Demystifying Attribution Models to Bring Clarity to Your Business

January 14, 2019 Jen Robinson

Today’s digital marketers are faced with a tall order: Deliver incredible, personalized customer experiences at every step of the customer journey, and tie every touchpoint, lead, customer, and dollar back to the marketing effort that created them. That’s why attribution is so business critical—it’s not easy, but organizations that do it well can assign and understand the value in the work they do, and ultimately prove elusive marketing ROI.

However, the customer journey is rarely linear, which makes it tough to know exactly how individual interactions influence or lead to desired business outcomes or customer actions. In fact, recent research from McKinsey turns the traditional customer buying journey on its head. More than ever, brands have the power to meet customer demands through technology and can actively shorten and shape customer decision journeys. Still, making sense of any customer buying decision requires cobbling together channel, content, and conversion data — both online and offline. It also requires alignment around how to weigh every touchpoint in the customer journey.

It starts with selecting the right attribution model for your organization’s shared business goals and objectives. Let’s break it down:

Selecting the Right Attribution Model

Marketing attribution, fractional attribution, or multi-touch attribution (MTA) are terms defining the same concept of using a rule, or set of rules, to determine how credit for sales and conversions is assigned to touchpoints in conversion paths. In other words: giving partial credit to interactions a customer is exposed to across their full journey path.

Customer journey touchpoints

  • Demonstrate a return on marketing investment  
  • Distribute spend of the marketing mix efficiently  
  • Optimize marketing campaign tactics
  • Eliminate double-counting  

Fundamentally, applying a marketing attribution model tells you how your marketing is doing.  

Simple attribution models

Simple attribution models are powerful, but the underlying methodology and algorithm are not universally agreed upon. There are many models an organization can choose from with advantages and disadvantages. Here’s a summary of the top models utilized in the industry:

First-touch - credit assigned

Some organizations use first touch, wherein the first marketing contact point would receive 100% of the credit. The rationale is that that first contact point is what lured the customer into the funnel, and the rest was just an inevitable journey toward conversion.

Pro: A good way to understand which interactions are most initially engaging. Also backs up the latest customer buying journey from McKinsey.

Con: Customers engage with an average of seven pieces of content before making a purchase decision.

Last-touch - credit assigned

All of the credit goes to the touch that happened just before the conversion event. The thinking here is that this is what convinced the person to actually pull the trigger and make the purchase, so that’s what we want to credit.

Pro: Shows you exactly which interaction made your customer raise their hand.

Con: Doesn’t consider previous interactions that may have prepped customer for the conversion.

Equal distribution - credit assigned

Also called linear distribution, equal distribution is as equalitarian as it sounds, and splits credit evenly across all of the channels. This model is easy to understand and an improvement on last and first touch, because it doesn’t ignore any of your channels.

Pro: Provides a clear view of the entire customer buying journey, from first touch to last touch.

Con: Every interaction is weighted equally, though interactions closer to conversion should often be given more credit.

Position-weighted - credit assigned

Sometimes called a U-shaped model, last and first touch are deemed the most important, but you don’t give them all the credit. You divide some of it between your middle touches. This model applies credit to all touchpoints and emphasizes the importance of the first and last touchpoints: If the first touchpoint didn’t pique their interest, they’d never have become a customer, and the last touchpoint must have had an impact as it lead to the final conversion.

Pro: A simple way to give credit to the first and last touch, which are deemed the most valuable touchpoints in a customer journey.

Con: A touch that influenced advancement in the customer journey (e.g. organic search or organic social) would not get the opportunity to get additional resources to produce valuable content if the credit given is nominal.

Decay - credit assigned

More recent touchpoints get the most credit, and as we move backward in time, our channels get less credit. The concept is fairly easily to grasp and seems like a logical approach.  

Pro: Placing more weight on recent behavior is a logical approach to understanding your customer’s current mindset.

Con: Undervalues the impact of early customer interactions.

Applying an Algorithmic Attribution Model

While simple attribution models get you closer to assigning value to all your marketing efforts, they still rely in part on qualitative insights, which leaves some room for subjective interpretation. The one advantage they serve is that you’re not double or triple counting conversions. However, there are more nuanced models to help you definitely attribute actions to business results.

Enter Algorithmic Attribution

This is the most accurate, and also the most difficult to understand. Using machine learning, this model analyzes every touchpoint, alongside the attributes and metadata(think: frequency, time of day, creative, publisher, etc...) that lead to a conversion or not. Each touchpoint is weighted based on your own data.

The illustration below demonstrates four different customer journeys: two that lead to conversions and two which do not. Using an algorithmic model, each touchpoint is evaluated by conducting thousands of simulations to determine which lead to a conversion.

Simply put, if a touchpoint often participated in a conversion, the algorithm might conclude that touchpoint had a positive impact on the customer’s buying decision and give it more weight or credit. Similarly, if a touchpoint often participated in a lack of conversion, the algorithm might determine that this touchpoint is a detriment to a conversion. It would assign lower (or no) credit.    

Looking at the example below, one might conclude that having a social touchpoint included helps lead to a conversion while having email doesn’t appear to have an effect on conversions. Methods like Shapley’s or Markov models apply this type of thinking.

Since it uses your organization's data, the positive AND negative outcomes (i.e. whether a conversion happened or not), it's the most accurate assessment of what contributes to a conversion. That said, it’s difficult for those who are less analytical to accept and understand. Many organizations lack the internal resources or experience to apply a model to built internally; using third-party solutions is far easier to implement than attempting to build your own.

Considerations for Implementing an Algorithmic Attribution Solution

Getting buy-in

Since there are multiple ways to apply attribution, some organizations are skeptical about how marketing assigns credit to each touchpoint in the customer journey. Much of this skepticism can be addressed through alignment around shared business goals and objectives.

Achieving cross-team alignment:

  • Dig deep into how marketing efforts are being tagged and tracked, and understanding the methodology for assigning credit.
  • Determine what success looks like across the organization. What are the shared business goals and KPIs that everyone can agree on? Aligning around those shared benchmarks ensures the marketing team is investing and tracking what matters most.
  • Show how a touchpoint directly and positively influenced an agreed-upon business outcome. This leaves little doubt about marketing’s impact on the bottom line.

Implementing a solution

Once you have buy-in, adoption can be another hurdle. It requires metrics collection discipline and governance. If each marketing tactic is tracked and collected in separate systems, assigning credit will be impossible.  

IT collaboration is also critical. In recent years, with the emergence of SaaS and cloud-based solutions, marketing has been less reliant on IT to implement new technology. However, more than ever, CIOs and IT teams are tasked with developing innovative business processes, implementing new technology, ensuring seamless integrations, and helping to demonstrate a return on technology investments. IT also has the most cross-organizational know-how into the functional and security needs of the entire organization.  

There are ways to manually manage all the variables and systems, and newer experience data platforms with built-in governance that makes unified tracking a lot easier and more reliable. By working with IT on implementation, marketing can deliver personalized, incredible customer experiences at every touchpoint—and tie those interactions back to real business results.

Using an Experience Data Platform  

So what exactly is an experience data platform? Simply put, it provides a unified view of your customer’s experience with your brand — namely their journey across each touchpoint, from paid and earned media channels to content to CRM and e-commerce data. In other words, you’ll know exactly how every piece of content and every customer interaction impacts the bottom line. With the ability to see each customer interaction in context, you can deliver even better customer experiences, create and do more of what’s working, better prove the impact of your marketing efforts, and deliver trusted ROI that you can tie directly back to sales — and revenue. From there, an XDP allows you to integrate and automate tracking across platforms and integrations to make governance reliable and easy.

The benefits of using an XDP are that it:

  • Provides structure and enables governance through utilizing consistent standards
  • Unifies channel tracking across teams and integrates technologies  
  • Expands beyond the channel by incorporating messaging or content in the mix
  • Tracks every customer touchpoint

By using an XPD, you’ll enjoy accurate data collection which leads to confident ROI calculation and the ability to deliver well-targeted messages to your customers and prospects through which will delight them—at the right time, wherever they are in their customer journey.

The Outcome

Whatever route you take, implementing a robust data collection and marketing attribution solution will not only help you better understand how your marketing affects the bottom line but will also deliver customer insights that will give you a leg up on the competition.

To learn more about how Strala by ObservePoint can help you integrate, attribute, and prove the ROI of your marketing efforts, schedule a demo with us.

Previous Article
Why Data Visualization Tools Aren’t Enough
Why Data Visualization Tools Aren’t Enough

Learn why maintaining a focus on data visualization instead of data quality is limiting your insights for m...

Next Article
How to Conduct Cross-Device Analytics Testing with LiveConnect
How to Conduct Cross-Device Analytics Testing with LiveConnect

Learn how LiveConnect can help you ensure positive customer experiences across devices by allowing you to t...