Progressive search marketers know conversion attribution should resemble actual purchase behavior while assigning credit to the event or events that actually influenced the purchase decision. Unfortunately, current attribution models fail to take into consideration the uniqueness of an individual business and the purchase processes its customers tend to follow. While many marketers know they need to move beyond the last-click attribution model, they often are equally challenged by other dominant industry models such as u-shaped, time decay, etc., which leave them similarly short-changed. These models rely on hard and fast rules for assessing value, thereby hindering their ability to adjust to constantly shifting environmental factors.
By relying on rules-based models, marketers limit their own ability to easily adapt to new and emerging value propositions. Rules-based models need continuous auditing to ensure consistent performance — a manual and time-consuming process and one that provides an imperfect attribution solution at best. The rules-based model that works best will be the most appropriate for most conversions, but no single rules-based model can afford the flexibility and performance of a value-based approach.
Marketers using rules-based models also need to consider performing rationality checks on their attribution programs to ensure they’re not giving a disproportionate amount of credit to campaign elements which remain largely irrelevant to the measured conversion. For example, are baseball bat ads being credited in a conversion path for dog leashes? If so, the attribution model needs adjusting yet again … or Major League Baseball has a significant new licensing opportunity at hand.
If the point of attribution is for marketers to gain a clear understanding of what drives the conversion process, how can marketers justify using models that consistently provide flawed data? Frankly, most marketers just have few-to-no better alternatives and sometimes even lack the time to explore what’s available beyond their current campaign management tools. As technology continues to advance, especially in data analytics, so too do the options available to marketers.
Value-Based Attribution and Its Four Core Elements
Value-based attribution offers real advantages over rules-based models which overlook that value changes not only for each unique business and customer but also over time — today’s highly-valued interactions may not be the same ones marketers value most in the future. Any number of factors can have a significant impact on the value of an interaction. To move beyond rules-based attribution, marketers need to examine each conversion and define the value of each interaction in advance of their marketing campaigns — only then can marketers begin to leverage value-based attribution.
Marketers looking to embrace value-based attribution should start by defining value, considering four core elements:
1. Causality and Synergy – Not every interaction leads directly to a conversion, but one interaction will sometimes lead to another which will eventually drive a conversion. It’s important to determine whether a causal relationship exists between interactions or if a single interaction occurs/takes place independently of others. By evaluating these interactions closely, marketers can assign the appropriate amount of credit to each in the conversion path.
2. Engagement Necessity — Considering whether any level of engagement, and therefore conversion, would occur naturally through factors like navigational intent helps inform what prompted the interaction. Removing naturally-occurring interactions reduces the possibility of false attribution.
3. Role of the Interaction in the Conversion Funnel – After removing naturally-occurring interactions, measure the value of each remaining interaction by weighing that value against other interactions based on its singular contribution to a conversion. This is the attribution value of the interaction—each value should be a fraction of a whole adding up to a total of one.
4. Value of Loyalty — Measuring customer loyalty adds an important variable to discerning the true value of a conversion; understanding how a customer’s previous experience or purchase impacts the likelihood of an additional conversion provides a critical data point. Marketers may need to propagate some of the credit back to earlier conversions and the interactions that precipitated them.
Finding the Right Approach
Rules-based and value-based attribution programs do have common elements between them. First, marketers need to align their attribution programs with the larger goals of the campaign. Next, to ensure they can measure the effectiveness of the attribution program, marketers need to establish a baseline — understanding where the program stands today provides a measuring stick for program success. Establishing a good baseline requires some depth of knowledge in the channel or channels involved.
While much of attribution focuses on analysis, marketers need to be sure to create actionable scenarios through the use of deep channel knowledge that can produce tangible results from an attribution model. Models that are not part of a scenario, linked to a system or some other means of action do not realize the full return on investment possible from attribution. Finally, attribution models should be constantly tested and evaluated to ensure maximization of campaign value and waste elimination.
Due to the complex nature of value-based attribution and the in-depth analysis required, large-scale implementation remains impractical without the right tools. The emergence of advanced algorithms and scalable machine learning, which can manage these complexities by analyzing data and adjusting programs at regular intervals, should help marketers implement value-based attribution programs.
A small number of vendors offer value-based attribution solutions, which provide channel-specific or holistic campaign views (including Kenshoo which just unveiled its cross-channel attribution focused SmartPath offering – full disclosure: I work for Kenshoo and market this product). More important than choosing an automated solution is choosing the right automated solution. For those seeking the most seamless automation, the best solutions will likely include automated bidding platform adjustments. Marketers should review the available options to find the one that best matches their value-attribution needs.
Image: Different Paths by Shutterstock