Lead scoring is a way of giving every lead a number, so sales knows who to contact first. The number comes from points for who the lead is and what the lead does.
This page defines the term, walks through a scored lead, and separates it from lead grading and routing. It ends with the LeadOcean filters that cover the "who they are" half.
Key takeaways
- Lead scoring adds and subtracts points to rank leads by how likely they are to buy.
- Two inputs matter: fit (who they are) and engagement (what they do).
- A score is only useful if it changes what a rep does next, such as a call, a sequence or a hand-off.
- LeadOcean supplies the fit data through search filters. Engagement signals come from your own site and email tools.
What it is
Lead scoring is a ranking method that assigns each lead a numeric value, based on fit and behavior, so the team can prioritize who to contact and when.
A lead with a high score gets a fast human reply. A lead with a low score goes to nurture or waits. The point is to spend rep time where it is most likely to pay off.
Fit scoring uses facts about the person and company: job title, seniority, company size, industry, country, tools in use. Engagement scoring uses actions: a pricing page visit, a reply, a demo request, an email click.
Scores can go down as well as up. A personal email address, a student title or a long silence are common negative signals.
How it works
Every model follows the same loop. You define a good customer, assign points to traits and actions, add them up, and act on the total.
- Look at closed-won customers and write down what they share: title, headcount, industry, region.
- Give each fit trait a point value. The traits that match your best customers get the most.
- Give each action a point value. A demo request outweighs a blog visit.
- Add negative points for bad signals, such as a competitor domain or an invalid email.
- Set a threshold. Above it, a lead is passed to sales. Below it, the lead stays in nurture.
- Review the model every quarter against what actually closed.
Worked example, with placeholder data. Jane Doe is VP of Sales at Acme, a 120-person software company in the US. She downloaded a guide and visited the pricing page.
| Signal | Type | Points |
|---|---|---|
| Job level is VP | Fit | +20 |
| Job function is Sales | Fit | +10 |
| 51-200 employees | Fit | +15 |
| Country is US | Fit | +5 |
| Visited pricing page | Engagement | +15 |
| Downloaded a guide | Engagement | +5 |
| Total | 70 |
Notice what the total hides. Fit gave Jane 50 points before she did anything. Her two actions added 20 more. A lead with the same fit and no activity would still sit at 50, close to the line.
If your threshold is 60, Jane goes to a rep today. The same person at a 5-person company with no activity would score 25 and stay in nurture. These point values are invented for the example. Yours should come from your own closed deals.
Engagement points should expire. A pricing page visit from last week matters. One from eight months ago does not.
Most teams subtract a few points for every 30 days of silence. They also cap points per action, so one person clicking the same email ten times does not outrank a real buyer. Without decay, old leads drift up the list and crowd out fresh ones.
Lead scoring vs lead grading
Lead scoring and lead grading get used as one word, but they measure different things. Grading is fit only. Scoring usually adds behavior on top.
| Lead scoring | Lead grading | Lead routing | |
|---|---|---|---|
| Question it answers | How ready is this lead? | How good a match is this lead? | Who should get this lead? |
| Inputs | Fit plus engagement | Fit only, often A to F | The score, grade, region, owner rules |
| Changes over time | Yes, with every action | Rarely | Only when rules change |
| Output | A number | A letter or tier | An assigned owner |
The confusion is natural, because many tools label one field "score" and compute it from fit alone. Check what a number contains before you trust it.
Some teams keep both. The grade says whether the lead is worth a rep's time at all, and the score says whether to call now. See lead routing for how the number turns into an owner.
If you only have time for one, start with fit. It is cheaper to build and it does not need weeks of behavior data.
When it matters
When sales has more leads than time
If reps cannot call everyone, they pick by gut. A score makes the pick consistent. It also stops good leads from sitting untouched while reps chase noisy ones.
When marketing and sales disagree on lead quality
A shared threshold gives both teams one definition of "qualified". Marketing stops passing every form fill, and sales stops ignoring the queue. Write the rule down and agree on it once.
When you buy or build outbound lists
Outbound leads have no engagement history yet, so only fit scoring applies. Score the list on title, size, industry and region before anyone sends a message. Contact the high-fit group first and test the rest at lower volume.
When you set up automation in a CRM
Most CRMs can compute a score from properties and trigger a workflow at a threshold. For the HubSpot steps, read how to set up lead scoring in HubSpot. For model design, read how to build a lead scoring model.
How LeadOcean handles it
LeadOcean covers the fit half of lead scoring. It does not track page visits, email opens or any engagement signal, and it has no scoring engine, so you apply the points in your CRM or a spreadsheet.
What it gives you is the fit data as search filters. The fields that map to a typical fit score are jobLevel, jobFunction, seniority, employeeRange, industry, country and technologies (LeadOcean OpenAPI, September 2026). Each one can be a scoring rule, or a way to pull only the leads that already pass your fit bar.
On 2026-10-01, leadocean_count_leads returned 17,133 mailable people with jobLevel VP, jobFunction Sales & Business Development, country US and employeeRange 51-200. That count uses the default email filter: verified, catch_all_valid and catch_all.
Sizing is free. Send count=true as a query parameter with limit=1, and no records are spent:
curl -X POST "https://api.leadocean.io/v1/people/search?count=true&limit=1" \
-H "x-api-key: $LEADOCEAN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"jobLevel": ["VP"],
"jobFunction": ["Sales & Business Development"],
"country": ["US"],
"employeeRange": ["51-200"]
}'Run it once per scoring tier to see how big each group is. Then pull the top tier with a normal search, where each person returned counts one record. The same filters work in the MCP server. They also work on the Exports page of the app at app.leadocean.io, which shows the record price before you start (LeadOcean exports docs, September 2026).
Enum values are exact strings, so check them with GET /v1/enums before you build rules. Pricing is two plans: Free (1,000 records, one-off, no card) and Pro at $499 a month. See pricing.
FAQ
What is a good lead score?
There is no universal number. A good score is one that separates leads who close from leads who do not, in your own data. Start with a threshold, then move it until reps agree that leads above it are worth calling.
What is the difference between a fit score and an engagement score?
Fit says whether the lead looks like your best customers. Engagement says whether the lead is showing interest now. A strong fit with no activity is a target for outreach. A weak fit with high activity is usually not worth a rep.
Do I need AI for lead scoring?
No. A rules-based model with 8 to 12 criteria works for most small teams. Predictive models need a large history of closed deals to learn from. Start with rules, and only move to a model when you have the volume.
How often should I change the model?
Review it every quarter. Compare the scores of leads that closed with those that did not, and move points toward the traits that actually predicted a win. Change one thing at a time so you can see what helped.
What data do I need to start?
Start with fit fields you already hold or can look up: title, company size, industry and country. Add engagement later, once you track it. A fit-only model is a fair first version, and you can size it against a database before you commit.
Size your best-fit leads before you build the score
Free to start. No credit card. 1,000 records to spend whenever you like.
Get your free API key →