Lead scoring
Lead scoring with the reason behind every point.
A score from 0 to 100 that compares each company with your ideal customer profile, criterion by criterion, and shows why it earned or missed each point. No model decides it.
(01)
What lead scoring is
Lead scoring ranks potential customers so the team spends its time on the best ones first. Fit scoring looks at who the company is (industry, country, size); engagement scoring looks at what it did (opened, replied, visited). Global Sales AI scores fit, against the ideal customer profile you write.
(02)
How the score is calculated
Each company is compared with the profile on six criteria. By default they weigh:
| Criterion | Default weight | What earns the points |
|---|---|---|
| Industry | 30 | the company's industry is one of the profile's |
| Country | 20 | the company's country is one of the profile's |
| Company size | 20 | the employee count or register band lies in the profile's range (partly, for a band that crosses a limit) |
| Keywords | 10 | the profile's keywords appear in the company's record |
| Buyer role | 10 | a contact holds one of the profile's roles |
| Data completeness | 10 | website, country, industry, size and a contact e-mail are known |
A criterion your profile leaves open does not count, neither for nor against. The score is the points earned divided by the points possible, from 0 to 100. The same data always gives the same score, and each criterion shows its reason next to the points.
(03)
Thresholds you set
Two thresholds in the profile turn the score into a decision: at or above the qualified threshold a company is qualified, at or above the review threshold it needs a person's look, below that it is set aside. The defaults are 65 and 40, and you can change them. A company in an excluded country or industry is set aside whatever its score.
(04)
A worked example (fictional)
Profile: software companies in Spain and France with 50 to 500 employees, keyword “ERP”, no buyer role set. Company: “Example Logiciels SAS”, a fictional French software company in the 50–99 employee band, with a website and no contact yet.
| Criterion | Points | Reason |
|---|---|---|
| Industry | 30 of 30 | software is in the profile |
| Country | 20 of 20 | France is in the profile |
| Company size | 20 of 20 | 50–99 lies inside 50–500 |
| Keywords | 0 of 10 | “ERP” not found in the record |
| Buyer role | not counted | the profile sets no role |
| Data completeness | 8 of 10 | missing: contact e-mail |
78 of 90 possible points gives a score of 87: qualified with the default thresholds. The reasons tell you what to check before writing — here, whether the company sells or uses an ERP.
(05)
Building your own scoring model
- Start from the customers you won: what do their records have in common?
- Put those traits in the profile: industries, countries, size range, keywords, roles.
- Score a list and read the reasons of the companies near each threshold.
- Adjust the profile or the thresholds when the reasons show a pattern you disagree with.
Questions and answers
Questions
What is lead scoring?
A way to rank potential customers by how well they fit what you sell, so the best ones are contacted first. Global Sales AI scores each company from 0 to 100 against your ideal customer profile.
How is the score calculated?
Six criteria — industry, country, company size, keywords, buyer role and data completeness — each with a weight. Criteria your profile leaves open do not count. The score is the points earned out of the points possible.
Can I change the thresholds?
Yes. The qualified and review thresholds are fields of each ideal customer profile; the defaults are 65 and 40.
Does AI decide the score?
No. The score is calculated by fixed rules from the company's data, so the same data always gives the same score. AI is used to draft messages, not to score.
Sources
Content checked on Nov 20, 2026.
Score your first list.
Create an account, write your ideal customer profile and score companies from the directory or from your own file. The plans are on the pricing page.