Attribution models sound complicated, and the industry does not help by naming them things like "U-shaped" and "time decay". Underneath the names, every model answers one simple question: when a deal closes, which of the recorded visits gets the credit?
This guide explains the three models that matter in plain words, what each one is useful for, and the one limitation they all share that nobody puts on the pricing page.
The Three Models in Plain Words
Imagine a buyer who visited your site four times before signing up: once from a LinkedIn post, once from a Google search, once from an email, and finally from a search for your company name. An attribution model is simply the rule that decides how those four visits share the credit.
How it works: The very first recorded visit gets all the credit.
Good for: Understanding what starts journeys and creates awareness.
Blind spot: Ignores everything that happened afterwards, and the true first touch is often invisible anyway.
How it works: The final visit before conversion gets all the credit.
Good for: Seeing what closes, such as branded search and pricing page visits.
Blind spot: Rewards the end of the journey and hides whatever created the demand in the first place.
How it works: Credit is shared across every recorded visit in the journey.
Good for: A fairer spread across the touches your tools can actually see.
Blind spot: Can only share credit between recorded clicks. Unrecorded influence still gets nothing.
The Limitation They All Share
Every attribution model, from the simplest to the most sophisticated, works with the same raw material: the visits your tools managed to record. The model can slice that credit in different ways, but it cannot give credit to anything outside the record.
In B2B, a lot of influence lives outside the record. The colleague who made the recommendation. The podcast episode listened to during a commute. The community thread where three people praised your product. None of these produce a recorded visit at the time they happen, so no model can credit them. The click record starts partway through the story, and the models can only argue about the chapters they have.
This is why switching models often feels underwhelming. You reshuffle credit between the same recorded clicks and the picture changes less than expected. The missing influence stays missing. We looked at where those recorded stories mislead in why your attribution dashboard is lying to you.
Changing your attribution model redistributes credit between the clicks you recorded. It cannot conjure up the recommendation, the conversation or the episode that actually started the journey.
A Setup That Works in Practice
You do not need a complicated model. You need two views you trust, read together.
Keep a simple click model. Last touch or a basic multi touch view is fine. Use it to understand the end of journeys: what people search for when they are ready, which pages close, where forms are abandoned. This is what click data is genuinely good at.
Add the buyer's own account. One open text question on your forms, "How did you hear about us?", captures the start of the journey in the buyer's own words. This is the part no model can reach. Together, the two views cover the whole story: the model shows how journeys end, the answers show how they begin.
When the two views disagree, do not average them. Investigate. The disagreement is usually the most valuable information in your whole reporting stack, because it points straight at influence you are underinvesting in. Our guide to demand creation and demand capture explains why that gap appears again and again.
Pick Simple, Add Honest
Teams spend months debating attribution models when the honest answer is that all of them see the same partial record. A simple model you understand beats a sophisticated one you do not.
Spend the saved energy asking buyers how they actually found you.
Frequently Asked Questions
Which attribution model is best for B2B?
None of them is best on its own, because all of them can only divide credit between the touches that were recorded. B2B journeys are long and much of the influence happens in places no tool records, such as conversations, communities and content consumed without clicking. The most useful setup is a simple model you understand, read alongside what buyers tell you directly in survey answers.
What is the difference between first touch and last touch attribution?
First touch gives all the credit for a deal to the earliest visit your tools recorded. Last touch gives all the credit to the final visit before the person converted. The same deal can tell two completely different stories depending on which model you pick, which is a strong hint that neither story is complete.
Is multi touch attribution worth the effort?
It can be, if you already have clean tracking and you treat the output as a rough map rather than the truth. Multi touch spreads credit more fairly across recorded visits, but it cannot see the conversations and recommendations that never produced a click. If your tracking is messy, fixing the basics and adding a survey question will teach you more than a sophisticated model will.
Why do my attribution model and my customers tell different stories?
Because they measure different things. The model measures recorded clicks. Customers remember what actually influenced them, which often happened before any recorded click: a recommendation, a podcast, a post they read. When the two disagree, the gap is usually the influence your tools could not see, not a fault in your customers' memory.
James Kevan is the co-founder of First Signals, which pairs your click-based attribution with what buyers say in their own words, so the two views can be read together.
Related guides: Why Your Attribution Dashboard Is Lying To You · Demand Creation vs Demand Capture · Why Is My Direct Traffic So High?
© 2026 James Kevan / firstsignals.ai. Share freely with attribution.