Glossary · B2BTOFU
Lead scoring
The short answer
Lead scoring assigns points to leads based on who they are (fit) and what they do (behaviour), so sales can prioritise the leads most likely to become customers.
A lead from a target industry who visited the pricing page three times scores higher than a student who downloaded one guide.
Two dimensions
- Fit — company size, industry, role, location, budget
- Behaviour — pages viewed, emails opened, demos requested, WhatsApp replies
Building a model that works
- Analyse your last hundred closed deals and a sample of lost ones.
- Identify attributes and behaviours that distinguished winners.
- Assign points accordingly — and negative points for disqualifiers.
- Set a threshold for MQL status.
- Check monthly: do high scores actually close more often?
In practice
Keep the first model simple — five to eight signals — and review it with sales monthly. Complex models with dozens of weighted fields are hard to explain and harder to trust. If high-scored leads don't close at a clearly higher rate than low-scored ones, change the model.
Example scoring model
| Signal | Points |
|---|---|
| Matches ICP industry | +20 |
| Decision-maker role | +15 |
| Visited pricing page | +10 |
| Requested demo | +30 |
| Opened three nurture emails | +5 |
| Student or job seeker | −30 |
See marketing automation.
Lead scoring in practice
A B2B company gives points for company size, job title and actions such as pricing-page visits. Leads above a threshold go to sales; others receive nurture emails.
Building a lead score
- Fit criteria (who they are)
- Engagement criteria (what they do)
- Negative signals (students, competitors)
- Regular calibration with sales outcomes
Lead scoring for Indian B2B and education
Education and B2B businesses in India often receive high lead volumes of mixed quality; scoring by fit, intent and engagement helps counsellors and sales teams prioritise.
Frequently asked questions
Should we use predictive lead scoring?
AI-based scoring can work well with enough historical data. With little data, a simple rules-based model is often more reliable.
Should lead scores decay over time?
Yes. Engagement from six months ago means less than engagement this week. Most scoring systems reduce behavioural points over time so stale leads fall down the list.
What threshold should send a lead to sales?
Set it with sales and adjust based on conversion data from high- and low-scored leads.
Can AI improve lead scoring?
Predictive models can help with enough historical data, but should be checked against real outcomes.
Should lead scoring use AI?
Predictive scoring can help with enough data; simple rules work well to start.
What are negative scoring signals?
Signals of poor fit, such as students applying for jobs, competitors or invalid contact details.
Related terms
Full glossaryTalk to a strategist
Thirty minutes. Your numbers. A straight answer.
Book a strategy call with the people who'd actually run your account. We'll tell you what we'd do — or that you don't need us yet.