Attribution and contribution marketing tools are part of the programmatic terms one needs to understand.
A lot of advertisers only consider post-click conversions. However, this is a very short-sighted position as they are missing out on a lot of insights to improve their conversion funnel.
The most advanced advertisers who use a data-driven marketing and attribution strategy are getting considerable benefits from those tools: up to 30 per cent in cost savings and 20 per cent in revenue growth.
What is the purpose of attribution/contribution tools?
These marketing tools allow advertisers to measure the way each of their marketing levers contributes
to events on their website. These events can be a visit or a conversion (lead generation or product purchase) according to one or several attribution models. ‘Marketing lever’ refers to the different budget lines of one’s media mix, like search, social, display, retargeting, affiliation or emailing.
For instance, a lever like display is very diversified, both in terms of formats (native ads, IAB, wrapping) and marketing goals (awareness, acquisition, performance, customer retention). To make relevant decisions, advertisers need to have the most comprehensive vision possible of their media mix and its resulting interactions. Moreover, the sum of interactions corresponds to a conversion path.
Interactions are associated with an event, more commonly called ‘conversion’. There are different types of conversions: online conversion (website visit, online purchase, subscription); visit in a physical store; purchase of a product in a physical store.
It’s important to differentiate ‘attribution’ and ‘contribution’
Attribution
This tool attributes an event (a visit or a transaction) to a unique lever. It’s a binary marketing tool. One lever is responsible for a transaction, or it is not. Historically, tools (such as Google Analytics) attribute the event to the lever that generated the last click. We can then talk about ‘last click attribution’.
Contribution
This tool attributes a weight of contribution to each of the levers that were involved in the conversion path. It then considers there is not only one unique cause to the conversion; it’s the accumulation of the different marketing levers that has generated this conversion.
Which attribution and contribution marketing tools are available on the market?
Here is a list of the major attribution tools: Easyence, Google Analytics, AT Internet, Eulerian Technologies, Commanders Act and Wizaly.
Adservers like Google Campaign Manager, Sizmek, or Weborama can also make attribution between marketing levers (tools measuring contribution can also measure attribution).
Bear in mind that attribution and contribution tools are not natively integrated into demand-side platforms (DSPs). Therefore, tracking elements from these platforms need to be integrated in order to measure the performance of programmatic campaigns. Then a trading desk will have to juggle between the monitoring on its platform and the performance observed on such a tool. This will have to be done according to the attribution model that was set up. Depending on the model, attribution can vary greatly. Advertisers must pay for these tools and cost depends on their complexity, website traffic and volume of events to be analysed.
How are attribution and contribution tools categorised?
There are three kinds of attribution and contribution tools:
First, attribution tools that only consider ‘click’ events, (Google Analytics and AT Internet). Those tools are the most widespread, with GA being free for advertisers.
Second, attribution tools that integrate ‘touches’ (all of the interactions with a lever, whether it is through a click or exposition). Those tools can measure the performance of levers whose business model is based upon impressions, such as display, social or affiliation.
Lastly, pure contribution tools that can be added on top of an attribution tool. (Easyence can analyse all the conversion paths of a website, to establish a custom-made model). This model aims at analysing the contribution of each lever as appropriately as possible, depending on the advertiser’s goals. It helps rationalise budgets between those different levers.
What are the main attribution and contribution models?
Many models exist. They’re either attribution models (the attribution of a conversion to a unique interaction), like the ‘last click’ model, or contribution models (which consider that several interactions have been useful to the conversion), like the ‘U model’.
Given the increasing complexity of digital marketing, the attribution model is increasingly challenged. Indeed, it does not consider the impact of other interactions along the conversion path. It is therefore mandatory to look at contribution models.
An interesting distinction to be considered in these models is the notion of ‘prior’ versus ‘post’. ‘Prior’ models (called rule-based) imply a bias in the analysis. This bias is linked to the person who chooses the attribution model. Thus, it will inevitably put forward one actor in the media mix compared to another.
The ‘post’ model (or statistical model), on the other hand, eliminates this bias. Statistical models focus their analysis on a past period of the campaign by using different theories (Markov, Shapley, linear regression). They highlight the optimal conversion paths to achieve maximum conversions.
The digital ecosystem is complex and technical, and it’s essential to understand it deeply – this is the reason why you need certified experts in this domain. They can assist advertisers to choose and better exploit these tools to optimise their campaigns. This is what we do at Gamned.
