Paid Media Intern, Tailwind
A key element of advertising is optimizing campaigns to improve performance, but in order to understand performance, we must first understand attribution. Many consumers encounter ads for the same brand or product throughout multiple different touchpoints on their conversion path. Once a conversion is completed, there are a variety of different ways for credit to be assigned to the content responsible for the conversion. These different forms are important to understand in order to calculate conversions more effectively.
Attribution & Attribution Models
The process of assigning value to the touchpoint is referred to as attribution. There are many different ways to calculate the value of a touchpoint for a conversion depending on which attribution model is being used. An attribution model is the rule or set of rules that determines how much credit is assigned to each of the touchpoints in the conversion path. Each attribution model comes with its own strengths and weaknesses, and in order to truly understand the performance of ads and ensure it lines up with the campaign strategy, it is important to first understand the types of attribution models and how to interpret them.
Single-touch attribution models have the simplest attribution breakdown, attributing one hundred percent of the credit for a conversion to a single ad or campaign. The common single-touch models are the first-touch and last-touch attribution models.
The first-touch attribution model attributes the entirety of the credit for a conversion to the very first touchpoint the consumer encountered the ads, whether that ad led directly to a conversion or not. The base reasoning behind using a first-touch attribution model is that a consumer cannot complete a conversion with a brand or product they do not know, therefore deeming the primary interaction the consumer has with the brand or product the most important. This model is simple to understand and interpret, but does not take the complete conversion path into account when attributing credit. Advertisers looking to broaden their reach and understand top-of-the-funnel behaviors can benefit from this attribution model.
In contrast to the first-touch attribution model, the last-touch model attributes the full value of the conversion to the very last touchpoint the consumer had an interaction with before ultimately completing a conversion. This is the most common form of attribution for paid media. Last-touch attribution is beneficial to advertisers measuring success based on which ads lead directly to conversions. This can offer insights into which content is driving direct conversions, but this model does not capture the importance of the path from the first touchpoint to the last.
Although single-touch attribution models are easy to understand and can provide simple insights into where conversions are coming from, they don't tell the entire story. In order to gain a more full understanding of the consumer’s conversion path, multi-touch attributions can be used. These models take more than one touchpoint into consideration, then weight their importance based on the attribution model. The four main multi-touch models are linear, time decay, position-based, and data-driven.
Linear attribution is the most simplified form of multi-touch attribution. Under the linear attribution model, the conversion value is split evenly between each touchpoint that the consumer engaged with before converting. Although this model recognizes each of ads that played a role in driving that conversion, this model is still oversimplified. All of the ads claim equal credit using this strategy, even though it is unlikely that each individual touchpoint played an equal part in driving that conversion. This model is best used to understand what content the consumer has seen before converting, without analyzing exactly the impact of each ad on the conversion.
Another common form of multi-touch attribution is time-decay attribution. The strategy of this model is to weight the value of the touchpoint based on how close the consumer interacts with that ad to the time of the conversion. The common reasoning behind using this kind of strategy is that the more a consumer sees the content, the closer they are to ultimately converting, with the last ad seen before a conversion being the deciding factor. This model can help understand consumer behavior later on in their conversion path, but the model is not perfect. Earlier touchpoints in the consumer’s journey are not deemed as important in this model, which is not necessarily the case. This model is best used to understand bottom-of-the-funnel consumer behavior, and should not be used when initial touchpoints are a major contributor to the conversion.
Under a position-based attribution model, the majority of the spend is attributed to the first and last touchpoints, without ignoring the impact of the ads seen towards the middle of a consumer’s journey. Forty percent of the conversion is credited to each the first touchpoint and the last touchpoint, while the rest is split evenly between the touchpoints in between. By placing emphasis on the first and last touchpoints, his model can help understand both top-of-the-funnel and bottom-of-the-funnel consumer behaviors. Although the central touchpoints are acknowledged in this attribution model type, the results can be misleading for campaigns in which touchpoints after the first but before the last hold great importance to the likelihood of a conversion.
Last but not least are data-driven models. This model of attribution utilizes machine learning with historical data to create custom rules of attribution for each touchpoint. This model is used by many big accounts and is the most effective under the right circumstances. Although data-driven attribution is the most powerful form of attribution, it is not the right for every account. In order for this attribution model to work effectively, the account needs to generate a very large amount of conversions. For smaller accounts, this is not a viable option as they will not generate enough data for the model.
There are many different attribution models of all levels of complexity. Single-touch models can be used to understand how consumers are converting off of a single touchpoint, and multi-touch offers more insights across the entire consumer journey. Each attribution model has strengths, each has weaknesses, and each has its place in the advertising world depending on the conversion types and the primary goals of the campaign. While there is no objectively correct or incorrect answer on which to use, understanding these models and the goals of each will help gain a fuller picture of campaign performance.
Current Open Positions
Tailwind is currently looking for the following positions to join our team:
Want to learn more about what fits your career trajectory best?
Find Your Tailwind.
With us behind you, you'll reach your destination faster.