Skip to content

Sales & CRM Systems

Lead scoring: call the right lead first, and know why

When there are more good leads than people to call them, something decides the order. Usually it's whoever arrived first. Scoring replaces that with a priority your team agreed on, and shows its working for every lead.
Discuss your systemSee eight leads re-ordered ↓
Arrival orderScored order1234567808:02Annual service plan3408:15Water heater leak, next day8108:31Bathroom remodel quote5508:40Drain backup, water on the floor9208:52Dripping faucet2809:03Existing customer, AC check6609:10Referral, four-unit building7409:22“Need a plumber”, no addressask first40

Arrival order is a priority system too. Just a bad one.

The same eight qualified leads from one morning: the order they arrived, against the order they should be called.

Scored order

  1. 108:40Drain backup, water on the floorwas #492
  2. 208:15Water heater leak, next daywas #281
  3. 309:10Referral, four-unit buildingwas #774
  4. 409:03Existing customer, AC checkwas #666
  5. 508:31Bathroom remodel quotewas #355
  6. 609:22“Need a plumber”, no addressask firstwas #840
  7. 708:02Annual service planwas #134
  8. 808:52Dripping faucetwas #528

The drain backup arrived fourth and goes first. The inquiry with no address isn't low priority; it's incomplete, so the first call asks for the address.

What goes into a score

Five kinds of signal cover most service and B2B businesses. The weights are yours to set, and to change.

  1. 01

    Fit

    How well the job matches what you do best

    Your highest-margin services, ideal job size, the areas you can reach fastest. Usually the heaviest weight.

  2. 02

    Urgency

    How soon it needs attention

    “Water on the floor” outranks “sometime this spring”. Often read from free text.

  3. 03

    Source

    Where it came from

    Referrals and repeat customers tend to book more often than cold directory leads. Your own outcome history should set this weight, not a guess.

  4. 04

    Relationship

    Whether you already know them

    Existing customers, past quotes, open opportunities.

  5. 05

    Completeness

    How much is known

    Not a penalty. A separate flag that changes the first action from “quote” to “ask”.

Same score, different certainty

A score built on missing information looks as confident as one built on facts. Recording what's unknown keeps the two apart, so a lead with little information gets a question, not a lower place in the queue.

Lead A · all signals known72
  • Fit +28
  • Urgency +24
  • Source (referral) +12
  • Relationship (past customer) +8
Lead B · two signals unknown72
  • Fit +30
  • Urgency (read from “ASAP”) +30
  • Source +12
  • Location unknown
  • Job size unknown

First action: confirm address and size before quoting.

Known signalUnknown, not scored

Scores you can argue with

  1. 01

    Every score shows its working.

    The signals and weights behind it are stored with the lead, so “why is this first?” has an answer.

  2. 02

    Owners can correct it.

    A correction is saved with its reason, the same as any other change.

  3. 03

    The weights get checked against what happened.

    Which high scores booked, which low scores would have. Weights move when outcomes say they should.

  4. 04

    Models help where language is messy.

    Reading “ASAP” or “water everywhere” as urgency is a good use of a model. Deciding the weights isn’t; people do that, and test changes before they go live.

What scoring hands off

In

Qualified leads
fields · free text
Relationship history
customers · past quotes · open deals
Agreed weights
set and changed by people
Scoring
signals × weights · completeness

Out

Score
with its contributing signals
Completeness flag
quote now, or ask first
Priority
read by routing and follow-up

Worth building when there's a queue

Good fit

  • More qualified leads arrive than the team can contact right away.
  • Some leads are clearly worth more, or more urgent, and the team can say which.
  • You have outcome history to check weights against, or are ready to start keeping it.

Not the right fit

  • Every lead already gets a reply within the hour.
  • Nobody can say how a higher score should change what happens next. Then a score is just a number.

Where scoring does its work

Next to scoring

Questions about lead scoring

What's the difference between lead scoring and lead qualification?

Qualification decides whether a lead fits: yes, no or needs review. Scoring orders the leads that fit. Scoring an unqualified list just ranks work you shouldn't do.

Do we need machine learning for lead scoring?

Usually not to start. A rules-based score with clear weights, checked against outcomes, beats a model nobody can explain. Models help with reading free text, and later with tuning weights once there's enough history.

How many points should each signal get?

Start with the order your best salesperson already uses, turn it into weights, and adjust as outcomes come in. The exact numbers matter less than agreeing on them.

Should a lead with missing information score lower?

No. Mark it incomplete and make the first action a question. Otherwise the leads that most need a call get the least attention.