Skip to content

Time-Decay Attribution

Time-Decay distributes conversion value across all touchpoints, with bigger weights for interactions closer to conversion. It captures the reality that recent touches usually push the decision more than ones from weeks ago.

This matters for short-cycle scenarios: limited-time promotions, impulse buys.

How It Works

Half-Life

Time-Decay uses half-life, borrowed from physics. It defines how long it takes for a touchpoint's influence to drop by half.

Mathematical Model

Touchpoint weight formula:

Weight = 2^(-t/half-life)

where t is days between touchpoint and conversion, half-life is your configurable parameter.

Weight Distribution

Temporal sensitivity controls how strongly time affects credit. Touches just before conversion get the most weight. Earlier ones get proportionally less.

Decay function smooths the transition. No sudden jumps between touchpoints.

Weight Example

7-day journey:

  • Day 1: Google Ads (search) → 10%
  • Day 3: Facebook (organic) → 20%
  • Day 6: Email campaign → 30%
  • Day 7: Direct visit → 40%

Half-life: 7 days

A touch 7 days before conversion gets half the weight of a same-day touch.

Compared to Other Models

Time-Decay vs Linear

Linear gives every touch the same value. Time-Decay weights recent touches higher. Linear assumes touches matter equally; reality often disagrees.

Characteristics:

  • Equal value distribution
  • Simple to understand and implement
  • Ignores timing
  • Good for overall channel contribution

Characteristics:

  • Recent interactions get priority
  • Accounts for proximity to conversion
  • Configurable half-life
  • Ideal when urgency matters

Time-Decay vs Position-Based

Position-Based hard-codes 40% to first and last, 20% across middle. Time-Decay distributes more smoothly without fixed percentages.

Setup

Pick a Half-Life

Match it to your sales cycle. Wrong half-life kills accuracy.

Half-Life Guidance

Short cycles (B2C):

  • E-commerce: 1-3 days
  • Impulse buys: 12-24 hours
  • Service subscriptions: 3-7 days

Long cycles (B2B):

  • SaaS: 14-30 days
  • Enterprise sales: 30-90 days
  • Consulting: 7-21 days

Lookback Window

The window defines how far back to look for touchpoints. Several types exist, like visit lookback (start of session containing the conversion).

Lookback TypeDescriptionUse Case
Visit-basedSession start to conversionIntra-session behavior
Time-basedFixed period (7, 14, 30 days)Standard marketing analysis
CustomConfigurableSpecific business cases

Technical Implementation

Setup:

  1. Conversion Settings → Attribution Models
  2. Pick Time-Decay
  3. Configure lookback window (default 90 days)
  4. Apply to your conversion actions

Configuration:

Exponential decay with customizable half-life, default 7 days.

  • Attribution Components → Time Decay
  • Set custom half-life
  • Configure lookback window

Steps:

  • Collect touchpoint data
  • Calculate temporal intervals
  • Apply 2^(-t/halflife)
  • Normalize weights to 100%

Urgency Modeling

Adapting for Urgent Campaigns

Time-Decay shines when conversions need to land fast. Tune the model for time-pressured scenarios.

Promotional Settings

For 24-48 hour campaigns:

  • Drop half-life to 1-2 days
  • Lookback window to 7 days
  • Boost weight on last 24 hours
  • Exclude touches older than a week

Seasonal Adjustments

Holiday periods shift consumer behavior. Reflect that in your parameters.

Adjustments by period:

  • Black Friday: half-life 6-12 hours
  • New Year holidays: half-life 2-3 days
  • Back-to-school season: half-life 3-7 days
  • Regular periods: standard settings

Industry Applications

E-commerce and Retail

Time-Decay helps optimize campaigns by surfacing recent high-impact touches.

graph TD
    A[Search Advertising<br/>-7 days: 5%] --> B[Email Campaign<br/>-3 days: 15%]
    B --> C[Retargeting<br/>-1 day: 35%]
    C --> D[Direct Visit<br/>Conversion Day: 45%]
    D --> E[Purchase $100]

SaaS and B2B

Long cycles with networks of interrelated touchpoints. Time-Decay reads them well.

B2B Specifics

Configuration considerations:

  • Multiple stakeholders per decision
  • 30-180 day cycles
  • Account-based marketing
  • Cross-device interactions

Lead Generation

Time-Decay helps optimize for quality leads. Surfaces channels that close.

Benefits and Limitations

Benefits

Realistic behavior model: Recent touches get proportionally bigger credit, reflecting their actual influence.

Configuration flexibility: Half-life adapts to your cycle.

Balanced credit: Unlike single-touch, every interaction gets considered, weighted by timing.

Limitations

Heavily dependent on accurate, complete touchpoint data. Gaps distort results.

Main Limits

Technical:

  • Sophisticated tracking required
  • Multi-source data integration is complex
  • Cross-device interactions are hard
  • Offline interactions are harder

Analytical:

  • Half-life selection is subjective
  • May undervalue awareness
  • Needs large data volumes for significance

Analyzing Results

Reading Reports

Reports show how much credit each touchpoint receives by proximity.

Key metrics:

  • ROI/ROAS by channel, weighted by Time-Decay
  • Cost per Acquisition with temporal adjustment
  • Customer Lifetime Value by source
  • LTV:CAC ratio for long-term effectiveness

Segmentation

Mobile vs Desktop:

  • Mobile: shorter decision cycles
  • Desktop: deeper research
  • Cross-device: attribution complexity

Regional differences:

  • Developed markets: longer consideration
  • Emerging markets: more impulsive
  • Local vs international campaigns

Generational:

  • Gen Z: shorter attention
  • Millennials: multi-channel
  • Gen X: more thorough research

Validate Against Other Models

Segment by source, campaign, device, demographics. Compare against other models.

Optimization

Budget Planning

Channels with high Time-Decay weight deserve more investment.

Strategies:

Increase budget for high-impact channels:

  • High Time-Decay weight
  • Short time-to-conversion
  • Strong urgency signal

Pull back from low-impact channels:

  • Low weight in final stage
  • Awareness-only sources
  • Long lag between touch and conversion

Creative Optimization

Adapt messages by proximity:

  • Early-stage: Brand awareness, educational content
  • Mid-stage: Product benefits, case studies
  • Late-stage: Urgency, limited-time offers

Timing

ChannelOptimal Touch TimingUrgency Level
Email2-24 hours before conversionHigh
Paid Social1-3 days before conversionMedium
Organic Search1-7 days before conversionLow
Display Retargeting6-48 hours before conversionVery High

Technical Aspects

Data Requirements

Quality data is foundational. Bad tracking, bad insights.

Minimum:

  • Customer ID for linking
  • Timestamp on every interaction
  • Source/Medium classification
  • Conversion events with exact timestamps
  • UTM parameters

Cross-Device Tracking

Strategies:

  • Deterministic matching: Email, user ID
  • Probabilistic matching: Device fingerprinting
  • Hybrid approach: Combined
  • First-party data: CRM integration

API Integrations

graph TD
    A[GA4 API] --> E[Attribution Engine]
    B[Facebook API] --> E
    C[Email Platform API] --> E
    D[CRM API] --> E
    E --> F[Time-Decay Model]
    F --> G[Attribution Reports]
    F --> H[Budget Optimization]
    F --> I[Campaign Insights]

What's Next

Machine Learning

Modern platforms automate Time-Decay parameter tuning. Data-Driven Attribution determines real channel contribution from historical data.

Directions:

  • Automatic half-life optimization
  • Predictive attribution
  • Real-time model adjustment
  • Cross-industry benchmarking

Privacy-First

  • Server-side tracking to bypass blockers
  • First-party data focus over third-party cookies
  • Probabilistic modeling to fill gaps
  • Cookieless attribution through advanced fingerprinting

Omnichannel Attribution

Combining offline and online touches into a unified Time-Decay model is becoming standard.

We're building tooling that combines digital, phone, and in-store interactions in one Time-Decay view. Plans include automatic urgency modeling that tunes parameters to campaign, period, and audience.

About AI participation in writing articles

This article, like many others on our site, was created, written and proofread by a team of developers. Of course, not without the participation of AI assistants. We don't hide this and believe that modern systems are already quite good at handling simple tasks and, relatively speaking, writing an article about Viewport yourself is quite strange. It won't come out significantly better and will take a lot of time. But providing basic understanding to beginner webmasters is necessary. Of course, after the article is written by assistants - there's always proofreading, and this is where not one or two people participate, and only after that the article is published.

Ready to Implement Time-Decay Attribution?

Sign up for our free trial and access advanced attribution modeling, including configurable Time-Decay with urgency optimization.


Ready to take control of your web analytics? Try Statable free for 30 days — no credit card required, full feature access, GDPR-compliant by default. Start your free trial or view a live demo.