disqualified records.
Agree on these rules with sales before launch. If marketing thinks a record belongs in the funnel and sales considers it automatically unqualified, the scoring problem starts with definitions rather than points.
4. Recency and frequency
Recency answers when something happened while frequency answers how often it happened.
Both improve engagement scoring.
A prospect viewing pricing four times this month presents a different level of activity from someone who viewed it once last year. HubSpot’s current tool lets teams include both time frames and event frequency in the scoring criteria.
Use those controls to reward current patterns rather than letting old activity accumulate indefinitely.
HubSpot Lead Scoring Example for a Mid-Market B2B Company
The following table shows how a mid-market B2B company could structure a 100-point qualification framework.
The point values are illustrative. Your company should set its own values using customer, opportunity, and sales data.
| Scoring Area | Criterion | Example Points |
| Fit | Target industry | +15 |
| Company within target employee range | +15 | |
| Decision-making or influential role | +10 | |
| Target geography | +10 | |
| Engagement | Demo or consultation request | +25 |
| Repeat pricing-page visits | +15 | |
| Case study or comparison-page visit | +8 | |
| Product-focused webinar attendance | +5 | |
| Marketing email click | +2 | |
| Negative | Competitor | -50 |
| Student or job seeker | -40 | |
| Unsupported region | -25 | |
| Unsubscribed from marketing | -10 |
The fit and engagement sections each have their own purpose.
A strong-fit company could reach 45 fit points without being ready for sales. If its main contact has only opened a newsletter, marketing may continue nurturing the account.
If the same company later returns to a pricing page several times and submits a consultation form, engagement changes quickly. That may justify immediate sales follow-up.
Now consider a student who attends webinars, downloads resources, and visits several pages. Their engagement may be high, but negative fit criteria should keep the record away from sales.
This is why a scoring model should not rely on a single total without showing where the points came from.
When Should You Use Multiple HubSpot Lead Scores?
One scoring model works well when the company has one target customer profile and one broadly consistent sales motion.
Multiple models become useful when the buying process differs enough that one set of criteria would distort qualification.
Consider a software company that sells an enterprise platform and a smaller self-service product. The enterprise buyer may be a VP at a 2,000-person company who reviews security, integrations, and implementation content over several weeks. The self-service buyer may be the founder of a 20-person company who visits pricing and starts a trial on the same day.
Those buyers should not necessarily share the same scoring rules.
Separate scores may also make sense for different product lines, direct and partner sales, new business and expansion, or regions where the buying process differs significantly.
HubSpot allows businesses to create separate scores and use their score properties in workflows, filters, segments, and reports.
Do not create separate models for minor differences.
Every additional score needs maintenance, documentation, reporting, and sales understanding. If two business units use similar qualification criteria, one governed model may work better than separate systems.
Use multiple models when the buying motions are genuinely different, not because each team wants its own version.
Manual vs. AI-Assisted Lead Scoring in HubSpot
HubSpot supports rules-based scoring and AI-assisted scoring, but they solve slightly different needs.
Rules-based scoring
Rules-based scoring works well when the company understands its ideal customer profile and sales process.
Marketing and sales can see the criteria, understand why a contact earned points, and adjust the model when business priorities change.
This approach is also useful when historical conversion data is limited. A company entering a new market or launching a new service may not have enough past outcomes to train a useful data-led model.
The trade-off is that the team needs to make the scoring decisions itself. If the criteria reflect assumptions rather than customer and opportunity data, the model will inherit those assumptions.
AI-assisted scoring
Marketing Hub Enterprise can create AI-assisted contact fit and engagement scores from historical conversion data.
HubSpot currently requires a minimum training sample of 50 contacts, including at least 25 converted and 25 non-converted contacts, before it can generate a score.
That minimum does not mean every database with 50 records will produce a useful model. Data quality still affects the result.
Conversion stages need consistent definitions. Lifecycle changes need to be recorded reliably. Contact and company properties need enough completeness for HubSpot to identify useful patterns.
For companies with strong historical data, AI-assisted scoring can help identify relationships the team may not have selected manually. For companies with inconsistent lifecycle data or changing sales motions, a transparent rules-based model may be easier to manage.
The decision should come from the quality of the underlying data and the way sales needs to use the score.
How The Automation Strategy Group Can Help
The Automation Strategy Group provides HubSpot consulting services for SMEs across the US, including implementation and onboarding, portal audits and optimization, workflow and email automation, CRM cleanup and data work, HubSpot integrations, and platform migrations.
We also work across Marketing Hub, Sales Hub, Service Hub, Content Hub, Data Hub, and Revenue Hub.
For companies introducing HubSpot, our implementation work covers CRM structure, pipelines, properties, user permissions, training, and handover. For companies moving from another platform, its migration services cover platforms including ActiveCampaign, Eloqua, Marketo, Salesforce, and other CRM or marketing automation systems.
The Automation Strategy Group has been an official HubSpot solution partner in the United States since 2016 and lists accreditations across CRM implementation, solutions architecture design, custom integration, service implementation, and CRM data migration.
Need help with HubSpot consulting services? Schedule a free consultation call with one of our HubSpot experts.
Frequently Asked Questions
What are the best practices for HubSpot lead scoring?
Start by defining your ideal customer profile, then separate fit from engagement. Give more weight to high-intent activity, use negative scoring for poor-fit records, apply decay to time-sensitive engagement, and set qualification thresholds using sales and opportunity data.
How does HubSpot lead scoring work in 2026?
HubSpot’s current Lead Scoring tool supports fit, engagement, combined, and certain deal scoring use cases. Teams can add or subtract points, set thresholds, use event frequency and time frames, apply score decay, and use the resulting score properties in workflows, segments, views, and reports.
What is the difference between fit and engagement scores in HubSpot?
Fit scoring evaluates who the prospect is based on properties such as industry, company size, geography, and role. Engagement scoring evaluates what the prospect does, such as visiting pages, submitting forms, attending meetings, or interacting with marketing. Keeping the two separate prevents high activity from making a poor-fit lead appear sales-ready.
What should I score in HubSpot?
Score criteria that help predict qualification or buying readiness. Fit criteria may include company size, industry, location, and seniority. Engagement criteria may include sales forms, pricing pages, product activity, meetings, webinars, and other actions tied closely to the buying process.
What is a good HubSpot lead score?
There is no universal number. A good threshold is one where contacts above it convert into sales-qualified leads and opportunities at a stronger rate than contacts below it. Build an initial threshold from historical customer and opportunity data, then adjust it as you collect more results.
How should I set an MQL threshold in HubSpot?
Review leads that became SQLs, opportunities, and customers, then compare their fit and engagement with leads that sales rejected. Use those patterns to define the first threshold. Sales and marketing should agree on what happens when a contact crosses it.
Does HubSpot support negative lead scoring?
HubSpot lets scoring rules add or subtract points. Negative scoring can reduce qualification for criteria such as unsupported geography, poor company fit, unsubscribes, or other conditions that make a record less useful to sales.
Does HubSpot lead scoring include score decay?
Engagement and combined score event groups can use native decay. HubSpot currently supports decay intervals of 1, 3, 6, or 12 months so older engagement contributes less to the score over time.
How often should a HubSpot lead scoring model be updated?
Review the model whenever your ICP, sales process, product mix, or territory structure changes. For a stable B2B business, a quarterly review is a sensible starting point. Check the workflows and reports that depend on the score before publishing major changes because HubSpot recalculates records when the scoring model changes.
