Multi-touch attribution
Also: 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.
See also
Frequently asked questions
What is multi-touch attribution?
Multi-touch attribution (MTA) assigns a fraction of the credit for each conversion to every marketing touchpoint a customer interacted with, instead of handing 100% to a single interaction. A 200 EUR sale preceded by a Facebook ad, a display impression, an email and a branded search click gets split across all four rather than credited entirely to search. The split follows a chosen model: linear, time decay, position-based or data-driven.
What is the difference between multi-touch attribution and media mix modeling?
MTA works at the level of the individual customer journey and needs touchpoint-level data stitched per user; media mix modeling (MMM) works on aggregate spend and outcome data without tracking individuals. Because cookie deprecation, consent rules and walled gardens degrade journey visibility, many teams run both, plus incrementality tests with holdout groups.
Who needs to understand multi-touch attribution, beyond the marketing team?
Finance and data leaders as much as marketing. MTA determines how spend is mapped to revenue, so it drives budget allocation arguments between the CFO and the CMO, and it depends on the data teams who stitch journeys across channels and devices. It is also relevant to anyone building algorithmic models, since data-driven attribution relies on logistic regression, Markov chains or Shapley values.
Which multi-touch attribution model should you start with?
Start with a rule-based model: linear gives equal credit to every touchpoint, time decay favours touchpoints closer to the conversion, and position-based (U-shaped) loads credit onto the first and last touches. They are simple to explain and to implement. Data-driven models are more accurate but require enough conversion volume and clean journey data to be statistically meaningful.
Why does last-touch attribution distort budget decisions?
Last-touch gives all the credit to the final interaction, which systematically over-rewards channels sitting near the end of the funnel such as branded search and retargeting, while awareness activity looks worthless. In the four-touch example above, last-touch credits the full 200 EUR to branded search; a linear model shows 50 EUR each for Facebook, display, email and search. The two views point the next euro of budget in opposite directions.