Explainer

How Marketing Attribution Works

Attribution tries to answer a deceptively simple question: which marketing effort deserves credit for a customer? This explainer covers how the data is collected, the common models and their limits.

  • 3 min read
  • Sept 2026 last updated
  • Independent educational content
On this page
  1. The question attribution tries to answer
  2. Step one: capture the source
  3. Step two: connect source to outcome
  4. Common attribution models
  5. Why attribution is imperfect
  6. Practical guidance
  7. Connecting to decisions
  8. Keep learning

The question attribution tries to answer

A customer rarely appears from a single source. They may see a social post, search for your name, read an article, click an advertisement and later book after receiving an email. Attribution is the practice of assigning credit for the outcome to those touchpoints. Its purpose is to inform budget decisions: where should the next pound or dollar go?

Step one: capture the source

Attribution is only as good as the data captured at the moment of contact.

  • UTM parameters are tags added to links, for example naming the source, medium and campaign, so that analytics tools can group visits correctly.
  • Referrer information shows the site that sent the visitor, although it is not always available.
  • Hidden form fields can carry the source into a form submission so that it is stored on the contact record.
  • Phone and booking questions ("How did you hear about us?") capture what tracking cannot, such as word of mouth.

Consistency matters more than sophistication. If the same campaign is tagged three different ways, reports fragment.

Step two: connect source to outcome

The source must follow the person into your CRM, and the outcome, such as a sale, must be recorded against the same record. That is why attribution is an integration problem as much as a marketing one: the website, CRM, calendar and payment tool must share the customer identity. Our guide to integrations, API and reporting covers the plumbing.

Common attribution models

Simplified customer path with several touchpoints
  1. Social post
  2. Search visit
  3. Advertisement click
  4. Email
  5. Booking
  6. Sale
  • First touch gives all credit to the first interaction. It highlights what introduces people to you.
  • Last touch gives all credit to the final interaction before conversion. It highlights what closes.
  • Linear splits credit equally among touchpoints.
  • Position-based gives more weight to the first and last touches.
  • Time decay gives more credit to touches closer to conversion.
  • Data-driven models use statistical methods to estimate contribution, usually needing large volumes of data.

No model is objectively true. Each is a lens that emphasises different stages.

Why attribution is imperfect

  • Privacy controls and consent choices mean some visits cannot be tracked.
  • Cross-device journeys are hard to stitch together: research on a phone, purchase on a laptop.
  • Offline influences such as referrals, events and phone calls often go unrecorded.
  • Platform reporting may claim credit for the same conversion in several places.
  • Long sales cycles stretch the path beyond the time windows tools use.

Treat attribution as a set of informed estimates, and combine it with common sense and direct conversation with customers.

Practical guidance

  • Agree a naming convention for campaign tags and stick to it.
  • Capture source on every form and booking page.
  • Store the source on the contact record, not only in analytics.
  • Compare source by customers won, not just leads generated.
  • Use two models side by side, such as first touch and last touch, and note where they disagree.
  • Add a simple "How did you hear about us?" question.
  • Review quarterly, and avoid large budget shifts based on tiny samples.

Connecting to decisions

Attribution should change what you do. If a channel produces many leads but few customers, examine the follow-up before cutting spend. If a small channel produces high-value customers, test giving it more. For applying this to paid campaigns, see advertising and campaign management and our explainer on how conversion rates work.

Keep learning

Keep learning

Explore the rest of the guides

Fifteen problem-first guides, a plain-language glossary and practical checklists, all free to read.

Browse all topicsOpen the glossary