You do not need a data science team to score leads. You need a short list of signals, a sensible number of points for each, and a threshold above which a human picks up the phone. This guide covers the two families of signal — behavioural and demographic — how to weight individual answers, and how to set a threshold you will actually trust.
Two families of signal
Every scoring signal is one of two types.
Demographic (who they are)
These describe the fit of the person or company: industry, company size, location, role, declared budget, vehicle or product interest. Demographic signals answer "are they the kind of customer we serve?" A lead from outside your service area is a poor demographic fit even if they behaved perfectly.
Behavioural (what they did)
These describe actions: which pages they viewed, how many funnel steps they completed, whether they returned, how quickly they responded, which high-intent option they picked. Behavioural signals answer "are they engaged right now?" A lead who picked "buying within 30 days" has signalled intent far louder than one who picked "just researching".
The useful rule of thumb: demographic tells you if they can buy; behavioural tells you if they will buy soon. A good score combines both. A perfect-fit demographic with zero behavioural signal is a cold prospect; a strong behavioural signal from a poor-fit demographic is a waste of a call. You want both high.
Per-answer weights, not per-question
The mistake most small businesses make is scoring at the question level: "the budget question is worth 10 points". That is too coarse. Score at the answer level, because the answers differ in meaning.
Take a single question, "When are you looking to buy?" Give each option its own weight:
| Answer | Weight |
|---|---|
| Within 30 days | +30 |
| 1–3 months | +15 |
| 3–6 months | +5 |
| Just researching | 0 |
Now a second question, "What is your budget range?"
| Answer | Weight |
|---|---|
| Over £10,000 | +25 |
| £3,000–£10,000 | +10 |
| Under £3,000 | 0 |
A lead who answers "within 30 days" and "over £10,000" scores 55 before any other signal. One who answers "just researching" and "under £3,000" scores 0. The two questions, weighted per answer, have already separated your hot leads from your cold ones — automatically, at the moment of capture.
This is exactly what our funnel builder does: you attach a weight to every option when you build the step, and the lead arrives pre-scored. No spreadsheet, no manual review.
A worked scoring example
Suppose your funnel asks four weighted questions. Weights shown per answer:
- Timeframe: 30 days (+30) / 1–3 mo (+15) / 6 mo+ (+5) / researching (0)
- Budget: >£10k (+25) / £3k–£10k (+10) / <£3k (0)
- Service area: in-area (+20) / bordering (+10) / out-of-area (0)
- Contact preference: call now (+15) / email (+5) / any (0)
Lead A — 30 days, >£10k, in-area, call now: 30+25+20+15 = 90.
Lead B — 1–3 mo, £3k–£10k, bordering, email: 15+10+10+5 = 40.
Lead C — researching, <£3k, out-of-area, any: 0+0+0+0 = 0.
Set your threshold at 60. Lead A is worked immediately, Lead B goes into a nurture sequence, Lead C is logged but not chased. Your salesperson's first call of the day is to Lead A, not to a random alphabetised list.
Setting the threshold without guessing
A threshold is only useful if it maps to your real capacity and your real close rate. Two ways to set it.
Top-down (capacity). If your team can work 30 leads a day and you get 120, you need the top 25% scored. Sort last month's leads by score and draw the line at the 30th. That threshold is defensible because it matches reality, not aspiration.
Bottom-up (close rate). Take last month's closed deals, look at their scores, and find the score below which almost nothing closed. If leads under 50 never converted, 50 is your floor regardless of capacity. Combine the two: the threshold is the higher of "what capacity demands" and "what history says converts".
Then leave it alone for a while. The point of a threshold is consistency — if you move it weekly you cannot compare this month to last.
Behavioural signals you can capture cheaply
You do not need to spy on people. A few honest, captureable behaviours carry most of the signal:
- Funnel completion. A lead who finishes all steps is more engaged than one who bails at step two. Weight partial completion.
- Return visits. Someone who comes back within 48 hours is warmer than a one-and-done.
- Response speed. A lead who replies to your first follow-up within an hour is signalling intent.
- High-intent option picks. "Call me now" beats "subscribe to newsletter" every time.
Common mistakes
- Scoring only demographics. You build a list of "perfect fit" leads who never intended to buy, and your team calls them to silence.
- Negative weights done wrong. Penalising out-of-area with a negative is fine, but don't let penalties stack so deep a single bad answer sinks an otherwise great lead. Cap the downside.
- Setting the threshold, then ignoring it. The score only helps if sales actually works high scores first. A score nobody reads is decoration.
- Never recalibrating. Markets shift. Revisit weights and threshold quarterly against what actually closed.
- Over-weighting one question. If "budget" alone can clear your threshold, the rest of the funnel is pointless. Keep weights spread so no single answer dominates.
From score to action
A score is not the output; the routing is. Map score bands to actions:
| Band | Action |
|---|---|
| 75+ | Call within the hour |
| 40–74 | Same-day email + nurture |
| 0–39 | Logged, automated sequence only |
This is the payoff: a small team stops debating who to call and just works the list in score order. Our lead tracking writes every enquiry to the database before processing, so a scored lead is never lost to a slow downstream step, and attribution tells you which source produced your high-scorers so you can buy more of it.
Getting started this week
You do not need to model this from scratch. Build a short funnel, attach a weight to each answer, set a threshold from last month's data, and route by band. The free plan supports unlimited users, so your whole team sees the same scores at no extra cost — and the 30-day Growth trial gives you scoring and the ROI dashboard immediately, no card.
Lead scoring is not about excluding people. It is about spending your limited time where it returns the most. For a small business, that discipline is the difference between a busy week and a profitable one.