---
title: "Negative Lead Scoring: Examples, Point Values, Best Practices"
description: "Negative lead scoring subtracts points for bounces, personal emails, competitors, students and inactivity. Get the point table, thresholds and CRM setup."
canonical_url: "https://www.leadpipe.com/blog/negative-lead-scoring/"
source_url: "https://www.leadpipe.com/blog/negative-lead-scoring/"
md_url: "https://www.leadpipe.com/blog/negative-lead-scoring.md"
content_type: "article"
pubDate: "2026-09-24T00:00:00.000Z"
last_updated: "2026-09-24T00:00:00.000Z"
---

# Negative Lead Scoring: Examples, Point Values, Best Practices

In this guide

Negative lead scoring subtracts points for bounces, personal emails, competitors, students and inactivity. Get the point table, thresholds and CRM setup.

Negative lead scoring is the practice of **subtracting points** from a lead when a signal suggests they are a poor fit, unreachable, or not actually buying. A hard email bounce, a competitor’s domain, a job seeker reading your careers page, or 90 days of silence should all push a lead down the queue, even if that same lead downloaded three ebooks.

Most scoring models are built only with positive points. They reward activity, so the leads who click the most float to the top: students writing papers, competitors doing research, and job candidates. Negative scoring is the counterweight that keeps your MQL queue honest.

This guide gives you a concrete point table, a worked example model with thresholds, the right way to handle personal emails like Gmail on a B2B form, and how to set it up in HubSpot, Salesforce and Pipedrive.

> **In one sentence:** Positive scoring tells you who is active; negative scoring tells you who to ignore, so sales spends its limited time on leads who can actually buy.

## Why negative lead scoring matters

Sales time is scarce. The [Salesforce State of Sales report (7th edition)](https://www.salesforce.com/news/stories/state-of-sales-report-announcement-2026/) found the average seller spends only **40% of their time actually selling**, based on a survey of 4,050 sales professionals. Every junk lead routed to a rep eats into that 40%.

The funnel math is also unforgiving. [Forrester](https://www.forrester.com/blogs/saying-goodbye-to-mqls-sweet-and-no-sorrow/) reports that **fewer than 1% of leads convert to closed deals** in a traditional lead-centric process. When the base rate is that low, filtering out the obvious non-buyers is one of the cheapest ways to raise MQL quality.

Data decay adds to the problem. ZeroBounce’s analysis of more than 11 billion addresses found that [nearly 23% of email addresses become invalid or risky within a year](https://www.prnewswire.com/news-releases/nearly-a-quarter-of-email-lists-decay-each-year-new-report-finds-302697945.html). And Gartner estimates that poor data quality costs organizations [at least $12.9 million a year on average](https://www.gartner.com/en/data-analytics/topics/data-quality). A scoring model that never subtracts points for bad data will slowly fill your CRM with records nobody can reach.

The payoff is measurable: in the same Salesforce research, **79% of high-performing sales teams prioritize data hygiene, versus 54% of underperformers**.

## Negative lead scoring point table

The table below is a starting template for a 0 to 100 point model. The values are illustrative: calibrate them against your own closed-won and closed-lost data.

| Signal | Type | Suggested points | Disqualify or subtract? |
| --- | --- | --- | --- |
| Hard email bounce (invalid mailbox or domain) | Data quality | −50 | Disqualify from email, re-verify contact |
| Soft bounce, 3+ in a row | Data quality | −10 | Subtract, retry later |
| Disposable or temporary email domain | Data quality | −40 | Usually disqualify |
| Spam-trap patterns (gibberish local part, keyboard mash, fake names like “test test”) | Data quality | −50 | Disqualify and suppress |
| Personal or free email domain (gmail.com, yahoo.com, outlook.com) on a B2B form | Fit (uncertain) | −5 to −15 | Subtract, then resolve identity (see below) |
| Role account (info@, sales@, admin@, support@) | Fit | −10 | Subtract |
| Competitor email domain | Fit | −100 | Disqualify |
| Student or academic title, .edu domain (if you do not sell to education) | Fit | −30 | Subtract or disqualify |
| Job seeker: careers page visits, “job” or “career” form reasons | Behavior | −15 per visit, cap −40 | Subtract |
| Unsubscribed from marketing email | Behavior | −20 | Subtract (never email again) |
| Marked email as spam | Behavior | −50 | Suppress entirely |
| No engagement for 30 / 60 / 90 days | Decay | −5 / −10 / −20 | Subtract |
| Non-ICP country (outside your served markets) | Fit | −25 | Subtract or route to partner |
| Non-ICP industry | Fit | −20 | Subtract |
| Company size far outside ICP band | Fit | −15 to −30 | Subtract |
| Existing customer (already won) | Routing | Remove from MQL scoring | Route to CS or account owner |
| Only viewed the support, login, or documentation pages | Behavior | −5 | Subtract |
| Bot-like behavior (dozens of pages in seconds, no scroll) | Data quality | −100 | Exclude from scoring |

Three rules keep the table from backfiring:

1. **Cap behavioral negatives.** A prospect who reads your careers page once while evaluating your company is normal. Cap repeat penalties so one signal cannot outweigh strong buying intent.
2. **Separate “unreachable” from “unfit.”** A hard bounce means you cannot email the person, not that the company is a bad fit. Keep the account alive and look for a valid contact.
3. **Never let negatives silently delete leads.** Negative scores should route and suppress, not destroy data you might need for attribution later.

## When to disqualify vs. subtract points

Not every bad signal deserves a point value. Some should flip a lead out of the scoring model entirely.

**Disqualify (set a status, stop scoring) when the signal is binary and reliable:**

- The email domain belongs to a competitor.
- The contact marked your email as spam or asked not to be contacted.
- The record is clearly fake, a spam-trap pattern, or bot traffic.
- The lead is an existing customer (route them, do not score them as new business).

**Subtract points when the signal is probabilistic:**

- Personal email domains, role accounts, careers-page visits, inactivity, and borderline firmographics all correlate with lower conversion but have real exceptions.

A practical way to implement this is a **“disqualified” checkbox plus a reason picklist** that sits alongside the score. Scoring then only runs for records where the checkbox is false. This keeps the numeric score meaningful and gives you clean reporting on why leads were excluded.

> **In one sentence:** Disqualify on certainty, subtract on probability.

## Email bounces, deliverability and negative scoring

Bounce handling is the most searched negative scoring topic for good reason: it protects your sending domain, not just your pipeline.

Google’s Gmail sender guidelines say senders should [keep their spam rate below 0.1% and never let it reach 0.3%](https://support.google.com/mail/answer/14229414?hl=en). Bounces and spam complaints feed into sender reputation, so a lead that bounced should not keep receiving nurture emails just because it still has a high engagement score from last year.

Best practices:

- **Hard bounce:** subtract heavily (−50), set email status to invalid, remove from all sends, and trigger a re-enrichment workflow to find a working address.
- **Soft bounce:** subtract lightly and only after repeated failures. Mailbox-full and temporary server errors often resolve on their own.
- **Catch-all domains:** do not penalize by default, but verify before bulk sending.
- **Spam complaint:** suppress permanently. No score adjustment brings this lead back into automated email.

## How to score personal emails (Gmail, Yahoo, Outlook) in B2B

This is where most negative scoring models make their most expensive mistake. They assign −20 or −50 to every gmail.com address and move on.

The problem: plenty of real buyers use personal email. Founders and consultants often do. People researching a tool before involving their team often do. Visitors on mobile often autofill their personal address. Auto-discarding them throws away leads who may be decision-makers.

A better approach has three steps:

1. **Apply a small penalty, not a disqualification.** −5 to −15 reflects the lower average quality without burying the lead.
2. **Resolve the person to their work identity.** Enrichment can often connect a personal email to a name, current employer, and job title. Once you know the Gmail user is a VP of Operations at a 300-person company in your ICP, the penalty should be reversed and fit points added.
3. **Score the resolved record, not the raw form fill.** Fit scoring should use the enriched company and title, not the email domain.

This is where person-level visitor identification helps. [Leadpipe](/product/identification/) matches website visitors to real people through deterministic matching against its own identity graph, returning professional details such as name, company and title for identified visitors. For non-US visitors, [IP-to-company identification](/product/ip-to-company/) adds the account-level picture. Instead of guessing whether “[jsmith1984@gmail.com](mailto:jsmith1984@gmail.com)” is a student or a buyer, you can see who they are and which pages they read.

[Try Leadpipe free with 500 leads →](https://dashboard.leadpipe.com/auth/signup)

## Negative signals from website behavior

Website behavior is a rich source of negative signals, especially once visitors are identified. Useful ones:

| Behavior | What it usually means | Suggested treatment |
| --- | --- | --- |
| Careers or jobs pages only | Job seeker | −15 per visit, capped |
| Login, support, or help center only | Existing user, not a new buyer | Route to CS, remove from new-business scoring |
| Blog-only visits over many weeks, never product pages | Researcher, student, or content consumer | Small negative or no positive points |
| Single bounce visit under 10 seconds | Accidental or low interest | 0 points (do not reward) |
| Investor relations or press pages only | Analyst, journalist, investor | Exclude from sales scoring |
| Visits from a competitor’s company (via IP-to-company or identification) | Competitive research | Disqualify from sales, flag for competitive intel |
| Abnormal speed or page counts | Bot or scraper | Exclude entirely |

The key insight: **absence of the right behavior** is also a signal. A lead who has read 20 blog posts but never opened pricing, product, or integration pages is informed, not in-market. See [lead scoring with visit behavior and intent data](/blog/lead-scoring-with-visitor-behavior-and-intent/) for the positive side of the same model, and [the dirty number on B2B bot traffic](/blog/b2b-bot-traffic-the-dirty-number/) for why bot filtering belongs at the top of the scoring pipeline.

## Score decay: the negative score that runs on a timer

Score decay reduces points over time when a lead stops engaging. Without it, a lead who visited your pricing page 11 months ago still looks hot today.

Two common approaches:

- **Event-level decay:** each scored event loses value as it ages. HubSpot’s lead scoring tool supports this directly: its [knowledge base](https://knowledge.hubspot.com/scoring/understand-the-lead-scoring-tool) describes score decay that “automatically reduces an individual event’s score based on how long ago a scored event occurred,” with decay intervals of 1, 3, 6, or 12 months.
- **Inactivity penalties:** a scheduled workflow subtracts points after 30, 60, and 90 days without a meaningful activity.

Decay should apply to **engagement** points only. A company’s industry or size does not become less relevant because they went quiet. Keeping fit and engagement as separate scores makes this easy.

## Example negative scoring model with thresholds

Here is a complete, hypothetical model for a B2B SaaS company selling to mid-market companies in North America and Europe. Fit and engagement are scored separately and then combined.

**Fit score (0 to 50)**

| Attribute | Points |
| --- | --- |
| Company size 50 to 1,000 employees | +20 |
| Target industry (software, professional services, financial services) | +15 |
| Director, VP, or C-level title in sales, marketing, or RevOps | +15 |
| Non-ICP industry | −20 |
| Outside served countries | −25 |
| Student or academic title | −30 |
| Role account | −10 |
| Personal email, unresolved | −10 |

**Engagement score (0 to 50)**

| Activity | Points |
| --- | --- |
| Pricing page visit (last 14 days) | +15 |
| Demo request | +30 |
| Integration or product page visits (3+) | +10 |
| Webinar attended | +5 |
| Careers page visit | −15 (cap −40) |
| Unsubscribe | −20 |
| No activity 30 / 60 / 90 days | −5 / −10 / −20 |

**Thresholds and actions**

| Combined score | Status | Action |
| --- | --- | --- |
| 70 and above | MQL | Route to sales within your follow-up SLA |
| 40 to 69 | Nurture, high priority | Targeted nurture, SDR research |
| 0 to 39 | Nurture | Standard marketing nurture |
| Below 0 | Recycle | Suppress from sales views, keep for reporting |
| Disqualified flag = true | Excluded | No scoring, reason captured |

Walk through two leads:

- **Lead A:** Director of RevOps at a 400-person software company, visited pricing twice, unsubscribed from the newsletter. Fit 50, engagement 15 + 10 − 20 = 5. Combined 55 → high-priority nurture, not an MQL yet. That is correct: they are a great fit who does not want more newsletters.
- **Lead B:** gmail.com form fill, 12 blog visits, 3 careers-page visits. Fit −10, engagement −40 (capped). Combined −50 → recycle. If enrichment later resolves the Gmail address to a VP at an ICP company, fit jumps to +50 and the lead re-enters the queue.

## How to implement negative lead scoring in your CRM

Keep this generic: menus change often, so check your vendor’s current documentation.

### HubSpot

HubSpot’s lead scoring tool lets you [add or subtract points](https://knowledge.hubspot.com/scoring/understand-the-lead-scoring-tool) and create separate fit, engagement, or combined scores, with score limits, decay, and High / Medium / Low thresholds. It is available on Marketing Hub and Sales Hub Professional and Enterprise. Build your negative criteria as property-based fit rules (email domain, industry, country) and event-based engagement rules (unsubscribe, careers-page views). For identified visitor data, see [the Leadpipe HubSpot integration](/blog/leadpipe-hubspot-integration/).

### Salesforce

Salesforce Account Engagement (formerly Pardot) separates **scoring** (engagement) from **grading** (fit against your ideal profile, expressed as a letter grade). Negative fit signals belong in [grading profiles](https://help.salesforce.com/s/articleView?id=sf.pardot_leadqual_grading.htm&language=en_US&type=5), while negative behavioral signals belong in automation rules or completion actions that adjust score. If you score in core Salesforce without Account Engagement, a formula field or a Flow that sums fit and engagement components works well. See [the Leadpipe Salesforce integration](/blog/leadpipe-salesforce-integration/).

### Pipedrive

Pipedrive’s native [Scores feature](https://support.pipedrive.com/en/article/scores) currently applies to **deals**, with criteria that are highly positive (+25), slightly positive (+10), or negative (−10), and is available on Premium plans and above. For lead-level scoring, most teams use a custom numeric field updated by automations or an integration. Pushing identified visitors into Pipedrive with company and title attached makes those rules far more accurate: see [the Leadpipe Pipedrive integration](/blog/leadpipe-pipedrive-integration/).

## KPIs to measure your negative scoring model

Track these monthly:

| KPI | How to calculate | What good looks like |
| --- | --- | --- |
| MQL → SQL conversion rate | SQLs ÷ MQLs in period | Rising after negatives go live |
| Lead quality score | Average combined score of leads passed to sales | Stable or rising |
| Sales acceptance rate | Accepted MQLs ÷ MQLs routed | Rising, fewer “junk” rejections |
| Disqualification reasons | Count by reason picklist | Informs targeting and form fixes |
| False-negative rate | Closed-won deals whose lead once scored below 0 | As close to zero as possible |
| Bounce rate on nurture sends | Bounces ÷ sends | Falling |

The false-negative rate matters most. If deals keep closing from leads your model buried, a penalty is too harsh (personal email penalties are the usual culprit).

## FAQ

### What is negative lead scoring?

Negative lead scoring subtracts points from a lead’s score when they show signals of poor fit, unreachability, or low buying intent, such as a hard email bounce, a competitor domain, a student title, careers-page visits, an unsubscribe, or long inactivity. It balances positive scoring so active but unqualified leads do not reach sales.

### What are examples of negative lead scoring criteria?

Common examples are hard bounces (−50), disposable email domains (−40), competitor domains (disqualify), student titles (−30), non-ICP countries (−25), unsubscribes (−20), careers-page visits (−15), role accounts like info@ (−10), personal email domains (−5 to −15), and inactivity decay of −5, −10, and −20 at 30, 60, and 90 days.

### Should I give negative points to Gmail or personal email addresses in B2B?

Give a small penalty, not a disqualification. Many real buyers use personal email. Enrich the record to find the person’s employer and title, then score the resolved profile. Person-level identification tools like Leadpipe can show who a visitor is and which pages they read, which settles the question with data instead of a blanket rule.

### How many points should a hard bounce subtract?

A hard bounce should subtract enough to drop the lead below your MQL threshold, typically −50 in a 100-point model, and set the email status to invalid. Remove the address from all sends to protect your sender reputation, then try to find a valid contact at the same company.

### What is score decay in lead scoring?

Score decay automatically reduces engagement points as activities age, so a pricing page visit from last year counts less than one from last week. Apply decay to engagement only, not to fit attributes like industry or company size.

### How do I score website visitors who never fill out a form?

Identify them first. Visitor identification turns anonymous sessions into named people or companies, so you can apply the same fit and behavior rules, including negative ones like careers-page-only visits. See [how to identify anonymous website visitors](/blog/how-to-easily-identify-anonymous-website-visitors/).

### What KPIs show whether lead scoring is working?

Watch MQL to SQL conversion rate, sales acceptance rate, average lead quality score, and the false-negative rate (closed-won deals that once scored below zero). Rising conversion with a near-zero false-negative rate means your negative scoring is filtering noise, not buyers.

## Related Articles

- [What Is Lead Scoring? Models and Best Practices](/blog/glossary-lead-scoring/)
- [Lead Scoring with Visit Behavior and Intent Data](/blog/lead-scoring-with-visitor-behavior-and-intent/)
- [ICP Fit Score: How to Qualify Leads Against Your ICP](/blog/icp-fit-score/)
- [MQL vs SQL](/blog/glossary-mql-sql/)
- [We Stopped Scoring Leads and Started Scoring Moments](/blog/stopped-scoring-leads-started-scoring-moments/)
- [Marketing Ops Visitor Data Hygiene](/blog/marketing-ops-visitor-data-hygiene/)

**Stop letting junk leads crowd out real buyers. Identify who is actually on your site and score them on real data: [start free with 500 identified leads, no credit card](https://dashboard.leadpipe.com/auth/signup).**