Multi-touch attribution
Aussi : MTA, Multi-touch attribution, Multi-touchpoint attribution, Fractional attribution, Attribution multi-touch, Attribution marketing multi-points de contact
A method that distributes conversion credit across all marketing touchpoints in the customer journey, rather than crediting only the first or last interaction.
What it is
Multi-touch attribution (MTA) is a measurement approach that assigns fractional credit for a conversion (a sale, signup, or lead) to each marketing touchpoint a customer interacted with along their journey. Instead of giving 100% of the credit to a single event, it spreads that credit across paid search, display, email, social, and other channels according to a chosen model.
Common crediting models include:
- Linear: equal credit to every touchpoint.
- Time decay: more credit to touchpoints closer to the conversion.
- Position-based (U-shaped): heavy credit to the first and last touches, less to the middle.
- Data-driven (algorithmic): credit derived statistically from observed conversion patterns, often using logistic regression, Markov chains, or Shapley values.
MTA contrasts with single-touch attribution (first-touch or last-touch) and with media mix modeling (MMM), which works at an aggregate level rather than the individual journey level.
Why it matters
- Budget allocation: it reveals which channels assist conversions, not just which one closed them.
- Fairer channel evaluation: upper-funnel activity (awareness) is not undervalued.
- ROI clarity: finance and marketing align on how spend maps to revenue.
Without MTA, last-touch bias tends to over-reward channels like branded search or retargeting that appear near the end of the funnel.
How it is used in practice
1. Collect journey data: stitch touchpoints per user across channels and devices.
2. Choose a model: rule-based for simplicity, data-driven for accuracy.
3. Assign fractional credit to each touchpoint.
4. Aggregate credited conversions and revenue by channel.
5. Act: reallocate spend, adjust bidding, and inform reporting.
Caveats: signal loss from cookie deprecation, privacy rules (GDPR, consent), and walled gardens limit journey visibility. Many teams now pair MTA with MMM and incrementality tests (holdouts) for a fuller picture.
Worked example
A customer converts on a 200 EUR purchase after four touchpoints:
- Facebook ad (first touch)
- Google display
- Branded search (last touch)
| Model | Facebook | Display | Email | Search |
|---|---|---|---|---|
| Last-touch | 0 | 0 | 0 | 200 |
| Linear | 50 | 50 | 50 | 50 |
| Time decay | 20 | 40 | 60 | 80 |
| U-shaped | 80 | 20 | 20 | 80 |
Last-touch credits search with everything, while MTA shows Facebook and email genuinely helped drive the sale. That insight changes where the next euro of budget should go.