Lead scoring: definition, model and examples to prioritize your prospects

Lead scoring means giving each prospect a score so you know who to contact first. That score combines two things: how closely the person matches your ideal customer, and how much interest they are showing right now. A good score does not predict the future; above all, it stops you from spending the week on contacts who will never buy.

This guide explains how lead scoring works, which criteria to keep, how to build a scoring model in six steps with a worked example, and why most models fail in outbound prospecting.

What is lead scoring?

Lead scoring is a qualification method that ranks prospects by their likelihood of becoming customers. In practice, you define a list of criteria, assign points to each one, and add them up. Above a certain threshold, the lead is considered ready for a sales rep to reach out.

It solves a simple problem: a sales team always has more contacts than time. Without prioritization, everyone works leads in the order they arrive, or on gut feeling. Scoring makes that sorting explicit and shared between marketing and sales.

It is often confused with lead qualification. Qualification is the goal (knowing whether a lead is worth working); scoring is the tool that partly automates it. A discovery call is still the best qualification there is, but scoring decides who deserves that call.

The two dimensions of a lead score: fit and intent

Every serious lead scoring model rests on two axes. Blending them into a single number without telling them apart is the most common mistake.

A VP of Sales at an 80-person company has an excellent fit. If they have posted nothing, visited nothing and have no recent news, their intent is zero: reaching out now is just cold outreach. Conversely, a student who downloads all your guides has strong intent and an off-target profile. The right prospect ranks high on both axes.

You will also hear about explicit criteria (what you know about the prospect: job title, company) and implicit criteria (what you observe of their behavior: visits, clicks, interactions). It is the same distinction under another name.

MQL and SQL: what the thresholds mean

The score is used to move a lead from one stage to the next. The most common labels:

In a small team, these labels can feel heavy. Keep the core idea: a threshold triggers a specific action, and everyone knows who does what when a lead crosses it.

Why traditional lead scoring fails in outbound prospecting

Most lead scoring guides are written for inbound: visitors land on your website, fill out a form, open your emails, and a marketing automation tool tallies the points. The model works when you have traffic.

In outbound, the problem is different. Your best prospects have never visited your site and do not open your newsletters, because they do not know you yet. A model based only on behavior on your own channels gives them zero intent points. And a model based only on fit brings you back to a cold list, sorted by job title.

The fix is to look for intent where it is actually expressed: outside your website. On LinkedIn in particular, many signals are public: someone commenting on a post about your topic, starting a new role, working at a company that is hiring, or engaging with a competitor's content. These signals replace website visits as the fuel for the intent side of the score.

How to build a lead scoring model in 6 steps

1. Start from your most recent closed deals

List your last 10 to 20 customers and, for each one, write down what was true at the time of first contact: industry, size, the contact's job function, and what had happened just before (hiring, new role, a post about a problem, an inbound request). The common traits are your criteria. Do not start from a theoretical list: start from what has already converted.

2. Pick 4 to 6 fit criteria

Keep the criteria that truly separate a good customer from a bad one. For a typical B2B offer, that often means industry, headcount, job function, seniority and location. Beyond six criteria, the model becomes impossible to maintain and nobody remembers why a lead has a given score.

3. Pick 3 to 5 intent signals

Keep observable signals you can actually capture, on your own channels and elsewhere: pricing page visit, webinar registration, LinkedIn comment on your topic, recent job change, hiring for a role related to your offer. A signal you have no way of detecting has no place in the model.

4. Assign points, including negative ones

Give more weight to what best predicts a purchase. Add negative points for what disqualifies: competitor, student, company too small, personal email address when you sell to businesses. Negative scoring is often more useful than positive scoring: it saves you wasted time.

5. Make signals decay over time

Intent has an expiration date. A comment posted yesterday is worth a message today; the same comment three months later justifies nothing. Plan for intent points to decrease or disappear after a set period, for example 30 days for an activity signal and 90 days for a job change.

6. Set a threshold and an action, then review monthly

Define the score at which a rep contacts the lead, and within what timeframe. Then, every month, look at the leads who replied and booked a meeting: what score did they have? If low-scoring leads close and high-scoring leads never reply, the model is wrong. Adjust the weights rather than adding criteria.

Lead scoring example: a sample B2B scoring model

Here is a model for a company selling a solution to sales leaders at small and mid-sized businesses. The points are indicative: what matters is the gap between criteria, not the absolute values.

With simple thresholds: below 30 points, do not contact; between 30 and 50, add the lead to a watch list; above 50, reach out within 48 hours and mention the signal.

Two examples to see the model in action. Claire, VP of Sales at a 60-person services firm, commented last week on a post about how hard it is to hire SDRs: 20 + 15 + 10 + 5 + 25 = 75 points, to contact this week. Mark, CEO of a company with the same profile but no visible activity: 50 fit points, 0 intent, and -10 if he has been inactive for months. He stays on the watch list until a signal appears.

Which lead scoring tools should you use?

The right tool depends mostly on where your signals live.

Start with a spreadsheet. Until you have checked on a few dozen leads that your model really separates good leads from bad ones, automating it only speeds up a mistake.

Lead scoring with intent signals: Jova's approach

Jova starts from intent rather than from a list. The agent detects people on LinkedIn who post or comment about your topic, attend events, change jobs or engage with your competitors. Each prospect gets a score based on fit with your target and the strength of the signal, then Jova finds their contact details and prepares the outreach, on LinkedIn or by email, built around that signal.

Pairing fit with timing is what makes the difference: on Jova campaigns triggered by hot signals, reply rates exceed 20%, compared with under 10% for classic cold outreach. At Aioba, 180 prospects were contacted in 14 days, with 70% positive replies and 9 meetings booked.

The limits are real. Jova relies mainly on LinkedIn data: if your target audience is not active there, there will be few signals to capture. And the tool is not built to assemble a massive contact database: it is meant to help you contact fewer people, better chosen, at the right moment.

Lead scoring mistakes to avoid

• Using too many criteria. A 25-line model never gets reviewed and ends up ignored by reps.
• Scoring fit only. You get a well-sorted cold list, not leads ready to talk.
• Scoring only behavior on your website. In outbound, your best prospects have never been there.
• Forgetting negative points. They save more wasted time than any positive criterion.
• Leaving signals undated. Six-month-old intent is no longer intent.
• Never checking the score against results. If meetings do not come from your top-scored leads, the model has to change.

Lead scoring FAQ

What is lead scoring in one sentence?

It is a method that assigns points to each prospect based on their fit and their interest, so you know who to contact first.

What is the difference between lead scoring and lead qualification?

Qualification is the outcome you want: knowing whether a lead is worth working. Lead scoring is the method that automates part of that sorting before a conversation with a rep, which remains the most reliable qualification step.

How many criteria should a lead scoring model have?

Between 8 and 12 in total is a good rule of thumb: 4 to 6 fit criteria, 3 to 5 intent signals and a few negative criteria. Beyond that, the model becomes hard to understand and to correct.

Can you do lead scoring without a marketing automation tool?

Yes. A spreadsheet is enough to get started and validate the model on a few dozen leads. A tool becomes useful when volume grows or when you want to capture signals automatically, on your website or on LinkedIn.

Does predictive lead scoring replace a manual scoring model?

Predictive scoring calculates a probability from your historical data. It needs enough lead and sales history to be reliable. For a small company that signs a few dozen customers a year, a simple model reviewed every month is often more accurate and easier to understand.