1. Sales Tracking

Types, Challenges, & Application of Attribution Modelling

attribution modelling

In digital advertising, attribution modelling allows marketers to capture actual ROIs for multiple marketing touch points, understand how and why their prospects converted into customers, and connect the right messages more precisely to the right consumers.


Imagine a prospect clicks on one of your display ads from his work computer and spends hours browsing your site, only to go home, clicking one of your Google Adwords ads and then making the purchase. Hence, the Adwords ad will get full credit for the conversion, driving up your customer acquisition costs and failing to validate your display ad with at least partial credit for the sale.

Multi-touch attribution (MTA) involves the problem of assigning credit to one or more sources for driving the user to making a purchase. Rather than giving all the credit to the last ad a user sees, multi-touch attribution allows more than one source to get the credit based on their corresponding contributions. To do so, MTA tracks how consumers interact with different channels and the actions they take after such exposure, which is crucial when multiple channels, such as search, display, social, mobile and video are involved.

 

Types of Attribution Modelling

There are multiple ways to attribute value to customer interactions:

· Linear Attribution Model: A basic model where all interactions receive credit equally. This model ignores the role that channels play in the conversion process and is typically used when no one particular interaction is known to convert customers.

linear attribution

· Position Based Attribution: This model attributes 40% of the credit to the first and the last interaction and the remaining 20% is divided evenly among the middle interactions. This model is best used for tracking drivers of awareness and action, as well as for companies with longer sales cycles.

position based attribution

· Time Decay: Most credit is given to the interaction closest to conversion, and the touch point prior to that will get less credit based on a smart and simple algorithm. This model is particularly useful for companies with very short sales cycles.

time decay attribution

· Customized Attribution Model: The analyst determines the credit to be given for each interaction, depending on knowledge of the product, customer base, and sales funnel. Hence, key factors include historical performance, current channel-mix and spend patterns.

custom attribution model

 

Challenges of Attribution Modelling

Attribution modelling involves three challenges:

1. The challenge of attributing offline impact. Revenue generated from the existing customer base, brand equity, and offline marketing efforts are difficult to track. Moreover, if spend, effort, or reach on certain channels are linked to discounts and promotions then the effect of advertising becomes unreliable due to price or promotional effects.

2. The challenge of attributing credit to all digital marketing channels that contributed to a particular conversion. For instance, the volume or likelihood of search queries for a particular brand or product depends on consumer interest at any given time, which is influenced by total marketing effort. This means that paid search impressions, and therefore clicks and spend, are difficult to disentangle from the indirect influence of other types of marketing and promotions on search queries.

3. The challenge of integrating data. Integration of data from web analytics, CRM software, and customer service teams is required in order to create a useful data set for multi-touch attribution. Hence, the task of data integration in a multi-device world becomes even more important.

Apart from being aware of these limitations, companies can cope with them by pursuing new models and algorithms. Moreover, almost any type of multi-touch attribution model is superior to a last-touch model in terms of reaching true marketing impact.

 

Attribution Modelling in Practice & Privacy Issues

Investigating multi-touch attribution, analysts need to ask themselves if the company has an attribution problem. Time lag and path length, which can be found in both Google Analytics and Adwords, reveal the percentage of conversions not occurring the same day as the impression or click. This indicates whether it is worth spending additional time and effort moving away from the traditional last-touch approach. For instance, the “Model Comparison Tool” in Google Analytics allows you to attribute credit to all digital marketing channels involved in conversions.

Nielsen (2014) demonstrates how a consumer exposed to a digital banner ad, subsequently ran an online search, which took her to a landing page, ultimately resulting in a purchase.

Nielsen attribution modelling

However, this example does not reflect that many companies lack the resources to successfully implement such a robust system based on both online and offline vendors. Nevertheless, getting online channels recorded through an analytics tool is realistic and could be done by most companies

In contrast to third party tracking platforms, Google Analytics does not natively provide visitor-level data. It has been argued that while the obfuscated unique visitor ID in the so-called UTMA cookie is personally identifiable, it is not a social security number, IP address, or name. Yet, bearing the increasing focus on privacy issues in mind, we should use tracking tools with due care.

 

References:

Kaushik, A. (2013). Multi-Channel Attribution Modeling. Retrieved 11 February 2015 from: www.kaushik.net

Nielsen (2014). Real Return: What Matter Most In Digital Advertising. Retrieved 11 February 2015 from: www.nielsen.com

Shao, X., & Li, L. (2011). Data-driven multi-touch attribution models. In Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining (pp. 258-264). ACM.

Wooff, D. A., & Anderson, J. M. (2014). Time-weighted multi-touch attribution and channel relevance in the customer journey to online purchase. Journal of Statistical Theory and Practice, 1-23.

Simon Raun Madsen

I am a marketing, business, and communications academic and practitioner with strong cross-disciplinary skills from courses and work experience within the fields of marketing and business intelligence. In particular, my passion for technology and several years of experience with online marketing have provided me with a flair for digital solutions, data analysis, and front-end development.
Simon Raun Madsen

Latest posts by Simon Raun Madsen (see all)