Marketing Attribution: Building a Framework

In the first part of our blog series focused on marketing attribution, we highlighted some of the more common attribution models employed by marketers and also introduced a few data-driven models designed to focus on conversion probabilities. One of the biggest questions many marketers face when it comes to attribution is….. where do I begin? While it seems relatively obvious, the first step to developing a successful marketing attribution plan is making sure your business is actually measuring baseline KPIs related to the effectiveness of your current marketing campaigns. Ask yourself these questions first:

Have we set clear goals?
Before you launch any campaign, it’s important to have a clear understanding of what you want to achieve (and alignment with your marketing team). Do you want to increase website traffic, generate leads, or boost sales? Setting specific, measurable, achievable, relevant, and time-bound (SMART) goals will help you determine the success of your campaign.

Are we using the right technology and measurement tools?
There are plenty of tools and tech available to help you track the performance of your marketing and advertising programs. While Google Analytics is a powerful and free option that provides insights on website traffic, conversion rates, and basic attribution models mentioned previously, there have been significant advancements in technology that allow you to also integrate data from your CRM, POS, and other sources to give you a more robust view of how customers interact with your brand and business. The last few years have seen significant growth in adoption of customer data platforms by brands and marketers alike. A customer data platform (CDP) is a software system that collects, consolidates, and manages data from various sources, such as website interactions, email communications, and purchase history. The goal of a CDP is to create a unified view of each customer across all channels and touchpoints, which can be used to improve the customer experience and inform marketing and sales strategies. CDPs typically include features such as data integration, identity resolution, and segmentation capabilities, and can be influential in developing your attribution model.

Are we tracking the right baseline metrics?
It’s important to track metrics that are directly related to the SMART goals you’ve previously established. In a simple example, if your goal is to increase website traffic, you might track metrics like pageviews, unique visitors, and bounce rate. If your goal is to simply generate more qualified customer leads, you might track baseline metrics like digital form submissions and lead conversion rate. Taking it a step further, segmenting simple conversions from actual qualified leads that enter your sales pipeline will start you on the path toward establish clear metrics for KPIs like cost-per-acquisition (CPA) and customer acquisition cost (CAC). Having your most important metrics clearly defined and measured will be critical to the success of your marketing attribution model.

Once you’ve established clear answers to the questions above, you can begin to evaluate which attribution model is likely to prove most beneficial to your marketing strategy and tactics. As we’ve highlighted previously, there are several commonly used attribution models you can begin with, each with their own strengths and weaknesses:

Last-click attribution: This model attributes all credit for a conversion to the last click before the conversion occurred. It is simple and easy to understand, but it may not accurately reflect the impact of earlier touchpoints in the customer journey.

First-click attribution: This model attributes all credit for a conversion to the first click that led to the conversion. It can be useful for identifying initial brand awareness, but it may not accurately reflect the full customer journey.

Linear attribution: This model evenly distributes credit for a conversion across all touchpoints in the customer journey. It is easy to understand and can provide a more holistic view of the customer journey, but it may not accurately reflect the relative importance of each touchpoint.

Time decay attribution: This model gives more credit to touchpoints that occurred closer in time to the conversion. It can be useful for understanding the impact of recent touchpoints on the customer journey, but it may not accurately reflect the impact of touchpoints that occurred earlier.

Position-based attribution: This model gives more credit to the first and last touchpoints in the customer journey, with a smaller portion of credit distributed evenly across the middle touchpoints. It can be useful for understanding the impact of initial and final touchpoints, but it may not accurately reflect the impact of touchpoints in the middle of the customer journey.

Data-driven attribution: This model uses data and machine learning algorithms to determine the relative importance of each touchpoint in the customer journey. It can provide the most accurate view of the customer journey, but it can be complex and time-consuming to implement.

When determining which attribution model and framework makes the most sense for your business, it’s also critical to give consideration to your prospective customers and their buying journeys. For example, if your business is relatively transactional (i.e., customer conversion is equal to the purchase of a cup of coffee) then a last-click attribution model and framework may be your ideal choice, as we can infer that a large percentage of your marketing budget is focused on lower-funnel tactics designed to engage the customer during an immediate moment of need. However, if your coffee shop has only recently opened its doors in your local market, then a position-based model may make more sense, since we can infer that a good portion of your marketing budget and tactics are also dedicated to driving brand/location awareness to generate upper-funnel customer awareness.

For brands and businesses that feature a much longer customer buying journey, one in which target audiences conduct a lot of research over a longer time period regarding your value prop, cost, competitors, personal preferences and much more, first-click and last-click attribution models do not provide much value. The decision-making processes and customer journeys for purchasing a new car, sending a child or loved one to college, or determining the best option for in-home care are most often very complex and involve a significant number of touchpoints, information and data prior to any purchase decision being made. In these examples, a data-driven attribution model that leverages machine learning, automation and incredibly large amounts of data will not only provide a window into the complexity of the customer journey but help you identify which marketing tactics and channels are most influential in moving your target audience down the path to conversion.

To learn how to approach implementing a data-driven attribution model, read Part 3 of of our blog series here.