A lead score combines two kinds of signal. Fit describes who the contact is: job title, company size, industry, country. Engagement describes what they do: page views, email clicks, form submissions, meetings booked. Points are added for positive signals and removed for negative ones, such as a competitor domain or months of inactivity. Depending on your subscription, HubSpot offers a score property or a dedicated lead scoring tool with separate fit and engagement scores.
The score is only useful if something happens because of it. The common pattern is a workflow that sets the lifecycle stage to MQL and assigns an owner when the score crosses a threshold. Sales views and segments can also sort by score, so reps start with the highest-ranked leads.
Scores break quietly. Rules written two years ago still award points for a webinar series that ended or a page that no longer exists. Engagement points pile up without decay, so a contact who was active once in 2023 still looks hot. And if fit properties such as industry or company size are mostly blank, the fit half of the score cannot work at all.
Treat a score as a hypothesis to test. Every quarter, compare the scores of contacts who became customers with those who did not. If they look the same, the rules are not predicting anything and should be rebuilt around the signals that did.