Building Multi-Touch Attribution Models That Actually Work

After building attribution systems for companies managing £5M-£16M in marketing spend, I’ve seen what works and what creates confusion. This guide covers how to implement attribution that actually improves decisions.

Why Last-Click Attribution Fails

Last-click gives 100% credit to the final touchpoint before conversion. This creates systematic problems:

• Brand search gets inflated credit: Someone searches your name after seeing five ads; brand search gets all credit
• Awareness channels get starved: Display and social that initiate journeys get zero credit
• Remarketing appears miraculous: It’s capturing demand, not creating it
• Budget allocation suffers: You cut what’s actually working and double down on closers

The Real Impact A B2B company using last-click nearly eliminated content marketing, attributing only 3% of conversions to organic content. Multi-touch analysis revealed content actually influenced 67% of pipeline, just not as the last click.

Multi-Touch Attribution Models Explained

Linear Attribution

Gives equal credit to every touchpoint in the journey.

• Pros: Simple, values all channels
• Cons: Doesn’t reflect reality: not all touchpoints matter equally
• Best for: Starting point when moving from last-click

Time Decay Attribution

Gives more credit to touchpoints closer to conversion.

• Pros: Recognises recency matters
• Cons: Still undervalues awareness channels
• Best for: Short purchase cycles

Position-Based (U-Shaped)

40% to first touch, 40% to last touch, 20% split among middle.

• Pros: Values both acquisition and conversion
• Cons: Arbitrary percentage splits
• Best for: Most B2B and considered purchases

Data-Driven Attribution

Uses machine learning to determine actual impact of each touchpoint.

• Pros: Most accurate when you have enough data
• Cons: Requires significant conversion volume
• Best for: High-volume accounts with clean tracking

Implementation Framework

Step 1: Fix Your Tracking

Attribution is only as good as your data:

• UTM consistency: Standardise UTM parameters across all channels
• Cross-device tracking: Connect journeys across mobile/desktop
• Offline conversion import: Connect CRM data back to marketing touchpoints
• Consent management: Ensure tracking respects privacy regulations

Step 2: Choose Your Model

Match model to your business reality:

• Short sales cycles (days): Linear or time decay
• Long sales cycles (weeks/months): Position-based
• High volume (1000+ conversions/month): Data-driven
• Complex B2B: Custom models with weighted touchpoints

Step 3: Build Reporting

Create reports that enable action:

• Channel contribution: How much does each channel influence conversions?
• Path analysis: What are common conversion paths?
• Assisted conversions: Which channels assist but don’t close?
• Model comparison: How do results differ by attribution model?

Step 4: Act on Insights

Attribution should change decisions:

• Budget reallocation: Shift spend to channels with true contribution
• Channel investment: Fund undervalued awareness channels
• Campaign optimisation: Focus on what moves prospects through the journey
• Cross-channel coordination: Optimise the full funnel, not individual channels

Common Attribution Challenges

Challenge 1: Walled Gardens

Facebook, Google, and LinkedIn each claim credit their own way:

• Use a third-party attribution tool for independent measurement
• Understand each platform’s attribution methodology
• Don’t expect platform numbers to match your analytics

Challenge 2: Long Sales Cycles

B2B purchases can take months:

• Extend your attribution window appropriately
• Track micro-conversions along the journey
• Connect CRM data for full-funnel visibility

Challenge 3: Offline Touchpoints

Events, calls, and meetings matter but are hard to track:

• Implement call tracking with source attribution
• Tag event attendees and connect to marketing campaigns
• Use CRM integration to capture sales touchpoints

Practical Advice Start simple. A basic position-based model with clean tracking beats a sophisticated model with messy data. Get the fundamentals right first, then add complexity as your data quality improves.

Measuring Attribution Success

Good attribution implementation should:

• Change your budget allocation: If nothing changes, attribution isn’t informing decisions
• Reveal hidden value: You should discover channels performing better than last-click showed
• Improve overall ROAS: Better allocation should improve return
• Reduce waste: Identify channels that looked good but don’t actually contribute