Guide
Lead Scoring Models: How to Build One That Actually Qualifies Leads
Most lead scoring articles stop at "lead scoring helps you prioritize prospects." True, but useless. This is about the actual model: which signals to score, how to weight them, and what happens automatically the moment someone crosses a threshold.
Quick answer
What is a lead scoring model?
A lead scoring model is a set of weighted rules that turns a prospect's answers into a number, and that number into a decision. A good model scores 4 to 6 signals (fit, budget, urgency, authority, readiness), weights the ones that predict a sale higher than the ones that just show interest, and maps each score range to a specific next action instead of leaving you to guess.
What is a lead scoring model?
A lead scoring model is the logic behind lead scoring: the rules that turn "this person filled out a form" into "this person is worth 82 points, so book them a call." Without a model, lead scoring is just a number with no explanation. With one, every point has a reason.
The model below is a starting point, not a formula to copy exactly. Budget 25 points for problem fit, 25 for budget, 25 for timeline, 15 for decision authority, and 10 for company size, and you get a lead who is either close to 100, ready now, or close to 0, not yet. That gap is the whole point. A scoring model should spread leads out, not cluster them in the middle.
A simple lead scoring model
| Problem fit | Exact problem you solve | +25 |
| Budget | €10k+ | +25 |
| Timeline | This month | +25 |
| Decision authority | Yes | +15 |
| Company size | Ideal range | +10 |
| Total | 100 | |
Build the scoring logic in FluoTest
FluoTest is a free scored quiz builder. You assign points to each answer, set score tiers, and trigger a different outcome, an email, a redirect, a calendar link, for each tier automatically. The model in this article isn't hypothetical. It's exactly what FluoTest's question weighting and disqualifier logic are built to run.
No code, no spreadsheet, no manually sorting through form submissions. A prospect answers 5 to 8 questions, gets scored in real time, and lands in the right bucket before you ever see their name.
Try it free at fluotest.com
How to build your lead scoring model
1. Pick 4 to 6 signals that predict fit, not just interest
A signal like "how did you hear about us" tells you nothing about whether someone will buy. Stick to fit, budget, urgency, authority, and readiness, and drop anything that only measures curiosity.
2. Assign point weights, not equal weights
Not every question deserves the same number of points. A weak signal might be worth +5, a strong one +20, and a disqualifying answer should subtract points or end the quiz outright. If every question is worth 10 points, you have a survey, not a scoring model.
3. Set score tiers before you see a single response
Decide the cutoffs first: what counts as qualified, what counts as a maybe, what counts as not yet. Set tiers after looking at your first batch of leads and you'll unconsciously bend them to match what already happened.
4. Map each tier to an action, not just a label
A score without an action is trivia. 80 to 100 should trigger a calendar link. 50 to 79 should trigger a case study or a nurture sequence. 0 to 49 should trigger a resource and nothing else. The model only earns its keep once it decides what happens next.
The 5 signals worth scoring
Most lead scoring models fail because they score 20 things instead of 5. Start here, and only add more once these are working.
Fit
Does this person actually need what you sell? Not "could they hypothetically use it" but "does their stated problem match your specific solution."
Budget
Can they realistically buy? Not the biggest possible budget, just enough to afford what you're offering.
Urgency
Do they need it this month, or "sometime next year, maybe"? Urgency predicts close speed better than almost anything else.
Authority
Can they actually make the decision, or do they need to go convince someone else first? Score the second case lower, not zero.
Readiness
Are they willing to take the next concrete step, a call, a demo, a trial, right now? Interest without a next step isn't a qualified lead.
Three example lead scoring models
The signals stay similar, the weighting and the next step change by business type. Use these as starting points, not final answers.
Consultant
- Problem fit25
- Budget25
- Timeline20
- Decision authority15
- Readiness15
70+ → Book a strategy call
40-69 → Send a useful resource
0-39 → Add to nurture
Agency
- Project budget25
- Scope clarity20
- Timeline20
- Existing assets10
- Decision maker15
- Service fit10
75+ → Route to a proposal call
45-74 → Send case studies, keep nurturing
0-44 → Add to newsletter, revisit in 90 days
B2B SaaS
- Company size20
- Current solution15
- Use case fit25
- Number of users15
- Pain severity15
- Buying timeline10
80+ → Route to sales for a demo
50-79 → Enroll in a nurture track
0-49 → Send self-serve resources
The biggest mistake: scoring everything equally
"We're a 50-person company" is a fact. "We're a 50-person company, we're spending €20k a month on this problem, and we need a fix this month" is a completely different lead. If your model gives both answers the same weight because they technically answer the same question, it isn't actually scoring anything.
Fix it by weighting specificity and urgency higher than a bare fact. A vague answer should score lower than a concrete one, even when they're answering the same question.
Turn the score into an action
A score of 82 isn't an endpoint, it's an instruction. The model should tell you exactly what happens next for every range, before a single real lead comes in.
80-100
Qualified
Book a call automatically
50-79
Potential fit
Send a case study, keep nurturing
0-49
Not ready
Send a free resource, revisit later
A score isn't a scientific measurement
An 82 out of 100 isn't a precise measurement of how good a lead is. It's a decision rule you wrote. The only thing that matters is whether that rule leads to better decisions than guessing. Recalibrate the weights against what actually closes, not what feels intuitively right on day one.
Why a scoring model beats a gut-feel qualification call
- -Every lead gets scored the same way, so "gut feeling" stops deciding who gets your time
- -The model runs the moment someone submits, before you ever open your inbox
- -New team members can qualify leads consistently without years of pattern-matching
- -You can see exactly why a lead scored low and adjust the model, instead of just feeling like something was off
- -The score doubles as a record: you can prove which signals actually predicted a close
Who should build a lead scoring model
- -Consultants who get too many discovery call requests to take them all
- -Agencies qualifying project inquiries before a proposal call
- -Coaches sorting serious applicants from browsers
- -Freelancers who need to protect their calendar
- -B2B SaaS teams routing demo requests to sales
- -Recruiters scoring candidate or client fit
- -Service businesses with a mix of good-fit and bad-fit inquiries
- -Sales teams drowning in unqualified form fills
Frequently Asked Questions
What is a lead scoring model?
A lead scoring model is a set of weighted rules that convert a prospect's answers into a number, and that number into a next action. It typically scores 4 to 6 signals like fit, budget, urgency, authority, and readiness, then maps score ranges to specific outcomes.
How do you assign points in a lead scoring model?
Weight signals by how strongly they predict a close, not equally. A weak signal might be worth +5, a strong one +20 to +25, and a disqualifying answer should subtract points or end the quiz. Decide the weights before you see real responses, then recalibrate against what actually converts.
What's the difference between lead scoring and lead qualification?
Lead qualification is the judgment call: does this person fit? Lead scoring is the measurable version of that judgment, a number built from weighted signals. Our lead qualification questions guide covers the individual questions to ask; this article covers the scoring logic that turns those answers into a decision.
How many signals should a lead scoring model use?
Start with 4 to 6: fit, budget, urgency, authority, and readiness. More signals don't make a model more accurate, they usually just add noise and slow down the form. Add a sixth only if it clearly predicts who closes.
Can a solo consultant or small business use a lead scoring model?
Yes, and arguably it matters more for solo operators, since every unqualified call costs a disproportionate share of your week. A simple 5-signal, 100-point model with three tiers, book a call, send a resource, nurture, is enough for most small businesses and consultancies.
Build your scoring model as a live quiz in FluoTest
Assign the points, set the tiers, and let FluoTest sort every lead automatically. Free, no credit card.
With FluoTest you can:
- - Weight each answer exactly like the models above
- - Set score tiers and trigger a different outcome per tier
- - Disqualify bad-fit leads before they book your calendar
- - Send qualified leads straight to a booking link or your CRM