This post was originally published in MarTech Advisor
Before the days of digital marketing, marketers and advertisers often faced the problem of not enough data. They sent things out into the world, crossed their fingers, and did their best to map business results to their marketing efforts.
Today the opposite is true: we are flooded with data. Each of the many tools in our martech, adtech, and overall technology stacks gives us a myriad of information to work with – and the number of tools that we use continues to grow
. This is a blessing and a curse. We now have the data we need to understand precisely how our efforts perform. But without some parameters around that data, it’s hard to find the signal through the noise.
My colleague Mariena Quintanilla, who has spent a decade focused on the quality of Ad Tech data, describes attribution as a story. She often tells our clients that if you’re not collecting the right dimensions of the right metrics, you can’t tell that story. In the case of marketing attribution, and frankly most digital marketing efforts, the narrative of that story is written based on user path data. We want to sort through our big data to understand who our users are, what they’re doing, when, how, and through our analysis, hopefully why. This process gives us the insight to reach our customers and use our budgets more effectively.
Back to the data. To create meaningful user paths and leverage them into successful multi-channel attribution, there are four types of data that are non-negotiable must-haves:
- User identifiers: User IDs, as the name suggests, are data that help us identify individual users. User IDs include cookies (on the web); device IDs (on mobile); and internal information, like login/account data. The more user ID data you can gather, the better. We use it to connect the dots across user paths. If you have login data, for instance, you can see that the same person engaged with you from this phone, this laptop, and this browser. Attribution systems then connect those dots to build a more complete picture of user behavior.
- Timestamps: One of the reasons attribution systems are so powerful is that they deal in event-level data instead of aggregate data, delivering measurement insights at a very high level of granularity. As such, you need to know exactly when each event (a click, impression, conversion) occurred. This data is often rather easy to gather, as it’s usually built in to reports from ad servers and the like.
- Media data: Media data provides the “what”: what a user saw and did. Was it version A or B of the creative, did he or she click or was it just an impression? The more detailed your media data, the more precise your attribution results can be. You can use naming conventions around your data, for instance, to designate not just the campaign, but the creative and keywords. This gives the attribution system more ways to slice the data, opening up greater optimization opportunities.
Channel data: To accurately track your campaigns, it’s critical to know where events took place: your channel data. Channel data can be surprisingly difficult to gather in an organized way. Different vendors and technologies track channel data in different ways, so you often end up with “mixed metaphors.” Like media data, a thoughtful taxonomy of naming conventions can help ensure that your channel data is clean and available.
These four data points, amid your sea of marketing data, form the backbone of your user path. They tell you who (user IDs), when (timestamps), what (media data), and where (channel data). This information offers clear insight into what’s working and what’s not, and how the pieces of your marketing puzzle fit together. It also enables opportunities for optimization. When you know, for example, what creative, as part of what campaign, is outperforming its cohorts on a specific channel, you can move budget accordingly. At the scale of some advertising programs, even one such adjustment can translate to hundreds of thousands of dollars in savings.
As a digital marketer, you have almost endless data at your fingertips. If impactful marketing attribution is your objective, you don’t need it all – just the right data. Gather (and govern!) the four types of data above to allow your attribution efforts to tell your story effectively