Score Inbound Leads

Outcome: a fit and intent score that computes on partial data and drives routing and priority — never whether someone gets a reply.

Surface: Sync GTM app
Level: Advanced
Uses: Formula columns
Credits: 0
Prerequisite: Lesson 05's enriched records

Inbound scoring is a different job

In outbound, the score decides who you spend money on. In inbound, the person has already raised their hand — the money is spent, and the score decides priority and routing, not eligibility.

That difference has one hard consequence, and it is the most important line in this lesson.

Never use the score to decide whether to respond. Everyone who asks a question gets an answer. A low score means a slower, lighter or more automated response — never no response. Companies that silently ignore low-scoring inbound leads are ignoring a founder using a personal email, or an enterprise buyer researching quietly, and they will never know.

Intent is the axis outbound does not have

Fit is the same as everywhere else: size, industry, geography, stack. Intent is specific to inbound and it is the stronger predictor.

Intent signal	Weight
Requested a demo or contact	Highest
Started a trial or signup	Highest
Submitted from the pricing page	High
Asked a specific question in the message field	High
Downloaded a bottom-of-funnel asset	Medium
Downloaded a top-of-funnel asset	Low
Newsletter signup	Lowest

The message field is underused. Someone who typed three sentences about their situation has given you more intent signal — and more personalization material — than any form field.

Score tolerantly

Roughly one in five leads will not enrich completely. A score that requires complete data will fail on exactly those rows, and they will look low-scoring rather than unscored.

Score each axis on what exists

Do not require all inputs.

Track completeness alongside the score

score_confidence = fields_present / fields_expected

Never treat missing as negative

A blank company field is unknown, not small. Scoring it as zero systematically penalizes the leads your enrichment could not reach — which correlates with segment, not with quality.

Route low-confidence rows to a human

A person can resolve in ten seconds what enrichment could not.

A model to start from


fit_score (max 10)
  employee_band in target                 +3
  country in target list                  +2
  industry matches ICP                    +2
  stack signal present                    +3

intent_score (max 10)
  demo or contact request                 +5
  trial or signup                         +5
  submitted from pricing page             +3
  wrote a message over 20 words           +2
  bottom-of-funnel asset                  +2
  newsletter only                         +0

priority = fit_score + intent_score, with score_confidence carried alongside

Here, unlike outbound, adding the two axes is defensible: you are ordering a queue rather than gating spend, and a high-intent poor-fit lead genuinely does deserve attention — just possibly from a different person.

What the score drives

Band	Response
High fit, high intent	Immediate rep response, phone if available
High fit, low intent	Nurture with a personal touch
Low fit, high intent	Fast response, but likely self-serve or a different product tier
Low fit, low intent	Automated response, low-touch nurture
Low confidence	Human review, quickly

Every band gets a response. Only the speed, channel and who sends it differ.

Do this now

Build the fit score

From enriched firmographics.

Build the intent score

Including the message-length signal.

Add score_confidence

Fields present ÷ fields expected.

Confirm missing never scores negative

Check a partially-enriched row by hand.

Define the five response bands

Confirm no band means “no response”

Test on 20 real historical leads

Do the bands match what a human would have decided?

Check your work

Fit and intent are separate columns
The score computes on partial data
Missing fields are unknown, never negative
Every band has a defined response
Low-confidence rows route to a human quickly

Where this breaks

A low score on an unenriched record is a data failure being read as a quality judgement. The leads most likely to enrich badly are small companies, non-English markets and personal email addresses — which is a description of a segment, not of intent. Carry score confidence next to the score, and route low-confidence leads to a human rather than to the bottom of the queue.

Further automation

Feed outcomes back into the weights. After a quarter, compare the score of leads that became opportunities against those that did not — the same converter-versus-non-converter analysis used everywhere else in this library. Inbound gives you this feedback faster than outbound because the volume is higher and the cycle is shorter.

Next lesson

07 — Route to the right rep, turning the score into an assignment.

Reference for this lesson: Tables, CRM integrations, GTM Engineering — scoring and gating, AI-Powered GTM.