Attribution
The methodology connecting marketing touchpoints to revenue outcomes, so investment can be allocated on evidence rather than on preference.
What it does and does not establish
Attribution distributes credit for conversions that occurred, under a rule you choose. It is routinely read as an estimate of what each channel caused. Those are different quantities and the difference is not a nuance — it determines whether the output can support a budget decision.
Credit and effect
The gap between them is selection. Channels appearing before conversions are often channels that people already intending to convert use — branded search being the clearest case. No credit rule applied to observational data separates the two.
The four choices inside any model
| Choice | Typical unexamined default | Effect of getting it wrong |
|---|---|---|
| Lookback window | Inherited from a tool | Systematically favours short or long cycles |
| Eligible touchpoints | Whatever the platform tracks | Untracked channels score zero by construction |
| Credit rule | Last touch, or a vendor default | Determines the ranking more than the data does |
| Conversion definition | Form fill or closed-won | Optimises for the wrong end of the funnel |
Source: Framework stated here
What attribution is good for
Describing journey structure — how many touches, over what elapsed time, in what order.
Detecting change under a fixed, documented rule. Movement then means something.
Comparing segments, whose journeys genuinely differ in shape.
Generating hypotheses to test — treating the model as a pointer rather than an answer.
When to run an experiment instead
Where the decision is material — a large budget shift, a channel exit — hold the channel out for a period exceeding your sales cycle, for a randomisation unit you control, and measure the outcome that matters rather than the proximate one. The gap between the attributed figure and the measured effect is the most useful number the exercise produces, because it calibrates how far to trust attribution everywhere else.
RELATED TERMS
Cost Per Lead (CPL)
Marketing spend divided by leads generated. The most immediately available demand-gen metric and the easiest to misread.
Customer Acquisition Cost (CAC)
The fully loaded cost of acquiring one new customer: sales and marketing spend for a period divided by new customers acquired from it.
Dark Funnel
The buying activity that happens where analytics cannot observe it — private communities, peer conversations, group chats, word of mouth.
Marketing Sourced Pipeline
Pipeline where marketing originated the first meaningful contact with the account, according to a documented rule.
COMMON QUESTIONS
- What is marketing attribution?
- A rule for dividing credit for a conversion across the touchpoints that preceded it. It describes observed sequences under a stated rule — it does not measure what each channel caused.
- What is the difference between attribution and incrementality?
- Attribution divides credit for conversions that happened. Incrementality asks how many would not have happened without the channel. Only the second answers a budget question, and it generally requires a holdout experiment.
- Which attribution model is best?
- The question is under-specified without the decision it supports. Every model encodes four choices — lookback window, eligible touchpoints, credit rule and conversion definition — and different decisions call for different settings.
- Why did our attribution numbers change without any campaign change?
- Usually one of the four embedded choices moved: the window, which touchpoint types count, the credit rule, or the conversion definition. Undocumented model settings are the most common cause of results that appear to shift on their own.
FURTHER READING
Learn how to apply this: RevOps 101: Revenue Operations Foundations
Definitions are the vocabulary. The courses are where you learn to operate it, with the interactive audit tools.
See the course