Last-touch Attribution

Run even two or three channels at once and a single purchase collects several touchpoints. Someone reads a blog post, clicks a search ad a few days later, and finally searches for the company by name and buys. When a $700 purchase arrives, the content marketer and the ad buyer each describe it as their result, and there is no number to put in the report.

That argument comes from a missing rule, not from missing data. The same record supports either claim until someone decides where credit goes. Last-touch attribution closes the question with the simplest rule available: everything goes to the last step. Nobody recalculates, and an ad platform can produce the report on its own.

Three steps.

1. Take one purchase and lay out that buyer's touchpoints in time order. A touchpoint is a single recorded contact — an ad click, a search visit, an email open.
2. Keep only the touchpoints inside the credit window. An attribution window is the deadline for how many days after a touchpoint still count, so a 30-day window drops a click from 40 days ago.
3. Give the full amount and the full conversion count to the latest remaining touchpoint. Everything earlier gets zero.

In the example above the blog post came 12 days out, the search ad 3 days out, and the branded search on the day of purchase, so the whole $700 lands on branded search. Most ad platforms use this rule as their default, which means a team that never changed a setting is already reading numbers calculated this way.

The limitation sits in the same place as the convenience. The last touchpoint is usually where someone who has already decided to buy happens to step. Branded search and retargeting come out looking stronger than they are, and whatever introduced the buyer in the first place scores zero.

"This report is last-touch, right? Then of course branded search is first."

The ranking is a product of the model, so it is weak ground for a budget decision.

"Blog posts stay at zero under last-touch, which makes the content budget hard to defend."

Only the closing touchpoint scores, which puts whoever works at the top of the funnel at a disadvantage.

"The default is last-touch, so let's confirm that before comparing the two tools."

Models have to match before a gap between two numbers means anything.

Ravi reports Pagedeck's monthly numbers on this basis, because the calculation is simple enough that the tools produce it unattended. When the budget gets reallocated, though, he attaches a first-touch table next to it, so that Grace does not read branded search sitting at the top as the whole story.