Welcome to the Oracle Modern Marketing Blog:
The latest in marketing strategy, technology, and innovation.

The importance of cross-channel attribution modeling and analysis

You may have heard John Wannamaker’s famous quote from about 100 years ago that “half the money I spend on advertising is wasted, the trouble is I don’t know which half!” Of course now in the age of cross-channel digital marketing, it's a lot more challenging to measure the effectiveness of marketing communications that touch the consumer in so many different ways.

As a result, many marketers are now paying close attention to the emerging field of attribution modeling and analysis to define better strategies to allocate their precious marketing budgets across different channels and programs.

I’d like to introduce you to the basic concepts of attribution modeling and analysis by using some simple examples. So let's start with how most of us may go about purchasing a product, say a new car.

The Customer Journey

Let's say you see a new car commercial on TV or a friend shares photos of her new car on Facebook. The first thing you probably do is use your computer or smart phone to search for similar cars or visit some auto review web sites. During this process you may get some targeted sponsored listings or display advertisements that entice you with attractive deals. After you click on one of those advertisements, you will see some relevant information and also may get invited to complete an online form to receive emails about special promotions and updates.

By completing one of those forms you may get a welcome email followed by regular emails personalized with special offers from local dealers about cars that you like. For several weeks you may have in-person and online dialogues with your family and friends about getting a new car, and finally you may come across an irresistible interest rate promotion via email or a display advertisement and decide to go to a local dealer for a test drive. After several hours of negotiation, you finally decide to trade in your old car, write a check, sign the papers and drive home in that shiny new car. A few days later you receive another email congratulating you on purchasing a new car and asking you to complete a short survey about your new car shopping and purchase experience.

This whole process is often referred to as “conversion funnel” which starts with broad awareness and research and gets narrowed down to a very specific product, offer and final conversion.

In this example, you received a number of marketing communications across different channels over the lifecycle of the conversion channel which finally led to purchasing a new car. The marketing team of that car company would really like to find out how you were influenced by those marketing communications so that they can effectively attribute your conversion (i.e., purchasing a new car) to those activities.

Having this information would in turn help that team to fine tune and optimize their future marketing communications for promoting similar products that may be of interest to people like you. For example, if you clicked on the low-interest rate lease offer in a promotional email just a couple of days before showing up in the dealership for a test drive, then that could be a very good indication that your conversion could be attributed to that email promotion.

Meanwhile, if you also saw multiple display advertisements with the same low-interest rate lease offer around the time that you clicked on your email links, then there is a strong possibility that the display advertisement may have also attributed to your conversion. Finally, during the negotiations at the dealership, you may have received a very attractive trade-in offer for your old car which finally pushed you over the edge to buy that new car. So the dealership would definitely like to also get attributed for their salesperson’s job well done.

Your interactions with each marketing communication throughout your entire car shopping and purchase experience may somehow get recorded — then get fed to a sophisticated attribution analysis program that would co-relate those interactions to come up with some suggestions about how each activity may have contributed to your final decision to sign on the dotted line.

Improper Attribution

Unfortunately, most marketers don’t have access to such sophisticated software or a team of analysts that could make sense of the analysis results. Therefore, in many cases the final conversion is attributed to the very last user communication, or a combination of the last X number of user interactions over a specific look-back period, like seven days, right before the conversion.

In my example, the conversion could most likely be attributed to the dealer’s attractive trade-in offer as well as the low interest rate lease promotions you saw in the emails and display advertisements.

According to a Forrester Research report on attribution modeling: “Traditional one-to-one, last-touch methods of allocating demand to marketing efforts are outdated and lead to a suboptimal marketing mix. Customer Intelligence (CI) professionals must adopt a cross-channel attribution model in order to optimize marketing budgets, accurately calculate customer value and acquisition costs, and develop a holistic view of the marketing ecosystem. Failure to embrace this new standard is expensive — firms will be plagued with continued channel conflict and an inefficient marketing budget.”

With more people shopping online, a lot of data is being collected about user interactions with different marketing communications across the Web, social media, mobile apps, etc. As a result, more and more marketers are getting interested to mine this data to gain better insights about the consumers’ cross-channel behavior.

In future blog posts I will discuss a number of emerging cross-channel attribution models and software companies that offer a wide variety of solutions for helping marketers better understand how their different marketing communications are positively influencing various segments of consumers.

For tips on evaluating web analytics tools, read my previous post on the topic.

Be the first to comment

Comments ( 0 )
Please enter your name.Please provide a valid email address.Please enter a comment.CAPTCHA challenge response provided was incorrect. Please try again.