Disclosure: MarTechMeter may earn a commission if you buy through some links on this page, at no extra cost to you. This does not change our recommendations. Learn more.
Quick answer: Lead scoring is a way to rank prospects by giving each one points for who they are (fit) and what they do (engagement), such as job title, company size, pricing-page visits or demo requests. Once a lead passes a set threshold, it goes to sales. It helps small teams call the best leads first.
Key takeaways
- Lead scoring turns “this lead feels hot” into a number your marketing and sales teams agree on, so reps stop guessing who to call first.
- Good models mix two things: fit (does this person match your ideal customer?) and engagement (are they showing buying intent right now?).
- Rule-based scoring, where you set the points yourself, works for almost any small business. Predictive or AI scoring needs a lot of past conversion data before it becomes reliable.
- Most CRMs gate scoring behind mid-tier plans: HubSpot Professional, Pipedrive Premium (from $59 a seat a month billed annually) and Zoho CRM Standard and up, as listed in November 2026.
- Start with five to eight rules, one threshold and a monthly review. Complex 40-rule models usually break before they help.
Terms used in this guide
- Lead score: A number attached to a contact, company or deal that estimates how likely it is to buy. Higher means sales should look sooner.
- Fit score: Points based on who the lead is: industry, company size, job title, location or budget.
- Engagement score: Points based on behavior: email clicks, page views, webinar sign-ups, demo requests, replies.
- MQL and SQL: A marketing-qualified lead has crossed your score threshold. A sales-qualified lead has been checked by a rep and accepted as a real opportunity.
- Score decay: Points that expire or shrink over time, so a lead who was active a year ago does not still look hot today.
What is lead scoring, in plain English?
Lead scoring is a points system for your prospects. Every lead in your CRM gets a number, and the number goes up or down based on two kinds of evidence: how closely the person matches the customers you actually win, and how much buying intent they are showing right now. When a lead crosses a threshold you choose, it moves from marketing to a salesperson.
The point is prioritization. A small sales team cannot call every one of the 300 people who downloaded a checklist this month. Scoring tells them which 20 to call first. Done well, it also tells marketing which leads need more nurturing before anyone picks up the phone.
We have reviewed the scoring tools in most of the CRMs we cover, including HubSpot Sales Hub and Zoho CRM. Our angle in this guide is practical: rather than theory, we build a complete starter model for a realistic small business, score four sample leads with it and show what you need to pay for to run it in common CRMs.
How does lead scoring work?
Almost every scoring system, from a spreadsheet to Salesforce, follows the same four steps:
- Define your ideal customer. Look at your last 20 to 50 closed deals. What do the best ones have in common? Industry, size, job title, region and budget are the usual patterns.
- List the signals. Split them into fit signals (facts about the person or company) and engagement signals (things they do, such as visiting pricing or booking a call).
- Assign points. Strong signals get more points. Signals that suggest a poor fit, such as a competitor’s domain or an unsubscribe, subtract points.
- Set a threshold and a handoff. Decide the score at which a lead becomes a marketing-qualified lead (MQL) and what happens next: an alert to a rep, a task or a deal created automatically.
The CRM then recalculates scores as data changes. Pipedrive’s Scores help article notes, for example, that its scores “are recalculated automatically when deal data changes,” and ActiveCampaign’s contact scoring guide lets points expire after a set period, which keeps old activity from inflating a score.
What is the difference between fit and engagement scoring?
This is the distinction that matters most, and the one beginners skip. A lead can score high for the wrong reason if you only measure one side.
- Fit answers “should we sell to this person at all?” A student who reads every blog post is highly engaged and a terrible fit.
- Engagement answers “are they interested right now?” A perfect-fit VP who has not opened an email in a year is a great fit with no current intent.
HubSpot’s lead scoring documentation reflects this split directly: HubSpot lets you build separate engagement scores, fit scores or combined scores for contacts and companies. We recommend tracking both and only passing leads to sales when both are reasonably high. If your CRM only allows one score, weight fit a little more than engagement, because behavior is easier to fake by accident (bots, curious competitors, job seekers).
What does a lead scoring model look like? A worked example
To make this concrete, we built a starter model for a fictional but typical business: a six-person US marketing agency selling retainers of around $2,500 a month to small and mid-size companies. We based the rules on the pattern we see in most service businesses, where the pricing page and the booking form are the two pages that matter most.
| Rule | Type | Points | Why we set it this way |
|---|---|---|---|
| Company has 10 to 200 employees | Fit | +15 | The sweet spot for a monthly retainer; smaller firms rarely have budget |
| Job title includes owner, founder, director or head of marketing | Fit | +20 | These people sign contracts; interns and students do not |
| Based in the US or Canada | Fit | +10 | The agency only serves North American clients |
| Personal email domain (gmail, yahoo and so on) | Fit | -10 | Often students or job seekers, though not always |
| Visited the pricing page | Engagement | +15 | The strongest buying signal on most service sites |
| Visited the pricing page 3+ times in 14 days | Engagement | +10 | Repeat visits usually mean they are comparing options |
| Downloaded a guide or checklist | Engagement | +5 | Interest, but early-stage |
| Booked a call or requested a quote | Engagement | +40 | Basically a hand-raise; this alone should push most fits over the line |
| Unsubscribed from email | Engagement | -25 | A clear sign of lost interest |
| No activity for 60 days | Decay | -15 | Keeps old leads from looking warm forever |
We set the MQL threshold at 60 points. That number is not magic. We chose it so that a decent-fit lead needs at least one real buying signal (pricing or a booked call) to qualify, while a perfect-fit lead with no activity does not.
Notice what is missing: email opens. Apple Mail Privacy Protection and similar features open emails automatically, so open-based points inflate scores with no human behind them. Clicks and replies are far more trustworthy.
How do the sample leads score?
Here is how four realistic leads land under that model:
| Lead | What we know | Score | Result |
|---|---|---|---|
| A: Marketing director, 40-person HVAC company in Ohio | Visited pricing twice, downloaded a checklist | 20 + 15 + 10 + 15 + 5 = 65 | MQL: passed to sales |
| B: Student with a Gmail address | Downloaded three guides, never saw pricing | -10 + 5 = -5 (guides scored once) | Stays in the newsletter |
| C: Founder, 8-person startup in Texas | Booked a call on day one | 20 + 10 + 40 = 70 | MQL: called within the hour |
| D: Owner, 120-person firm in London | Visited pricing 4 times, then went quiet for 60 days | 15 + 20 + 15 + 10 – 15 = 45 | Below threshold; back to nurture |
Lead B is the classic case scoring is designed to catch. Without a model, a busy agency might call the person who downloaded three guides first. With it, the founder who booked a call (Lead C) and the director who keeps checking pricing (Lead A) jump the queue. Lead D shows why decay matters. A strong-fit lead that went quiet is worth a nurture email, not a sales call. If they come back to the pricing page, the points return and they cross the line again.
Run this kind of test on 20 of your own recent leads before switching anything on. If your best customers would not have scored above the threshold, change the rules first.
What is predictive or AI lead scoring?
Predictive scoring flips the process around. Instead of you choosing the points, software studies leads that converted and leads that did not, then builds the model itself. Salesforce describes Einstein Lead Scoring this way, and HubSpot offers AI-built contact scores on Marketing Hub Enterprise.
The catch is data. Salesforce’s Einstein Lead Scoring requirements says Einstein uses a global model until you have at least 1,000 leads created and at least 120 converted in the last six months. HubSpot’s AI option needs a sample of at least 50 contacts, 25 converted and 25 not converted, before it can suggest criteria. Many small businesses close a handful of deals a month, which is not enough history for a model to find reliable patterns.
Our view: start rule-based. You will learn what actually predicts a sale in your business, and you can always layer AI on top once you have the volume. Predictive scores are most useful for teams with thousands of inbound leads a month and a sales team that cannot possibly touch them all.
Which CRMs include lead scoring, and what does it cost?
Lead scoring is rarely on a free or entry plan. Here is what the tools we cover require, based on each vendor’s own documentation in November 2026:
| Tool | Scoring type | Plan you need (Nov 2026) | Notable limits |
|---|---|---|---|
| HubSpot | Rule-based contact, company and deal scores; AI contact scores | Marketing Hub or Sales Hub Professional or Enterprise; AI scoring needs Marketing Hub Enterprise | Contact scores need Marketing Hub; deal scores need Sales Hub. AI needs at least 50 sample contacts (25 converted) |
| Pipedrive | Rule-based deal scores | Premium (US$59 a seat a month billed annually) or Ultimate (US$79) | Deals only; one active score per pipeline; 5 custom scores on Premium, 10 on Ultimate |
| Zoho CRM | Rule-based scoring rules, including email and signal rules | Standard ($14 a user a month billed yearly) and up | One rule per module on Standard and Professional; multiple rules from Enterprise ($40) |
| Salesforce | Predictive Einstein Lead Scoring | Sales Cloud Einstein: included in Performance and Unlimited, add-on for Enterprise | Uses a global model until you have 1,000+ leads and 120+ conversions in six months |
| ActiveCampaign | Rule-based contact scores; deal scores | Contact scoring on marketing plans; deal scoring with the CRM (Pipelines) add-on | Scores can expire automatically, which handles decay for you |
A few practical notes. HubSpot’s costs climb fast: Sales Hub Professional lists at $100 a seat a month ($90 billed annually) and Marketing Hub Professional at $890 a month, or $800 a month billed annually, for three core seats, before onboarding fees. Zoho is the budget pick, since Zoho CRM’s scoring rules documentation shows scoring rules from the Standard edition, with Enterprise needed for multiple parallel rules. Pipedrive scores deals rather than raw leads, which suits teams that turn inquiries into deals quickly. Our Pipedrive pricing breakdown covers what else Premium adds, and our ActiveCampaign pricing guide explains which add-ons bring its CRM features.
Lead scoring pros and cons
Pros
- Reps spend time on leads that match your customer profile instead of whoever filled in a form last
- Marketing and sales share one definition of a “qualified” lead, which ends a lot of arguments
- Hot leads can trigger instant alerts, so speed-to-lead improves
- Scores feed automation: nurture low scorers, route high scorers, clean out dead ones
- Easy to start: a handful of rules in a CRM you already pay for
Cons
- Bad rules are worse than no rules, because they send reps chasing the wrong people with confidence
- Scoring is often locked behind pricier CRM tiers
- Engagement data depends on tracking, which privacy settings and blocked cookies weaken
- Models drift as your product, pricing and market change, so they need regular review
- Predictive scoring is a black box for many teams and needs volume most small businesses do not have
What are the most common lead scoring mistakes?
- Building it without sales. If reps do not agree with the threshold, they ignore the score. Agree on the definition of an MQL in one meeting before you build anything.
- Too many rules. A 40-rule model is impossible to debug. Start with five to eight and add more only when you can explain why.
- Scoring email opens. Automated opens make this signal unreliable. Score clicks, replies and form fills instead.
- No negative points. Competitors, students, job seekers and unsubscribes should pull scores down.
- No decay. Without expiring points, your database slowly fills with “hot” leads who left months ago.
- Never reviewing it. Check monthly: how many MQLs became sales-qualified, and how many deals closed from each score band? If high scorers are not closing, adjust the rules.
How did we evaluate lead scoring tools?
When we test a CRM’s scoring feature, we look at five things that matter more than the marketing page suggests:
- Fit and engagement together: can you score both, or only one? HubSpot and Zoho handle both; Pipedrive focuses on deal fit.
- Negative points and decay: can scores go down, and can points expire? ActiveCampaign’s expiring points are the cleanest version we have seen.
- Transparency: can a rep see why a lead scored 70? Pipedrive shows the contributing factors on hover, which builds trust in the number.
- Automation: can a threshold trigger alerts, tasks or deals without extra tools?
- Plan cost: how far up the pricing ladder do you have to climb to get it?
If you are still picking a CRM, our guide to the best CRM software for small business weighs these alongside the rest of each platform.
Who should use lead scoring?
- Good fit: B2B companies and service businesses that get more inbound leads than their team can follow up on quickly.
- Good fit: teams with a separate marketing and sales function that need a shared definition of a qualified lead.
- Good fit: businesses running outbound alongside inbound, where scoring helps decide which replies get a call. Prospecting tools like Apollo.io add data that makes fit scoring more accurate.
- Poor fit: very low-volume businesses that get five leads a week. Just call them all.
- Poor fit: pure ecommerce stores, where purchase behavior and segmentation do the job better than a sales-style score.
For teams that already score leads and want to act on them faster, our roundup of sales engagement software covers the sequencing tools that pick up where scoring leaves off.
The bottom line
Lead scoring is a simple idea: give each prospect points for fitting your ideal customer and for showing buying intent, then send the high scorers to sales first. The best small-business models are short, mix fit with engagement, subtract points for bad signals and let old activity expire. Start with a rule-based model in the CRM you already use, test it against your recent deals and review it every month. Save predictive AI scoring for when you have the lead volume to feed it.
Frequently asked questions
What is lead scoring in simple terms?
Lead scoring is a points system that ranks your prospects. Each lead earns points for matching your ideal customer and for actions such as visiting your pricing page or booking a call. Leads above a set score go to sales first.
What is a good lead score?
There is no universal good score, because every business sets its own scale. A common approach is a 0 to 100 range with a sales threshold around 50 to 70, tuned so that your recent best customers would have crossed it.
What is the difference between an MQL and an SQL?
A marketing-qualified lead (MQL) has reached your score threshold. A sales-qualified lead (SQL) has been reviewed by a salesperson and accepted as a real sales opportunity. Scoring decides MQLs; people decide SQLs.
What are the two main types of lead scoring?
Fit scoring rates who the lead is, such as industry, company size and job title. Engagement scoring rates what they do, such as clicks, page views and demo requests. The strongest models combine both.
Does HubSpot have lead scoring on the free plan?
No. HubSpot lead scoring requires Marketing Hub or Sales Hub Professional or Enterprise. Contact scores need Marketing Hub, deal scores need Sales Hub, and AI-built contact scores need Marketing Hub Enterprise.
Is predictive lead scoring better than rule-based scoring?
Only with enough data. Salesforce Einstein, for example, needs at least 1,000 leads and 120 conversions in six months for a model built on your own data. Most small businesses get better results from a simple rule-based model first.
How often should you update a lead scoring model?
Review it monthly and rebuild it whenever your pricing, product or target customer changes. Check whether high-scoring leads actually become customers, and adjust points for any signal that is not predicting sales.
Sources: HubSpot: Overview of the lead scoring tool; HubSpot: Build lead scores; Salesforce Einstein features datasheet; Zoho CRM: Scoring rules editions and limits; Pipedrive: Scores; Pipedrive pricing; ActiveCampaign: Contact scoring. Prices checked on official pages in November 2026. Vendors change pricing and promotions often, so confirm before you buy. This article is general information, not tax or legal advice.
