Most B2B founders treat every lead the same.
This post covers exactly how to build a lead scoring model that tells you who deserves your time and who doesn't.
What a lead scoring model actually is
A lead scoring model is a point-based system that ranks every lead in your pipeline by how likely they are to buy.
You assign points to the signals that matter
job title, company size, intent behaviours, engagement
and every lead gets a total score.
The higher the score, the closer they are to being ready.
The lower the score, the further back they sit in the queue.
When it's built properly, your scoring model functions like a filter: you stop guessing who to call first and start working a ranked list.
The lead scoring MISTAKE
The common mistake is building a model around the leads you want not the leads most likely to close.
Most founders start with demographic signals only: job title, company size, industry. That's half the picture.
A CEO at a perfectly matched company who has never engaged with anything is not the same lead as a Head of Sales who has visited your pricing page three times this week.
Demographic fit tells you who they are. Behavioural signals tell you when they're ready and you need both.
The 2 signal types every lead scoring model must include
A complete lead scoring model is built on two layers.
1. Demographic signals
The structural characteristics of the lead and their company.
Job title, seniority level, company size, industry, geography, and revenue range. These signals tell you whether this person is worth pursuing at all before you look at anything they've done.
2. Behavioural signals
What the lead has actually done.
Visited your pricing page, requested a demo, opened three emails in a row, booked a call, downloaded a resource.
Behavioural signals are the strongest predictor of near-term buying intent.
A lead that matches your ICP and has hit your pricing page twice in a week is a different conversation than one that hasn't.
Build both layers into your model and your scoring stops being a guess. It becomes a ranked list grounded in fit and timing simultaneously.
How to assign points
Not all signals are equal. The goal is to weight each signal by how strongly it predicts a closed deal.
A simple starting framework:
- High-value signals (7–10 pts): Requested demo, booked a call, visited pricing page, started a free trial
- Mid-value signals (3–5 pts): Visited case study, clicked email link, attended webinar, returned to site 3+ times
- Low-value signals (1–2 pts): Opened email, engaged on LinkedIn, visited homepage
- Demographic fit (1–10 pts): Scale by how closely the lead matches your ICP title, company size, industry, geography
Start conservative. It's easier to lower thresholds once you see how leads score in practice than to rebuild a model that's handing MQL status to everyone.
What a finished model looks like
Here's a simple example based on Flywheeler's own scoring:
- Founder / CEO title: 7 pts
- 5–20 employees: 5 pts
- $500K–$5M ARR: 5 pts
- Based in Australia: 3 pts
- Visited pricing page: 10 pts
- Requested demo: 10 pts
- Opened 3+ emails: 3 pts
- Clicked email link: 5 pts
MQL threshold: 25 points
Any lead scoring 25 or above gets contacted the same day. Anything under 15 goes into a nurtured sequence. Everything in between gets a follow-up within 48 hours.
That's the model. It's not complicated it's just explicit.
If you want to build yours directly, use the Flywheeler Lead Scoring Builder →
It walks you through demographic signals, behavioural signals, and MQL threshold, and lets you test a sample lead against your model before you use it. Takes about 10 minutes.
How to use your score in practice
A scoring model is only useful if it changes what you do.
Step 1: Score every new lead on entry.
Any lead that enters your pipeline, inbound or outbound gets scored immediately against your model. Don't let unscored leads sit in your CRM.
Step 2: Set your MQL threshold.
Decide the minimum score a lead must reach before sales touches them.
Most B2B SaaS teams set this at 40–60% of the maximum possible score. Start tighter and loosen it if your sales team needs more volume.
Step 3: Route by score, not by gut.
Leads above the threshold go to sales immediately.
Leads below go into nurture sequences until they score higher through further engagement.
This stops your best reps wasting time on leads that aren't ready yet.
Step 4: Review monthly.
Lead scoring is a hypothesis. The first version of your model will be wrong in some places. Check monthly: are the leads you're flagging as MQLs actually converting? If not, find the mismatch and adjust the weights.
Skip the build. Get the meetings.
If you'd rather skip the build entirely and have a fully managed outbound system running for your SaaS business ICP defined, leads scored, sequences built, follow-up handled that's what we do at Flywheeler.
We deliver qualified meetings to your calendar. Your only job is to show up and close.
