A multi-touch attribution model that assigns increasing credit to marketing touchpoints as they get closer in time to the conversion event, based on the idea that recent interactions have more influence on the final decision.
Marketers or analysts reviewing marketing attribution reports alongside other modeling techniques.
01What it is and how it works
Time Decay Attribution uses a mathematical decay function — often exponential — to weight each touchpoint. The weight decreases the further back in time the touchpoint occurred from the conversion. A common implementation sets a half-life parameter, say 7 days: a touchpoint 7 days before conversion gets half the weight of one that happened on the conversion day. Touchpoints beyond that get progressively less. The model sums the weights across all touchpoints in a path and distributes 100% of the conversion credit proportionally. For example, if a user sees an ad 14 days before purchase, then clicks a search ad 2 days before, the search ad gets significantly more credit. This model is often used when the sales cycle is long and the last few interactions are considered decisive.
It means that the last few clicks or ads before a purchase get most of the credit, while earlier ones get less, using a decay function like a half-life.
02What to do about it
First, set the half-life to match your actual sales cycle. If most conversions happen within 3 days of the last touch, use a short half-life (e.g., 1 day). For longer cycles, use 7 or 14 days. Second, compare Time Decay results with other models (e.g., linear, position-based) in your analytics platform to see if it changes budget allocation. Third, apply it to campaigns where recency matters — like retargeting or email drip sequences. Fourth, use it in combination with offline conversion tracking to avoid over-crediting last digital clicks. Finally, document the chosen half-life and review it quarterly as customer behavior shifts.
03How it is measured or noticed
You see Time Decay Attribution in action by looking at attribution reports in tools like Google Analytics or Google Ads. The report shows each channel’s attributed conversions and revenue under this model. Compare the share of credit for early-stage channels (e.g., display, social) vs. late-stage channels (e.g., branded search, direct). If early channels get very low credit, the decay is too aggressive. You can also compute the average time lag between first touch and conversion; if it’s long, Time Decay may underweight top-of-funnel efforts. A quick check: run a path analysis and see the weight distribution across touchpoints in a 30-day window.
04Common mistakes
- Using the default half-life without analyzing your actual conversion time lag.
- Applying Time Decay to short sales cycles (e.g., same-day purchases) where it behaves almost like last-click.
- Ignoring offline touchpoints — if a user sees a TV ad then clicks a search ad, the model only credits the digital touch.
- Not adjusting for multiple conversions from the same user; Time Decay can over-credit recent touchpoints across different journeys.
- Assuming Time Decay is always better than linear — it depends on whether recency truly drives conversions.
05Limits
Time Decay Attribution is not suitable for brand awareness campaigns where early exposure is critical. It also fails when the sales cycle is extremely short (e.g., impulse buys) because it collapses to near last-click. It is often confused with 'last non-direct click' but it gives some credit to earlier touchpoints. The model assumes a monotonic decay, but real influence may be non-linear (e.g., a touchpoint 3 days ago might be more influential than one 1 day ago if it was a key piece of content). It also does not account for cross-device or view-through conversions unless explicitly tracked. For B2B with multiple decision-makers, Time Decay on a single user path misses group dynamics.
06Worked example
A user first clicks a display ad 20 days before purchase, then an email link 10 days before, then a branded search ad 2 days before, and finally converts. With a 7-day half-life, the weights (relative to the last touch) are: display: 2^(-20/7) ≈ 0.14, email: 2^(-10/7) ≈ 0.38, search: 2^(-2/7) ≈ 0.82, last touch (direct or search? assume search is last): 1.0. Normalized: total = 2.34, so search gets 1.0/2.34 ≈ 43%, email gets 16%, display gets 6%. The remaining 35%? Actually the last touch is the conversion itself? Typically the conversion is a separate event. Let's adjust: if the last touch is the search click, then the conversion gets no credit. So the three touchpoints: display 0.14, email 0.38, search 0.82. Sum=1.34. Search gets 61%, email 28%, display 10%. This shows how recency dominates.
Frequently asked questions
How is Time Decay Attribution different from Last Click Attribution?
Time Decay Attribution gives partial credit to multiple touchpoints, with more weight to recent ones, while Last Click Attribution gives 100% credit to the final touchpoint. Time Decay is better for understanding the influence of earlier interactions, especially in longer sales cycles.
Should I use Time Decay Attribution for a long sales cycle?
It depends. Time Decay Attribution works well for sales cycles where recent interactions are more influential, but if early touchpoints are critical for awareness, it may undervalue them. Consider using a custom decay rate or a different model for very long cycles.
How do I set up Time Decay Attribution in Google Analytics?
In Google Analytics, you can select Time Decay as the attribution model in the Model Comparison Tool or apply it to conversion paths. You can also adjust the half-life parameter to match your sales cycle.
Does Time Decay Attribution still work if my customers often convert after a long delay?
It can still work, but the decay function will heavily discount early touchpoints. If early exposure is crucial, you might need a longer half-life or a different model like Linear Attribution.
What happens if I use Time Decay Attribution for brand awareness campaigns?
Time Decay Attribution will undervalue early awareness touchpoints, making it seem like they have little impact. This can lead to underinvestment in brand-building activities. Use a model that gives more credit to first interactions instead.
How long does it take for Time Decay Attribution to show reliable results?
Reliability depends on the volume of conversions and the length of your sales cycle. You need enough data to see patterns across touchpoints. Typically, a few weeks to a month of data can start showing trends, but it's best to compare with other models.
Asked out loud
spoken, not typedThe same term in the words somebody uses speaking to an assistant rather than typing into a box — written from the situation, which is why each one carries the situation it came from.
That's because the report uses a time decay model, which gives more weight to touchpoints closer to conversion. If you want to see earlier contributions, try a linear or position-based model.
You should use a multi-touch attribution model like time decay if your sales cycle has multiple interactions and recent ones are more influential. But if early awareness is key, consider a first-click or U-shaped model.
Tell them that last-click ignores all earlier touchpoints, while time decay gives a more balanced view by crediting recent interactions more. It helps us understand the full customer journey.