Free tool · Lead generation Lead scoring model builder: test your rules on real leads
A lead score adds points for who someone is (title, company size, industry) and for what they did (pricing page, demo request, webinar), and lets the behavior points fade over time. Edit the rules, paste leads exported from your CRM or a spreadsheet, and see who crosses your MQL line before you change anything in the CRM.
- Fit and behavior rules
- Score decay by half-life
- Leads stay in your browser
Rules, decay and a test run
Each rule looks at one column of your lead list. Fit rules describe the person and company and do not decay. Behavior rules describe activity and fade with time since the last activity.
Scoring rules
| Type | Column | Condition | Value | Points | Cap | |
|---|---|---|---|---|---|---|
Separate alternatives with | (vp|director). Text comparisons ignore case. "Points per unit" multiplies a count column (page views, clicks) by the points, up to the cap; yes / true / 1 count as one. Negative points are allowed.
Decay and thresholds
After this many days without activity, behavior points are worth half. Leave blank for no decay.
A date (2026-11-03) or a number of days since the last activity.
Used when the activity column holds dates.
Leads at or above it are marketing-qualified.
Leads over the MQL line but under this fit score are flagged "Check fit" instead.
First row is the header. Commas, tabs (Excel, Google Sheets) and semicolons are detected. The first column is used as the lead's name. Up to 2,000 rows; nothing leaves your browser.
Scored leads
7 leads · 2 MQL · 1 to check for fit · 4 to nurture
| # | Lead | Fit | Behavior | Days since | Decay | Behavior after decay | Total | Status |
|---|---|---|---|---|---|---|---|---|
| 1 | Maya Chen | 40 | 66 | 3 | ×0.93 | 61.6 | 101.6 | MQL |
| 2 | Priya Shah | 15 | 53 | 1 | ×0.98 | 51.8 | 66.8 | Check fit |
| 3 | Grace Kim | 30 | 54 | 20 | ×0.63 | 34 | 64 | MQL |
| 4 | Tom Becker | 30 | 30 | 12 | ×0.76 | 22.7 | 52.7 | Nurture |
| 5 | Luis Ortega | 40 | 22 | 45 | ×0.35 | 7.8 | 47.8 | Nurture |
| 6 | Sam Lee | 40 | 0 | 90 | ×0.13 | 0 | 40 | Nurture |
| 7 | Anna Kowalski | -30 | 45 | 2 | ×0.95 | 43 | 13 | Nurture |
Model as Markdown (for your CRM admin or a team review)
# Lead scoring model
| Type | Column | Condition | Value | Points | Cap |
| --- | --- | --- | --- | --- | --- |
| Fit | title | contains any of | director\|vp\|head\|chief\|founder | +20 | |
| Fit | title | contains any of | manager | +10 | |
| Fit | company_size | ≥ | 50 | +10 | |
| Fit | company_size | ≥ | 500 | +10 | |
| Fit | industry | equals any of | software\|fintech\|ecommerce | +10 | |
| Fit | email | ends with any of | gmail.com\|yahoo.com\|hotmail.com\|outlook.com | -15 | |
| Fit | title | contains any of | student\|intern | -30 | |
| Behavior | demo_request | is filled | | +30 | |
| Behavior | pricing_views | points per unit | | +5 | 20 |
| Behavior | webinar | is filled | | +10 | |
| Behavior | email_clicks | points per unit | | +3 | 15 |
Behavior points halve every 30 days since the last activity (column: days_since_active). Fit points do not decay.
MQL threshold: total score ≥ 60.
Leads over the threshold with a fit score below 20 are reviewed before hand-off.
Version v1.0Updated Update history
What the score tells you
A lead score is a sorting rule, not a prediction. It decides who sales calls first and who stays in nurture.
Fit score
How close the person and company are to the customers you win. In the example, a VP at a 320-person software company gets 40; a student gets −30 and never reaches MQL however active.
Behavior score
Buying signals: a demo request (30), pricing page views (5 each, up to 20), a webinar, email clicks. These say someone is looking now.
Decay
Behavior loses value with time. With a 30-day half-life, 22 behavior points from 45 days ago are worth 7.8 today. Without decay, last year's webinar keeps an inactive lead at the top.
MQL and "Check fit"
Total ≥ 60 is an MQL. Priya Shah reaches 66.8 on behavior alone, but her fit score (15) is below the 20 minimum, so she is flagged for a human to look at instead of going straight to sales.
The points and thresholds in the example are a starting template, not a standard. Calibrate them on your own history: score last quarter's leads and check that the ones who became customers scored higher.
Four steps
Build the model here, check it on real leads, then set the same rules in your CRM or marketing automation tool.
-
1
Export leads with the columns you score on
Title, company size, industry, email, page views, form fills, last activity. A CSV export or a copied spreadsheet range both work.
-
2
Edit the rules
Fit rules for who they are, behavior rules for what they did. Use | for alternatives and negative points for disqualifiers.
-
3
Set decay and thresholds
A half-life for behavior, an MQL line and, if you want, a minimum fit score.
-
4
Check and export
Sort by score, check the top and the bottom against what you know, then download the scored CSV or copy the model as Markdown.
Score = fit + behavior × decay
Fit points are added for every fit rule that matches. Behavior points are added the same way and then multiplied by a decay factor based on days since the last activity.
- Decay factor = 0.5 ^ (days since last activity ÷ half-life). 30 days at a 30-day half-life is ×0.5; 60 days is ×0.25
- "Points per unit" = points × the number in the cell, capped at the cap (by absolute value)
- Status: total ≥ MQL threshold → MQL; if a minimum fit is set and fit is below it → Check fit; otherwise Nurture
- The split into fit and engagement scores with decay on engagement is common in marketing tools; HubSpot, for example, offers fit, engagement and combined scores and can decay scores by how long ago an event happened (HubSpot knowledge base)
Fit
Σ points of matching fit rules
VP (20) + 320 employees ≥ 50 (10) + software (10) → 40
Behavior
Σ points of matching behavior rules
demo 30 + 4 × 5 pricing views + webinar 10 + 2 × 3 clicks → 66
Total
fit + behavior × 0.5^(days ÷ half-life)
40 + 66 × 0.5^(3 ÷ 30) = 40 + 61.6 → 101.6
Seven leads, scored
The default list and rules, MQL at 60, minimum fit 20, half-life 30 days. Fictional leads.
| Lead | Fit | Behavior | Days since | After decay | Total | Status |
|---|---|---|---|---|---|---|
| Maya Chen, VP Marketing | 40 | 66 | 3 | 61.6 | 101.6 | MQL |
| Priya Shah, Founder (gmail, 8 staff) | 15 | 53 | 1 | 51.8 | 66.8 | Check fit |
| Grace Kim, Marketing Manager | 30 | 54 | 20 | 34.0 | 64.0 | MQL |
| Tom Becker, Operations Manager | 30 | 30 | 12 | 22.7 | 52.7 | Nurture |
| Luis Ortega, Head of Growth | 40 | 22 | 45 | 7.8 | 47.8 | Nurture |
| Sam Lee, Director of Sales | 40 | 0 | 90 | 0 | 40.0 | Nurture |
| Anna Kowalski, Student | −30 | 45 | 2 | 43.0 | 13.0 | Nurture |
Sam Lee is a strong fit with no recent activity: a target for outreach, not an MQL. Anna Kowalski is the most active lead in the list and the least likely buyer; the student rule keeps her out of the sales queue.
Where lead scores go wrong
Most scoring models fail quietly: the scores look reasonable and stop matching who buys.
Points by gut feel only
Score last quarter's leads and compare won and lost. If customers did not score higher, the rules need changing.
No decay
Without decay a lead that was busy a year ago outranks one who requested a demo yesterday.
Behavior without fit
Students, competitors and job seekers are often the most active visitors. Negative fit rules and a minimum fit keep them out of the sales queue.
Too many rules
Thirty rules with similar points are hard to explain and to maintain. Start with five to ten that sales agrees with.
The events a score is built from
TapCub does not calculate lead scores. It records the activity and outcomes a score uses: form views and submits, chat consultations, and qualification and deals sent back from your CRM, on the same person.
Lead-gen events
lead_form_view, lead_submit, consult_start from your site; lead_qualified, deal_closed, lead_lost from your CRM.
Learn moreSegments on lead properties
Add budget or urgency as properties of lead_submit and build segments or custom metrics on them.
Learn moreChat-captured leads
A number captured in chat is hashed and sent as lead_submit with channel=chat, in the same funnel as form leads.
Learn moreLead scoring questions
Still have a question?
Chat with the team behind TapCub — we usually reply within a few hours on business days.
How do you build a lead scoring model?
List the attributes of customers you win (fit) and the actions that come before a purchase (behavior). Give each a number of points, add negative points for disqualifiers, set a threshold for MQL, and test the model on past leads before switching it on.
What is the difference between fit and behavior scores?
Fit describes who the lead is: title, company size, industry, region. Behavior describes what they did: pages viewed, forms, events, email clicks. Fit changes rarely; behavior goes stale, so it is the part that decays.
What is lead score decay?
A rule that lowers behavior points as time passes without new activity. With a half-life of 30 days, points from a month ago count half. It keeps the score about current interest.
What score should count as an MQL?
There is no standard number; it depends on your points. Pick the threshold that captures the leads sales actually wants: score past leads, look at where customers ended up, and set the line just below most of them.
Can I use this with HubSpot, Salesforce or another CRM?
Yes, as a sandbox. Export leads as CSV, design and test the rules here, then recreate them in your tool's scoring settings. The Markdown export lists every rule for whoever configures the CRM.
Is my lead list uploaded?
No. Parsing and scoring run in your browser. The list is not sent anywhere and is not put into the page address.
How is lead scoring related to cost per lead?
Scoring defines a qualified lead. Once you have it, cost per qualified lead by channel shows where good leads come from; the cost per lead calculator does that.
Update history
New features, fixes and data refreshes for this tool, newest first.
-
· v1.0
LaunchedFirst release.
Missing something or found an error? Tell us — requests that come up often go into the next update. All tool updates
Collect the behavior your score needs
Install the lead-gen template to record form views, submits and chat leads, and send qualification and deals back from your CRM, so the score has data to work with.