Lead Scoring & Data
Inbound lead scoring with Clay: enrich, score and route every lead
How to enrich every inbound lead in Clay, score fit and intent, and route it to the right rep in HubSpot or Salesforce within minutes.
7 min · Ayhan K. Isaacs · November 2026
- Signal cohorts
- Captured
- Difficulty
- Operator
- Time to run
- About 2 weeks to the first routed lead
- Tools
- Clay, HubSpot, Salesforce, n8n, RB2B, Warmly, Slack, BuiltWith
Most inbound scoring models score activity, so the people who read the most get the highest scores. Students, competitors and job seekers open every email and visit every page. The buyer who filled in one demo form waits behind them.
Inbound lead scoring with Clay works the other way round. Clay enriches each lead first and scores fit from that data, and a second score for intent decides how fast a person responds. A lead that fits and asked for a call reaches a rep in minutes. Everything else waits its turn.
Who this is for
This playbook is for SaaS teams with more inbound than their reps can answer well. Leads come from demo requests, free sign-ups, content downloads and chat, and they land in HubSpot or Salesforce.
It fits when you have a rep or SDR team waiting on leads and sales has started to say the leads are bad. It also fits when nobody can read every lead by hand any more.
Who this is not for
This is not for teams with a handful of inbound leads a week. Read those yourself and call them.
It is also not for teams without a CRM owner. A score needs someone who will change the weights when sales pushes back.
The signal it engineers or captures
Inbound lead scoring works on captured signals. A form or chat inquiry is the strongest one: the buyer wrote to you. A pricing page visit and a product sign-up from a target account come next.
All three only exist on your own site and product, so no competitor sees them. The playbook adds fit data from enrichment, so you know which of these signals deserve a rep today.
The workflow
01 Send every new lead to one Clay table
Point every inbound source at one Clay table: the demo form, the sign-up form, the chat and content downloads. Use a Clay webhook as the source, or a HubSpot or Salesforce trigger when a new contact is created.
One table means one scoring logic. Keep the raw fields: email, name, company as typed, form name, page URL, UTM source and timestamp.
02 Fix the record before you enrich it
Personal emails (Gmail, Outlook, iCloud) hide the company, and free product sign-ups are full of them. Add a formula column that flags personal domains, then run a person lookup on name plus email to find the work profile and the employer domain. If nothing matches, mark the lead "unresolved" and keep it out of the fast lane.
Deduplicate on email and on domain against the CRM. A lead from an account with an open opportunity goes to that owner, not to the queue.
03 Enrich the company and the person
Add these columns in this order: company domain, company name, employee count, industry, HQ country, funding stage and last round date, tech stack, job title, seniority, department and LinkedIn URL. Tech stack comes from BuiltWith or a Claygent prompt on the company site.
Run providers as a waterfall, so the cheapest one goes first and the next only runs when it returns nothing. Stop enriching a lead once it fails a hard filter. That saves credits on leads you will never route.
04 Score fit with a formula
The fit score is a formula column out of 100, so anyone can read why a lead scored what it did. Example weights to tune: employee count in your band 25, industry on your list 20, a country you sell in 15, seniority of director or above 20, a department that buys 10, and a tool in their stack that you integrate with or replace 10.
Hard filters set the score to 0: students, agencies selling to you, competitors, and countries you do not serve. Use a Claygent or Use AI column only for inputs a formula cannot read, such as "Does this company sell to businesses? Answer yes, no or unclear" from the homepage.
05 Add intent as a second score
The intent score sits apart from fit. Example points to tune: demo request 50, pricing page visit in the last 7 days 20, product sign-up 20, second visit within 14 days 10, content download 5. Pull visits from Clay web intent, RB2B or your product analytics.
Keep the two scores apart. A perfect-fit account reading a blog post and a poor-fit company asking for a demo need different handling, and one blended number hides that.
06 Route by the two scores
Write the routing as four lanes and show them on the CRM record. Lane 1 is fit 70 or more with a demo request or pricing visit: assign to the account owner, or round-robin to an AE when there is none, alert them in Slack, and call within an hour.
Lane 2 is fit 70 or more with low intent: the SDR gets a task and sends a personal email the same day. Lane 3 is fit 40 to 69: a nurture sequence and retargeting, rescored when intent changes. Lane 4 is under 40 or a hard filter hit: no rep time, marketing email only.
These thresholds are starting points. Move them after the first month of rep feedback.
07 Write it back to the CRM
Push fit score, intent score, lane, the enrichment fields and a one-line "why" summary to HubSpot or Salesforce properties. In HubSpot, a workflow on the lane property rotates the lead to an owner. In Salesforce, use assignment rules or a Flow.
n8n covers what the native integrations cannot, such as a Slack alert with the lead summary and a link to the record.
08 Review weekly with sales
Once a week, pull every Lane 1 and Lane 2 lead from the last seven days with the rep's outcome: meeting, no reply or wrong fit. When reps reject a Lane 1 lead, find the field that misled the score and change its weight.
Log every change with a date. Later you will want to know which version of the model produced which pipeline.
What it costs and how long it takes
Tools: Clay, your CRM, Slack, and n8n if you need custom routing. Clay credits grow with the number of enrichment columns and the lead volume. Visitor data adds RB2B, Warmly or Clay web intent.
Time: plan two to four days to build the table and the CRM properties. Then run a week in shadow mode, where scores are written but routing does not use them yet, and switch routing on after that. Upkeep is about an hour a week for the review with sales.
The numbers
We have not published an inbound scoring result yet. Measure these from day one: minutes from form fill to first rep touch in Lane 1, meetings held from Lanes 1 and 2, the share of Lane 1 leads reps accept, and pipeline created by lane after 90 days.
Set your targets from the first month's baseline. If Lane 3 creates as much pipeline as Lane 1, the score is wrong.
Where it breaks
Enrichment coverage is the first thing that breaks. Small companies and companies outside the US return thin data, and a missing field quietly scores as zero. Treat missing data as unknown, and send unknowns with a high intent score to a person.
Score drift is the second. Marketing adds a form or a new campaign brings in students, and nobody updates the hard filters. The weekly review catches it.
The third is trust. If sales does not believe the score, reps ignore the lane. Set the weights with the sales lead in the room and show the "why" line on every record.
Speed is the last one. A perfect score is wasted if a Lane 1 lead waits a day. Alert a named person, not a channel, and track the minutes.
Hand-off to sales
A Lane 1 lead reaches the rep as a Slack message and a CRM task with the summary: who they are, company size and stack, what they did on the site, the fit score and why. The rep calls first, then emails.
The first line refers to what the buyer asked for, not to the score. Example: "You asked about pricing for a team your size, so I put the two options side by side. Is a 20-minute call this week useful?" If the rep marks the lead as wrong fit, that tag feeds the weekly review.
FAQ
Related
- Website visitor identification: from content to booked meetings
Publish one piece, promote it with outbound and LinkedIn ads, identify the companies and people who read it with RB2B, Warmly or Clay, then follow up.
- Outbound reply handling: classify, route and answer every reply
How to sort every outbound reply by intent, route interested ones to an SDR within the hour, answer the questions, and catch the referrals.
- Simplified ABM with Clay: qualify 5,000 accounts and score them
A simplified ABM playbook: take 5,000 accounts through Clay for tech stack, news, hiring and org chart checks, score them, and work the top tier by hand.