LEAD SCORING MODELS: TYPES, REVENUE BRACKETS AND FORMULAS

• Score revenue in four to six bands, and give the most points to the band with your highest win rate, which often isn't the biggest one.

• A weighted sum of industry, revenue, headcount and region gives you a fit score out of 100 that you can build in any CRM.

• Only route leads to sales when fit and engagement both clear a threshold, so engaged students and perfect-fit accounts that haven't engaged stay out of reps' queues.

• B2B buyers decide in groups, mostly before they fill in a form, so pair your lead score with a view of the whole buying group in each account.

A lead scoring model is the set of rules that turns what you know about a prospect into a number your sales and marketing teams both trust. Each attribute or action earns points (or loses them), and the total decides what happens next, whether that's a sales call, a nurture track, or the bin.

Most models have two halves.

When a lead's combined score crosses an agreed threshold, it becomes a marketing qualified lead (MQL). If sales accepts it and confirms there's a real opportunity, it becomes a sales qualified lead (SQL).

Done well, a lead scoring model gives you:

• A shorter queue for sales, filled with leads that are more likely to close

• Focused conversations, because reps can see what a prospect has already been interested in

• A shared definition of "qualified" 

• Evidence for where to spend budget, because you can see which sources produce quality, high-scoring leads

There are five main types of lead scoring models, and most B2B teams combine at least three of them. They split into explicit scoring, based on hard facts a lead gives you or that are publicly known, and implicit scoring, based on behavior you observe.

In B2B, demographic scoring should focus on the job someone does, so skip personal attributes like age and gender. A VP of Marketing at a target account is worth more points than an intern at the same company, and both are worth more than a student.

Negative scoring

Points should come off as well as go on. Negative scoring stops leads with no buying intent from drifting over the threshold on activity alone. Common deductions include:

• Competitor email domains, since they're probably checking out your pricing

• Careers page visits, because they may want a job rather than your product

• Personal email addresses (Gmail, Yahoo) if you sell B2B

• Email unsubscribes, as they've told you they don't want to hear from you

• Company size well below your smallest deal

• Long inactivity. Let behavioral points decay, for example halving them after 90 days without engagement, so last year's webinar attendee doesn't look hot today.

To score revenue brackets, split annual company revenue into four to six bands, then give each band points based on how closely it matches your ideal customer profile. The bands that close fastest and biggest get the most points. Then weight revenue against other firmographics so it shapes the score without dominating it.

Here's a sample scale for a B2B SaaS company whose best customers sit in the mid-market and lower enterprise. Swap the points to fit your own closed-won data.

Notice the top band doesn't get the most points. Bigger isn't always better. Score the band that wins most often for you, and let your CRM data set the curve.

How to weight revenue against other firmographics

Revenue is one signal among several. Score it alongside industry, employee count, and region, assigning a weight to each attribute based on how strongly it predicts a closed deal. A simple approach is to use a 10-point weight allocation.

Review the weights every quarter. If revenue band turns out to predict wins better than industry, shift a point across.

A simple lead scoring formula for revenue brackets

Most revenue-bracket scoring models use a weighted sum. Score each attribute from 0 to 10, multiply by its weight, then add the results together.

Fit score = (industry points × 3) + (revenue points × 3) + (employee points × 2) + (region points × 2)

Worked example for a $120M financial services company with 800 employees, based in a core region:If you run scoring in a CRM like HubSpot or Salesforce, you can set this up with a calculated property or score property that maps revenue ranges to points. Most marketing automation tools let you build the same logic with if/then rules.

Score engagement by how close an action sits to a buying decision. A blog read earns a point or two, and a demo request earns enough to route a lead to sales on its own. Map your signals to four stages and give each stage a points range.

The content you serve should help people move to the next stage:

• Awareness: useful, shareable content on the problem, plus surveys and polls that tell you what your audience is wrestling with

• Consideration: thought leadership, case studies, and personalized content that makes your approach relevant to their industry

• Evaluation: honest comparison pages, clear navigation, and personalized pages for target accounts

• Decision: social proof, pricing and comparison tables, and your strongest testimonials

Don't write off low scores

Most of the engagement you score will be early-stage, and that's normal. As Kerry Cunningham, Head of Research & Thought Leadership at 6sense, puts it in his foreword to The Intent Activation Blueprint, "roughly 60% of a well-defined ICP is not in-market at any given time. Of the remaining 40%, about three-quarters are in early journey stages, months or quarters from buying, or even from taking a meeting."

A low engagement score from a high-fit account means "not yet". Keep those accounts in a nurture track built to help them define the problem, because that's when buyer preference forms. Sending a BDR after them to book a meeting rarely works.

Treat fit and engagement as two separate thresholds, and only pass a lead to sales when both are cleared. A common setup is a fit score of 60+ and an engagement score of 50+. That stops a highly engaged student or competitor from becoming an MQL, and stops a perfect-fit account that hasn't engaged from clogging a rep's queue.

Plot every lead on a simple grid and give each box its own next step.

The high-fit, low-engagement box is the one most teams neglect. These are the accounts you'd most like to win, and they haven't noticed you yet. That's a marketing job, and it's where account-based programs earn their budget.

In account-based marketing (ABM), the same fit formula works at account level. Score every company on your target account list for fit, using revenue band, industry, and headcount, and you've got a ranked list you can split into tiers.

- 1:1 (Tier 1): your highest-scoring accounts, with bespoke content for each one

- 1:few (Tier 2): the next band, grouped into clusters by industry or use case

- 1:many (Tier 3): everyone else, reached with scaled programs

Engagement needs one rule change for ABM. Roll every lead up to its account, and never discard a second or third lead from the same company as a duplicate. Several people from one account engaging is the strongest signal you'll get that a buying group is forming. Look at who they are and what they're reading, as well as how many of them there are.

Every business's lead scoring model will look different, depending on your market, deal size, and sales cycle. The process for building one is the same.

1. Pull your closed-won and closed-lost data. Export 12 to 24 months of deals from your CRM and note the firmographics and behaviors behind each one. This is the evidence your points should come from. Clean the data first, because duplicate accounts and missing revenue fields will skew every score that follows.

2. Agree definitions with sales. Sit down with Sales and Customer Success, the people closest to your customers, and agree what an MQL and an SQL mean, plus how fast sales should follow up. Write it down. A lead scoring model only works when both teams trust the number.

3. Build your fit score. Choose four or five firmographic and demographic attributes, score each from 0 to 10, and weight them using the formula above. Check the result against Step 1, because your recent wins should score highly.

4. Map engagement to buying stages. List every trackable action, from blog reads to demo requests, and give each one points based on its stage. Map your customer journey first if you haven't already, so you know which content sits where.

5. Add negative scoring and decay. Set deductions for competitors, job seekers, unsubscribes, and poor-fit companies. Add a decay rule so behavioral points fade after a set period without activity.

6. Set thresholds and routing. Decide the fit and engagement scores that trigger an MQL, and what happens in each box of the fit and engagement grid. Make sure every route has an owner.

7. Build it in your CRM. HubSpot, Salesforce, and Marketo all come with lead scoring tools you can customize. Turtl's integrations with HubSpot and Marketo send content engagement data, like read time, straight into your scoring properties.

8. Review it every quarter. Compare MQL-to-SQL conversion, sales rejection reasons, and win rate by score band. If leads that cleared the threshold keep getting rejected, something is weighted too heavily. That's exactly how we found a fix in our own model.

For years, we ran a behavioral and engagement lead scoring model in HubSpot. It leaned on how people engaged with our content, because read time and depth told sales how familiar a prospect already was with Turtl, and what they didn't need explaining twice.

What happened when we adjusted our lead scoring model

A batch of inbound MQLs had passed the lead score threshold, then been rejected by sales because the timing was wrong. When we reviewed them, read time was the attribute weighted too heavily.

We'd been awarding +4 for reading a Turtl Doc for 60 seconds or more, and nothing below that. So we split the bracket into smaller steps.

A full minute of reading still earns +4, because it's a metric we believe in. Readers who stop at 30 seconds now get scored for the attention they gave, too. The result was more leads scored, and fewer not-yet-ready MQLs passed to Sales.

The lesson travels. When sales rejects leads that cleared your threshold, look for a single attribute doing too much of the work, and break it into smaller steps before you rebuild anything.

Where we've gone since

We've since moved to an intent scoring model. Lead scoring taught us which content signals mattered, and that experience is what made the switch possible. The bigger shift was in what we score. Our intent model puts more weight on signals that show a buying group forming, and less on one person crossing a points line. The next section explains why.

Lead scoring is a reliable way to prioritize the people who raise their hand. It can't see the buying group behind them, and that's where most B2B decisions get made.

The timing works against the MQL. According to 6sense's 2025 B2B Buyer Experience Report, buyers first contact sellers 61% of the way through their journey, buy from their Day One shortlist 95% of the time, and initiate 79% of engagements themselves. Marketo co-founder Jon Miller puts the consequence bluntly: "If you only wait for somebody to raise their hand, you are by definition only talking to the 5% in market."

So by the time a lead crosses your threshold, the group has often formed a favorite. And a lead-based process can punish you for reaching that group. On Turtl's podcast Real Seekers After Truth, Kerry Cunningham described teams binning the second and third leads from an account because a BDR couldn't earn credit for qualifying them. One account converts, the report shows 25% lead conversion, and the campaign that reached the whole group looks average.

"The better you do your job, the worse the stats are going to look." Kerry Cunningham, 6sense

Swapping lead scores for account scores won't fix this on its own. A "hot account" flag is still one number standing in for a decision several people are making. The extra value comes from reading the group itself.

Where Turtl fits

Keep your lead scoring model for triage. Turtl adds the buying group view beside it, so you can see how a decision is forming before anyone fills in a form.

• See accounts before they raise a hand. Account Reveal identifies companies from anonymous website visits, so you can spot activity in target accounts early.
• Read what each person cares about. Turtl's content analytics track engagement down to the section, so you can see whether a reader spent their time on ROI, security, or implementation, and what the group is debating.
• Give every function its own content. With 6sense or Demandbase buying stage and persona data, alongside Dynamic Personalization, Turtl serves a business case to the CFO and an implementation guide to the practitioner from the same campaign.
• Feed the model you already run. Engagement flows into HubSpot and Marketo, so your scoring keeps working, with a far richer picture of each account sitting alongside it.

A lead scoring model gives sales a shorter, sharper queue. Score revenue brackets on your own win data, require fit and engagement together, and review the weights every quarter. Then go further than the MQL. Your buyers decide as a group, mostly before they talk to you, and the teams that win are the ones reading that group while it forms.

Get your copy of The Intent Activation Blueprint and learn how to turn intent signals into buying group engagement

What's a good lead score threshold for an MQL?

A good MQL threshold is the score most of your past SQLs reached and most of your rejected leads didn't. Set it from your own conversion data rather than a benchmark. On a 100-point scale, many B2B teams start around 60 for fit and 50 for engagement, then adjust quarterly. If sales rejects lots of MQLs, raise it. If good deals arrive without ever crossing it, lower it.

What's the difference between an MQL and an SQL?

An MQL is a lead marketing judges ready for sales based on its score, and an SQL is a lead sales has accepted and confirmed as a real opportunity. The gap between them is your best health check. A low MQL-to-SQL conversion rate usually means your scoring rewards activity that doesn't signal buying intent.

Is predictive lead scoring better than rule-based scoring?

Predictive lead scoring beats rule-based scoring once you have enough clean data, usually a few hundred closed deals. Below that, the model doesn't have enough patterns to learn from and can produce confident-looking scores that mean little. Rule-based scoring is also easier for sales to understand and trust, so many teams start there and add predictive scoring later.

What's the difference between lead scoring and intent data?

Lead scoring ranks known individuals using data you collect, while intent data shows which accounts are researching a topic, often before anyone from them visits your site or fills in a form. They work best together. Intent data flags the accounts to watch, and lead scoring prioritizes the people from those accounts who engage with you directly.

How do you know if your lead scoring model is working?

Your model is working when higher score bands convert to SQLs and closed deals at clearly higher rates than lower bands. Check MQL-to-SQL conversion, sales rejection reasons, and win rate by score band every quarter. If a 90-point lead is no more likely to close than a 60-point lead, your weights need rethinking.