Attribution & Conversion Counting

Position-Based Attribution vs Time-Decay Attribution

In short: Both models move past the one-touch-gets-everything logic of last-click, but they reward different moments in the journey. Position-based fixes 40 percent on the first touch and 40 percent on the last, treating discovery and close as the two moments that matter; time-decay gives progressively more credit the closer a touch sits to the conversion, with no special regard for how the journey began. Neither is available in Google Ads or GA4 anymore, both were retired in 2023. Use position-based when you can defend how we found them and what closed them as the two moments worth rewarding; use time-decay when you believe recency itself, not position, is what signals real intent.

By the AdFlint research team · Fact-checked against current Google and Meta platform behavior · Last reviewed July 2026

Position-Based Attribution

Assigns forty percent each to the first and last touchpoints and splits the remaining twenty percent across everything in between.

Also called U-shaped, this model rewards the two moments most advertisers care about, discovery and close, while acknowledging the middle without letting it dominate. It fits multi-step B2B paths where an early content touch and a final branded search both clearly matter. Google Ads and GA4 retired it in 2023. The weakness is that the weights are arbitrary conventions, not estimates derived from how your customers actually behave.

Full definition

Time-Decay Attribution

Weights touchpoints by recency, so interactions closer to the conversion receive exponentially more credit than ones earlier in the path.

Credit decays on a half-life, commonly seven days, so a touch a week before the conversion carries half the weight of one on the closing day. It suits long consideration cycles where recent activity signals intent. Google Ads and GA4 dropped it from reporting in 2023. The misunderstanding is assuming the half-life is tuned to your business; it is a fixed default, not something the model learns from your data.

Full definition

Side by side.

The differences that actually change what happens in your account.

	Position-Based Attribution	Time-Decay Attribution
What earns the biggest share	The first touch and the last touch, fixed at 40 percent each no matter what happened between them.	Whichever touch sits closest to the conversion date, regardless of whether it was first, last, or somewhere in the middle of a repeat path.
How a path's first touch is treated	Guaranteed 40 percent, even if it happened months before conversion.	Can shrink to almost nothing if it happened long before the decay window, no guarantee at all.
How a touch right before conversion is treated	Falls into the shared 20 percent bucket if it is not the very first or last touch.	Can receive a large share if it is the touch closest to conversion; position in the path does not matter, only timing.
Configurability	None in the standard preset, 40/40/20 is fixed.	A half-life you can typically adjust, though most tools ship a default, often seven days, that is rarely revisited.
Best-fit journey shape	Clear discovery-to-close funnels, especially considered B2B purchases with a defined start and finish.	Journeys with repeat engagement where the most recent interaction is the best signal of buying intent.
What it ignores	The order and timing of everything except the very first and very last touch.	Where in the funnel a touch happened; a first touch and a fifth touch are judged purely by how recent they are.
Failure mode	A coincidental first or last touch, like a stray direct visit, can lock in 40 percent it did not earn.	A short but genuinely important early touch gets buried if it falls outside the decay window, even in an otherwise short path.

What actually separates them.

Position-based decides credit by where a touch sits in the sequence, first, last, or middle; time-decay decides credit purely by how many days separate it from the conversion, ignoring sequence.

In position-based, first touch always gets 40 percent no matter its age; in time-decay, an old first touch can be discounted to near zero regardless of its structural importance.

Position-based's split total, 80 percent to the two anchors, does not change with path length; time-decay's distribution changes shape depending on how spread out in time the touches are.

Time-decay has a tunable half-life parameter that directly changes results; position-based's 40/40/20 preset has no equivalent dial.

A path where the first and last touch happen on the same day collapses time-decay's timing advantage, since everything is recent, while position-based still rigidly reserves 40 percent for each of those two touches even if they are nearly simultaneous.

Which one should you use?

Use Position-Based Attribution when

You can point to a specific first-touch channel, like a webinar or content download, and last-touch channel, like a demo request or branded search, that your sales team already treats as the meaningful moments.
Your funnel has a clear beginning and end with a genuinely different middle stage you want to de-emphasize.
You want the model's assumptions to be easy to explain to a non-technical stakeholder in one sentence.
Path length varies a lot across customers and you do not want that variance to change how much credit the anchors get.

Use Time-Decay Attribution when

Your buyers re-engage multiple times close together right before converting, and that clustering is a real signal.
You do not have a clean first-touch-equals-discovery story; your paths are noisy and recency is the more reliable signal you have.
You are willing to pick and defend a half-life that matches your actual sales cycle length, not just accept a default.
You want a channel that only ever shows up right before conversion, like remarketing or branded search, to be rewarded for that positioning.

Common questions.

Which model gives more credit to remarketing, position-based or time-decay?

Time-decay, in most cases, because remarketing typically fires close to the conversion and time-decay rewards recency directly. Position-based only rewards it heavily if it happens to be the very first or very last touch in the path, which remarketing often is as the last touch, but it will not get a boost just for being recent if there is another touch after it.

Can I combine position-based and time-decay logic?

Not as a built-in preset in either Google's tools, which no longer offer either, or most third-party platforms, but some attribution vendors let you build a custom weighted model that blends positional and recency factors. Outside of a custom build, you pick one rule set or the other.

My conversion path is only two touches long. Does the model choice even matter here?

Less than you would think. On a two-touch path, position-based reduces to roughly a 40/40 or 50/50 split depending on the tool's edge-case handling, and time-decay will heavily favor whichever touch happened closer to conversion. The bigger the gap in timing between the two touches, the more they will disagree.

Or stop choosing between them.

AdFlint picks the setting, writes the ads, and keeps optimizing inside the Google and Meta accounts you already own.

Try AdFlint free See how it works

Related comparisons

All comparisons