Lead Scoring in CRM: How to Prioritize Leads for Sales
Practice4 minInnovialabs editorial team
In short
Lead scoring adds up points for how closely a lead matches your buyers and how actively they engage; leads above the threshold get handled by sales first.
The short answer
Lead scoring is a number the CRM assigns to each lead so a sales rep knows where to start. Points come from two parts:
- Fit — how closely the lead resembles your existing customers: industry, company size, city, job title, budget.
- Behavior — what the person does: opened the proposal, viewed the pricing page, submitted a second request, replied in a messenger.
High-scoring leads get a call first, medium ones get nurturing, and low ones get automated emails.
How to pick criteria from won deals
Do not invent criteria. Use your CRM history.
- Export closed deals for a period that reflects your current business, split into won and lost.
- For each attribute, compare how often it appears in won versus lost deals: source, industry, company size, region, how fast the client first replied.
- Attributes noticeably more common in won deals get positive points. Those typical of lost deals get zero or negative points.
- Make sure the attribute is actually filled in. A criterion that reps enter only half the time will break the score.
If you have few closed deals, start with the sales team's expert judgment and refine the weights as data accumulates.
Example model
| Criterion | Type | Points |
|---|---|---|
| Industry from the target list | Fit | +20 |
| Job title — manager or owner | Fit | +15 |
| Region you serve | Fit | +10 |
| Opened the proposal or price file | Behavior | +15 |
| Repeat request within a short period | Behavior | +20 |
| No response to several touches in a row | Behavior | −15 |
| Student or "just curious" request | Fit | −30 |
The points are arbitrary: what matters is the ratio between weights, not the exact numbers. Keep the model simple — 5-8 criteria work better than 30.
Thresholds
Split leads into groups:
- Hot — above the upper threshold: priority call, task created for a rep immediately.
- Warm — between thresholds: email, message, follow-up in a few days.
- Cold — below the lower threshold: an automated sequence without a rep.
Set thresholds so the number of hot leads matches what your team can actually handle. If there are more hot leads than reps can process, the threshold is too low.
How to implement it in amoCRM
- Create a numeric Score field on the deal or contact, plus separate fields for fit criteria (dropdowns instead of free text).
- Use the digital pipeline: triggers on events (stage change, field filled, incoming message) update the field value or launch a Salesbot that recalculates points.
- Send behavioral events from the website and email tools through webhooks or an API integration.
- For complex formulas, an external handler is easier: the CRM sends data, a script calculates the score and writes it back through the API.
How to implement it in Bitrix24
- Add a custom Score field to the lead or deal.
- Configure automation rules on stages and a workflow (business process) that recalculates the total with variables and conditions when fields change.
- Use automation rules for routing: above the threshold — a task for the responsible rep and a notification; below — start an email sequence.
- Pass external events (website, messengers) through inbound REST API webhooks.
How to review the model
- Every quarter, compare conversion to payment by group: hot leads should convert noticeably better than warm and cold.
- If won deals often come from the cold group, the weights are wrong.
- Collect feedback from reps: which "hot" leads turned out to be junk and why.
- Change one or two criteria at a time so you can see the effect.
Common mistakes
- Too many criteria that nobody fills in.
- Points only for fit, none for behavior — the score misses that a lead has gone cold.
- No negative points or decay: old activity keeps inflating the score.
- Setting the model up once and forgetting it.