Lead Scoring

Lead scoring is the practice of assigning numeric values to leads based on how well they fit your ideal customer and how they engage with your marketing. The score ranks every contact in a database so sales works the highest-value leads first. When a score crosses a set threshold, the lead is handed to sales as a marketing qualified lead.

Why Lead Scoring Matters

Lead scoring solves a routing problem. Most databases contain far more leads than a sales team can call, and volume alone gives no signal about which ones are worth the call.

Without a score, reps work leads in the order they arrive. A student downloading a report for a class project gets the same follow-up as a VP of Operations who visited the pricing page three times.

Scoring also settles the argument between marketing and sales. When both teams agree what a 75 means, the handoff stops being subjective.

How Lead Scoring Works

Every scoring model combines two categories of data, then compares the total against a threshold.

1. Explicit data (fit). What the lead tells you: job title, company size, industry, country. This answers whether they match your ideal customer profile.
2. Implicit data (behavior). What the lead does: pages viewed, emails clicked, demos requested, webinars attended. This answers whether they are interested right now.
3. Point assignment. Each attribute and action carries a value. A director-level title might be worth 15 points, a pricing page visit 20, a whitepaper download 5.
4. Threshold. Once the total crosses the agreed number, the platform changes the lifecycle stage and alerts a rep.

Most teams use a 0 to 100 scale, though the ceiling is arbitrary. What matters is that the threshold reflects real conversion behavior, not a round number someone liked.

Lead Scoring Criteria

Scoring criteria fall into four groups. A model using only the first two will overrate stale, poor-fit contacts.

Demographic and firmographic criteria

• Job title and seniority (a VP scores higher than an intern)
• Company size and revenue band
• Industry, matched against your best-performing verticals
• Country or region, matched against where you sell

Behavioral criteria

• Pricing page or demo page visits (the strongest single signal in most B2B models)
• Repeat visits within a short window
• Email clicks, not just opens
• Bottom-funnel downloads such as comparison guides or ROI calculators

Negative criteria

• Free email domains (gmail.com, yahoo.com) on a B2B model
• Job titles containing “student”, “intern”, or “professor”
• Competitor domains
• Careers page visits, which usually signal a job seeker

Score decay
Interest expires. Most platforms support degradation rules that subtract points after inactivity, such as removing 10 points after 30 days with no site visit or email click. Without decay, a lead who was hot last year keeps a high score forever and clogs the sales queue.

Acquisition source belongs in the list too. A lead from a branded search campaign behaves differently from one captured by a giveaway. Scoring by channel only works if campaign links carry consistent UTM parameters your form passes into the CRM, which is what saved UTM templates enforce.

Types of Lead Scoring Models

• Rule-based scoring. A human defines every rule and point value. Transparent, easy to explain to sales, and the right starting point for most teams.
• Predictive scoring. Machine learning analyzes closed-won and closed-lost records to find the attributes that predicted revenue, then scores new leads against that pattern. Salesforce Einstein Lead Scoring and HubSpot’s predictive scoring work this way, and both need a volume of historical conversions to train on.
• Two-dimensional scoring (fit and engagement). Adobe Marketo Engage and Salesforce Marketing Cloud Account Engagement, formerly Pardot, separate a letter grade for fit from a number for behavior. A lead is an A1 or a D4, which distinguishes a perfect-fit lead who is quiet from a poor-fit lead who is very active.
• Account-based scoring. Points aggregate across every contact at the same company rather than per person. Used in ABM, where a buying committee each contributes signal.

How to Build a Lead Scoring Model

1. Analyze closed-won deals. Pull the last 6 to 12 months of customers and list the attributes and actions they shared before purchase. Repeat for closed-lost leads.
2. Score fit and behavior separately. Keep the two columns apart even if you sum them at the end. It makes the model debuggable.
3. Set point values on evidence. If pricing page visitors close at four times the rate of blog readers, the point spread should reflect that ratio.
4. Add negative and decay rules on day one. These are the rules teams skip, and their absence is what makes scores drift upward until they mean nothing.
5. Agree the threshold with sales in writing. Define who gets notified, how fast they respond, and what returns the lead to nurture.
6. Review quarterly. Compare close rates by score band. If leads scoring 40 close at the same rate as leads scoring 80, the point values need rebuilding.

Common Lead Scoring Mistakes

• Scoring email opens. Apple Mail Privacy Protection pre-fetches images, which fires open events for people who never read the message.
• Threshold set by capacity, not conversion. Lowering the bar because sales has spare time produces qualified leads that are not qualified.
• Never validating the model. A score that has never been checked against closed revenue is a guess with decimals.
• Untracked acquisition source. If campaign data never reaches the CRM, you cannot score by channel or prove which campaigns produce leads that convert.

Frequently Asked Questions

What is lead scoring in simple terms?

Lead scoring gives each lead a number showing how likely they are to buy. Points go up when a lead matches your ideal customer or takes a buying action, and down when they do the opposite. Sales calls the highest numbers first instead of guessing.

What is a lead scoring model?

A lead scoring model is the documented set of rules deciding how many points each attribute and action is worth, plus the threshold at which a lead goes to sales. Models are either rule-based, where a marketer sets every value manually, or predictive, where machine learning derives values from historical conversion data.

What are good lead scoring criteria?

The strongest criteria are the ones your own closed-won data supports. In most B2B models that means pricing and demo page visits, job title seniority, company size fit, and repeat visits in a short window. Pair those with negative points for free email domains, competitor domains, and career page visits.

What is the difference between lead scoring and lead grading?

Grading measures fit and scoring measures behavior. Grading produces a letter (A through D) based on who the lead is, while scoring produces a number based on what they did. Marketo and Account Engagement report both together, so an A1 is a perfect-fit lead who is highly engaged.

What score should trigger a sales handoff?

Set the threshold where your historical data shows conversion rates jump, not at a round number. Group past leads into score bands and find the band where close rates rise sharply. Then confirm the resulting lead volume matches what your sales team can work each week.

To score leads by acquisition channel, tag every campaign link consistently with the free UTM builder at linkutm.