Revenue Attribution Model

A revenue attribution model is the specific method used to assign credit for closed revenue across the marketing and sales touchpoints that contributed to a deal. Where revenue attribution is the practice, the model is the rule that decides how the credit is split, and that choice shapes what the data appears to say.

Because a typical B2B deal involves many touches, the question "which touch gets the credit?" has no single right answer. Different models answer it differently, and each tells a different, partial story, which is why understanding the models is essential to using attribution well rather than being misled by it.

What a revenue attribution model is

An attribution model is a set of rules for distributing credit for a conversion or closed deal across the touchpoints in the customer journey. Some models give all the credit to a single touch; others spread it across many. The model you choose determines which channels and campaigns look valuable, so it is less a neutral measurement than a lens, and choosing it deliberately matters.

A worked example: one deal, five models

A $30,000 deal had four recorded touches before it closed: a prospect found a blog article through search, later attended a webinar, then replied to an outbound email sequence, and finally booked a demo from a retargeting ad. The illustrative credit under each model:

The W-shaped column assumes the webinar was where the lead was created and the demo was where the opportunity opened, with the remaining credit on the sequence. The same deal makes content look like the whole story, like nothing, or like a quarter of it, depending on the rule. That is the reason to choose a model on purpose and to look at more than one.

What each model is good for

• First-touch answers "what brings new people in?" It is useful for judging awareness channels and poor for judging anything that converts.
• Last-touch answers "what tipped the decision?" It is simple and often built into ad platforms, but it overstates channels that capture existing demand, such as branded search and retargeting.
• Linear is a fair starting point when you do not know which touches matter more. Its weakness is that it treats a passing ad view the same as a sales meeting.
• Time-decay suits short, sales-led cycles where recent touches really do carry more weight.
• W-shaped and other position-based models reflect B2B milestones: first touch, lead creation and opportunity creation.
• Data-driven models learn weights from converting and non-converting paths. They are the most realistic when data is plentiful and the least transparent. Google, for instance, explains how its attribution models work in Analytics, including the data-driven option.

What attribution models cannot see

Every model only distributes credit among touches it can see, which is why journey analytics starts with capturing the journey itself. Word of mouth, a podcast mention, a peer recommendation or a conversation at an event often leave no digital trace, yet they drive many B2B purchases. Two practical fixes help. First, ask: a "How did you hear about us?" field on forms and in discovery calls captures what tracking misses. Second, test: for major channels, a holdout or geographic test measures whether the channel creates revenue that would not have happened otherwise, which no attribution model can prove on its own. See marketing causal inference and revenue lift.

Attribution also depends on sales activity being recorded. If calls, emails and meetings are not logged, the model will credit marketing touches for deals that sales created. Automatic activity capture in the CRM is therefore an attribution prerequisite, not a separate project, see CRM reporting and marketing attribution.

Common attribution models

Single-touch vs multi-touch models

The biggest divide is between single-touch models (first- or last-touch), which credit one interaction, and multi-touch models (linear, time-decay, W-shaped, data-driven), which distribute credit across the journey.

Single-touch models are simple but blunt, last-touch ignores everything that created the opportunity; first-touch ignores everything that closed it. Multi-touch models, the focus of multi-touch attribution, are more realistic but more complex and data-hungry. The right choice depends on your sales motion and the data you can reliably capture.

Why the model matters

• It shapes decisions. The model determines which channels look effective, and therefore where budget flows.
• It can mislead. The wrong model systematically over- or under-credits parts of the journey.
• It enables comparison. A consistent model lets you compare channel performance fairly over time.
• It frames the conversation. Marketing and sales align (or argue) based on what the model credits.

Choosing an attribution model

There is no universally correct model, only the one most useful for your situation. Short, simple sales may be served fine by last-touch; long, multi-touch B2B journeys need a multi-touch or data-driven model to reflect reality. Many teams view several models side by side, since the differences between them are themselves informative. The key discipline is to treat the model's output as directional, within the broader practice of revenue attribution, rather than as exact truth, since no model captures every influence.

Common attribution model mistakes

• Defaulting to last-touch. Using it by default credits only the final touch and undervalues everything upstream.
• Treating the model as truth. Any single model's output is a lens, not reality; read it directionally.
• Over-complex models on thin data. A sophisticated model fed poor data produces confident nonsense.
• Switching models constantly. Changing the model breaks comparability and invites cherry-picking.

A revenue attribution model is the rule that decides how credit for revenue is shared across touchpoints, and the choice profoundly shapes what your data seems to show. Chosen to fit your sales motion, applied consistently, and read as a directional lens rather than gospel, the right model turns attribution from a source of arguments into a guide for where to invest.