Multi-touch attribution: understanding and choosing the right model

Why multi-touch attribution is a game changer

In "last click" attribution (the default for Google Analytics and most tools), all sales credit goes to the last touchpoint before conversion. It is like giving all credit for a goal to the scorer, ignoring the 15 passes that built the play.

Multi-touch attribution distributes credit among ALL touchpoints in the customer journey. Result: you understand the true contribution of each channel, campaign, and ad at every funnel stage.

Metrikia offers 9 attribution models : more than any other tool in this price range. Here is how to choose the right one.

The 9 attribution models in Metrikia

1. First Touch

How it works: 100% of credit goes to the first touchpoint.

When to use it:

• You want to understand which channels DISCOVER your brand
• Your goal is awareness and new prospect acquisition
• Short sales cycle (< 7 days)

Example: A prospect sees a Meta ad > clicks a Google ad > fills a form. First Touch gives 100% credit to Meta.

2. Last Touch

How it works: 100% of credit goes to the last touchpoint before conversion.

When to use it:

• You want to understand which channels CONVERT
• Your focus is bottom-of-funnel
• Direct comparison with Google Analytics (same model)

Example: Meta > Google > Form. Last Touch gives 100% to Google.

3. Linear

How it works: Credit is distributed equally among all touchpoints.

When to use it:

• You do not want to favor any touchpoint over others
• First exploratory analysis when you do not yet know your funnel
• Medium sales cycle with 3-5 touches

Example: Meta > TikTok > Google > Form. Each channel receives 33%.

4. Time Decay

How it works: The closer a touchpoint is to conversion, the more credit it receives. Recent touches count more.

When to use it:

• Long sales cycles (> 30 days)
• Recent actions are more deterministic in your business
• Time-limited promotions and offers

Example: Meta (D-30) > TikTok (D-14) > Google (D-1) > Conversion. Google receives ~60%, TikTok ~30%, Meta ~10%.

5. U-Shaped (Position-Based)

How it works: 40% to first touch, 40% to last touch, 20% distributed among intermediate touches.

When to use it:

• You value both discovery and conversion
• Good compromise for most businesses
• The default recommended model in Metrikia

Example: Meta (40%) > TikTok (10%) > Email (10%) > Google (40%).

Pro tip: U-Shaped is the best starting point if you do not know which model to choose. It balances acquisition and conversion well.

6. W-Shaped

How it works: 30% to first touch, 30% to lead creation, 30% to last touch, 10% distributed among others.

When to use it:

• Your funnel has a distinct "lead creation" moment (form, signup, demo request)
• B2B with clear qualification stages
• You want to understand the role of "mid-funnel"

7. Full Path

How it works: 22.5% to first touch, 22.5% to lead creation, 22.5% to opportunity creation, 22.5% to close, 10% distributed among others.

When to use it:

• Complex B2B sales cycles with multiple distinct stages
• Your CRM tracks stages: lead > MQL > SQL > Opportunity > Deal
• You want a complete funnel vision

8. Markov Chain

How it works: Probabilistic model that calculates each channel's contribution by simulating what would happen if it were removed from the journey. Based on observed transitions between channels.

When to use it:

• You have sufficient conversion volume (100+ per month recommended)
• You want a data-driven model, not based on arbitrary rules
• Quarterly strategic analysis of your channels

Unique advantage: Unlike rule-based models, Markov adapts to YOUR data. If in your business, TikTok is always followed by Google before conversion, the model detects it automatically.

9. Shapley Value

How it works: From game theory, this model calculates the marginal contribution of each channel by considering all possible channel coalitions. It is the most mathematically fair model.

When to use it:

• You want the fairest and most sophisticated model
• Sufficient data volume (100+ conversions)
• Strategic annual budget allocation decisions

Unique advantage: Shapley is the only model that mathematically guarantees that the contribution attributed to each channel is proportional to its true added value.

How to configure attribution in Metrikia

1. Go to Attribution in the main menu
2. Select the attribution model from the selector in the top right
3. The view refreshes instantly with recalculated credits
4. Compare multiple models by switching between them
5. Use date filters to analyze different periods

Comparing models

Metrikia lets you see side by side how different models attribute credit. This reveals channels "undervalued" by last click:

• If a channel scores well in First Touch but not Last Touch, it is a discovery channel
• If a channel scores well in Time Decay and Last Touch, it is a conversion channel
• If a channel is strong in Shapley but weak in Last Touch, it plays a crucial assist role

Recommendations by business type

Pro tip: Start with U-Shaped, then switch to Shapley or Markov when you have 3+ months of data and 100+ conversions. Ask Diana "Compare U-Shaped vs Shapley for last quarter" for an automatic summary of differences.