What really caused the customer to buy? Was it the display ad seen while reading an article online? Could it have been the paid search ad they saw on Google? How about the print ad in the Sunday paper? For marketers that run media across channels and audiences, they routinely face the challenge of trying to attribute success to a particular campaign or media. This insight allows more effective investment and allocation of marketing resources for future campaigns. However, most marketers can only compare performance across platforms or media, without making a direct comparison of investment performance.
There have been other approaches to rationalize and compare performance across the digital channel. Approaches such as “first-click” (giving credit to the first click a user made) and “last-click” (giving credit to the last click/interaction a user made) both provide views of user behavior, but neither account for the impact of multiple interactions. And none of these include offline interactions.
Attribution modeling provides a new way for marketers to use algorithmic modeling to take into account all the different media a customer has seen, and provide a relative impact value for each media interaction. This data-driven approach elevates the level of insight available to measure campaign performance, enabling organizations to better optimize both spend and media mix. For marketers who run multiple campaigns, and have a large mix of media, even a minor shift could have a major return.
Running a successful attribution modeling program is contingent on four key steps:
Vendor selection. For most companies, this is too specialized of a project to handle internally. Selecting the right attribution modeling vendor sets the foundation for success. While each of the leading vendors all offer the basics, some are more effective at integrating offline data, others do a better job at delivering an easy-to-use interface while others are more focused on scenario modeling. Upfront identification of program goals and key measures will help ensure a successful vendor selection.
Tracking governance. One of the most critical aspects of a successful attribution modeling program is ensuring all the data is tracked consistently. Many organizations are fragmented in their approach to digital media delivery – often using multiple vendors, media servers, and measurement programs. The easiest way for all digital media to be viewed consistently in the data model is to be served through one central ad server. If you decide against this route, cookie matching is possible but adds significant complexity to this endeavor. This doesn’t mean replacing vendors or media partners, but you will need to go through a process to ensure standardization of tagging.
Data integration. For the model to address all spend sources, all marketing data needs to be incorporated into the reporting stream. While the single ad server mentioned above will support the digital data, offline data – from TV to print to outdoor advertising – must all be accounted for. This means identifying the media sources, establishing a standard for how the media will be tracked, and developing a process for how the data will be routinely fed into the data model.
Actionable follow-through. With the model set up and test data flowing, organizations must look internally to develop a structure that supports modifying behavior based on the model. Specifically, there must be a team or individual who has both the responsibility and authority to shift investments based on performance.
As marketers get new insights into campaign performance, and can make more informed cross-channel comparisons, a more clear view of the customer will emerge. This insight can lead to more informed marketing investment decisions – and hopefully, a more engaged customer.
Time is running out to feature your company in our inaugural Mobile Vendor Reader Survey.
Marketers create personas to better understand their target audience and what it looks like. If marketers can understand potential buyer behaviors, and where they spend their time online, then content can be targeted more effectively.
What’s behind a successful data-driven marketing strategy?
Audience targeting can be challenging in social media, especially when brands make quick assumptions about their target users. How can you avoid generalisation and what are the real benefits of it?