What is lead attribution? And how do you actually track it.

Lead attribution is the practice of crediting the marketing channel, campaign and source that produced a lead or a paying customer. Every inbound enquiry came from somewhere: a paid search ad, an organic result, a referral from another site, a link in an email. Attribution names that somewhere precisely and records it against the lead record.

Without it, the field in your CRM says 'Direct' or 'Web', which is the entire internet. With it, you know which campaign is worth renewing and which one to cancel. That is the whole point of lead attribution: connecting spend to outcome rather than guessing.

Attribution models: first touch, last touch, and what sits between

An attribution model is the rule for deciding how much credit each marketing touch receives. Most businesses need two numbers: what created the lead and what closed it.

First-touch attribution

All credit goes to the first interaction: the channel or ad that introduced the lead to your business. Good for understanding what fills the top of the funnel. Tells you nothing about what closes. See first-touch attribution for the full breakdown.

Last-touch attribution

All credit goes to the final interaction before the form was submitted. Good for seeing what closes. Ignores every influence that ran before it, including the blog post, the ad, or the referral that created the interest three months earlier.

Multi-touch attribution

Credit split across every touch in the journey. Theoretically more complete. In practice, most service businesses do not have the data volume or the tooling to make it reliable. It is the right model for an e-commerce business with thousands of transactions; it is overkill for a law firm getting twelve enquiries a month.

Where attribution goes wrong in practice

Most businesses have attribution in theory. The data that arrives in the CRM is another matter entirely. Three things break it consistently:

UTMs vanish

Ad blockers strip them. Visitors copy links and share them. The visitor comes back direct three days later on a different browser. The UTM that was present at first click is gone by the time the form is submitted, so the source arrives as 'Direct' rather than the Google Ads campaign that actually sent them.

The CRM says 'Web'

Free-text lead source fields collect forty spellings of the same channel and a lot of creative naming. 'Google', 'google ads', 'Adwords', 'PPC'. None of it aggregates into a reportable number. Normalising the data later costs more than capturing it correctly at the start.

Last click wins by default

Most standard attribution setups record whatever the visitor's last session looked like. The blog post that created the interest three months ago gets no credit. The branded search they did right before submitting the form does. That makes content look useless and branded search look like a growth channel, when in reality it is often just the last door someone walks through.

How lead attribution works technically

A tracking script on your website reads the referrer, the UTM parameters, and the page the visitor landed on when they first arrived. It stores those values, typically in a first-party cookie or browser storage, and attaches them to the form submission when it happens. The result is a source record that travels with the lead into your CRM.

The gap in most implementations is that the script captures the current session, not the original visit. If the customer first arrived via a Google Ad in January but returns direct in March to submit the form, a session-level implementation records 'Direct'. A first-touch implementation that stores the original visit in a long-lived cookie records the Google Ad. Both numbers are real; they answer different questions.

Server-side capture is more reliable than client-side capture for one reason: ad blockers, browser privacy settings and cookie restrictions apply to client-side scripts. A server-side implementation that captures at the moment of form submission rather than in the browser is not affected by those signals being stripped.

What to do with attribution data

The most useful question attribution data answers is: which channels produce customers, not just leads? A channel that sends a hundred enquiries and converts two is not the same as a channel that sends twenty enquiries and converts fifteen. Attribution that only captures the source of the enquiry is the starting point; connecting that source through to the closed deal is the full picture.

In practice, most service businesses start with source-per-lead and find that alone shifts where they put the budget. If you have been spending equally across Google Ads, Facebook, and a directory listing, and attribution shows that ninety per cent of closed jobs came from Google Ads and the directory produces none, that is a budget decision, not a reporting exercise.

For a deeper look at how to track this in practice, Lead source tracking captures first touch, last touch and the full page journey on every form submission, and the source lands on the lead record in your CRM.