Attribution and Contribution: The Key to Advertiser Success

In digital marketing, where customer purchase journeys are increasingly fragmented and complex, advertisers need a solid grasp of attribution, contribution and deduplication. These analytical methods are essential to accurately assess campaign performance with a full-funnel approach, optimize marketing budgets and, ultimately, maximize return on investment (ROI). This article takes an in-depth look at these concepts, illustrated with concrete examples from the TrackAd marketing attribution solution.

Attribution: Breaking Down the Customer Journey in Depth

Attribution is at the heart of understanding the customer journey. It determines which touchpoints in a conversion path should receive credit for a given conversion. Attribution falls into two main categories: multi-touch attribution and single-touch attribution, also known as deduplication.

Multi-Touch Attribution

Multi-touch attribution distributes conversion credit across all the touchpoints in a user journey. It is essential for giving you a complete picture of how effective your various interactions are throughout the customer journey. There are several multi-touch attribution models, each with its own characteristics and specific use cases.

Linear Model:

This model splits conversion credit equally across every touchpoint in the journey. For example, in a conversion path with five touchpoints, each one receives 20% of the credit, or 0.2 conversion. This model is easy to implement and useful for highlighting the cross-journey impact of your sales-pressure campaigns, such as Performance Max, retargeting and affiliate marketing.

Time Decay Model:

This model gives more credit to touchpoints that occur closer in time to the final conversion. Interactions that happened just before the conversion are therefore considered more influential than those that occurred earlier in the journey. This model is rarely used, but it can be valuable for channels whose touchpoints gain influence as the user gets closer to converting, mainly bottom-of-funnel channels (Performance Max, branded search, retargeting, affiliate marketing) or activation campaigns targeting visitors to your editorial content.

Markov Chain Model:

More advanced, this model uses probabilities to calculate the impact of each touchpoint based on its ability to generate further interactions in the conversion path. For example, by analyzing the sequence of touchpoints, the model can determine which source contributed most to the conversion, taking into account the alternative paths that might have been followed without a given interaction. This model is especially useful for complex campaigns where interplay between channels plays a key role in conversion, mainly mid- and top-of-funnel campaigns (branding, SMA, display / RTB).

Data-Driven Model:

Based on the Shapley value method, this model assigns conversion credit according to each touchpoint's actual contribution, calculated from historical data. It is a powerful model because it uses all available data to assess the relative impact of each touchpoint in a given context. The Shapley value method is particularly effective at modeling contributions when multiple touchpoints interact in complex ways within the conversion path. It is a complex model whose results are close to those of the Markov chain.

Data-Driven and Markov chain models are appealing because they give due credit to mid- and top-of-funnel channels, which are all too often penalized by the last-click models of analytics solutions. However, they suffer from a "black box" effect: their complexity forces you to accept their results without really understanding how they are calculated.

Single-Touch Attribution, or Deduplication

Single-touch attribution, or deduplication, focuses on a single touchpoint and gives it 100% of the conversion credit. It is used when the goal is to simplify analysis by highlighting one specific touchpoint in the customer journey. These models are factual, delivering objective results, in contrast to the output of Data-Driven and Markov models.

First Click:

This model gives all the conversion credit to the first touchpoint in the journey. It is particularly useful for identifying the channels that play an important role in the initial acquisition of future customers. It gives due credit to mid-funnel and upper-funnel channels whose goal is to win new customers (branding, SMA, display / RTB). For a complete analysis, this model should be combined with the channel's view-through (post-view) conversion impact.

Last Click:

Very widespread, this model gives all the conversion credit to the last touchpoint before the conversion. It is misleading, however, because it ignores the impact of earlier interactions that may have paved the way for the conversion. It is only truly relevant for very bottom-of-funnel channels, such as cashback or promo code sites.

Last Paid Click:

This model gives all the credit to the last paid touchpoint before the conversion, deprioritizing unpaid channels such as organic traffic. It is particularly well suited to bottom-of-funnel channels (Performance Max, branded search, retargeting, affiliate marketing) to highlight their impact as sales-pressure levers.

Post Click:

This model is used by every media platform in its interface (Google Ads, Meta Business Manager, etc.). It attributes the conversion to a source present in the conversion path when the time between the click and the conversion is shorter than the cookie duration (the "cookie time," which is 7 days for Meta, for example). Any other channels in the conversion path between that click and the conversion have no influence on the result. Since media platforms have no access to other channels' touchpoints, they are inherently limited to this model.
Post-click deduplication is also occasionally used by affiliate platforms to settle commissions between affiliates.

TrackAd lets you compare the results of different attribution models, so you can apply the right model to each channel based on its profile: top, middle or bottom of funnel / branding or performance.
Limiting yourself to a single attribution model would inevitably disadvantage some channels and undermine the objectivity of your decision-making.

Contribution: Understanding Where Each Channel Fits in the Conversion Path

Unlike attribution, contribution analysis involves no model. It is a factual account of the value each channel or campaign adds to the conversion path. This analysis captures not only direct interactions but also the supporting role certain touchpoints play in the conversion process. The goal of contribution analysis is to identify gaps in your media mix across lower-funnel, mid-funnel and upper-funnel channels so you can rebalance them for stronger performance.

Segmentation by channel role

To understand the impact of each traffic channel, you need to consider all of its participations, meaning every order for which it was part of the conversion path.
The analysis then counts those participations as conversion path Initiations (first click), Closings (last click), Support (neither first nor last click) and Autonomous (both first and last click).
This segmentation is essential for understanding how each channel contributes to overall campaign performance.

Initiation:

Channels that initiate conversion paths are essential for capturing the attention of future new customers. Branding, SMA and display / RTB channels usually play this key role by introducing a user to a brand for the first time.

Closing:

Closing channels are the ones that complete conversion paths, turning purchase intent into an actual purchase. They complement initiating channels. Retargeting and cashback campaigns are typical closing channels that capitalize on interest the user has already shown.

Autonomous:

Some channels are autonomous, meaning they both initiate and complete the conversion path. The classic example is a promotional Facebook campaign leading to a dedicated landing page that not only draws the user to the site but also drives the conversion without any other channel stepping in.

Support:

Support channels come into play between initiation and closing. They keep your prospects engaged and help them move down the conversion funnel. For example, a series of marketing emails or retargeting ads on social media can play this role by reminding prospects of products or providing additional information to win them over.

TrackAd lets you visualize these roles in terms of both traffic and visibility. Filters let you analyze contribution among your new customers and your existing customers, to distinguish each channel's contribution to acquisition and to retention.

Analyzing the balance between channels helps advertisers make more strategic resource allocation decisions. Take a channel such as SMA that shows poor last-click ROI in the analytics solution. If contribution analysis showed that this channel played a strong initiating role among new customers, and its first-click ROI turned out to be good, it would make sense to increase investment in it to introduce your site to as many future new buyers as possible.
The same would apply to a source with a strong support profile that accompanies many initiating and autonomous channels. It would be worth keeping to help convert the visitors those channels bring in, making the whole mix profitable.

Conclusion: The Importance of a Holistic, Flexible Approach

Attribution and contribution are essential components of marketing performance analysis. By mastering them, you can not only gain a finer understanding of how effective your media mix strategy is across all your customer journeys, but also significantly increase the revenue generated by your digital acquisition.

In a digital environment where purchase journeys span multiple channels and devices, this technical, holistic approach has become essential to maximize return on investment and ensure the success of marketing strategies.

For advertisers looking to get the most out of their campaigns, TrackAd provides the tools and expertise needed to navigate this constantly evolving landscape effectively.

TrackAd stands out for its ability to deliver personalized, granular analysis tailored to each advertiser's specific needs. Whether through multi-touch attribution or detailed contribution analysis, TrackAd enables advertisers to make decisions based on accurate, relevant and customized data.

Attribution and Contribution in 6 Questions