Establishing an attribution model requires you to determine what you want to report on. Then, identify the data needed to help you determine that information, decide on which attribution model best fits your needs, and then analyze by applying the data to your attribution model.
Determine what information you want to get from your attribution model that will support decisions and drive actions. The results of your attribution model should influence future marketing decisions. Marketing metrics could include insights about final actions from the target audience, such as sales, conversions, responses or donations.
Identify the data needed for your attribution model, as well as how you will collect, clean, and review your data sets. Needed data sets may include both customer-specific and more generalizable information.
Customer-specific data would include information that relates to a particular individual or their actions, such as time, date, purchase, spend, and touchpoints that led to a specific conversion. General data is not associated with one specific customer. Instead, the data serves as proxies for a set of customers. It may include information such as all of the ways customers can engage with your organization, clicks on a website and any demographic information associated with those clicks.
Decide which of the four or five things you will measure and which attribution model can be used to most effectively lead you to actionable results. [Click here for additional information on Approaches to Attribution.] Some attribution models can be conducted internally, while the more complex models often require the assistance of a vendor. [Click here for a list of questions to consider asking while interviewing and selecting a vendor.]
Apply the data you have collected to your attribution model and analyze the results. Use that information to support decision making and drive actions.