Digital marketing campaign classifications — metadata — are the wild wild west of marketing data due to a lack of industry-wide protocol. The names very widely: metadata, dimensions, or classifications. We refer to the organization of such as taxonomy hierarchy, or schemas.
What exactly does all of that mean?
Simply put, metadata is information that describes other information. In marketing and digital analytics, metadata helps us dig into our campaign performance, describing metrics or events like impressions, clicks, conversions, and calculating ROI. For example, you could drill into impressions, clicks, and conversions through metadata like channel, source, or campaign name to better understand the performance of a given channel or publisher. And channel-specific metadata is often the foundation for attribution models and analysis.
As digital marketers and analysts, we need a system that creates order out of a very complicated marketing and advertising platform landscape. But that system needs to do the seemingly impossible: keep it simple while still supporting a marketer’s needs to have a consistent naming convention for complex metadata and associated values.
Here are 5 tips for campaign classification:
The last thing marketers need is an absurdly complex marketing and advertising platform landscape, and yet that’s what currently exists. The solution here is to standardize metadata. Use mutually exclusive values for each category and associated values with customized picklists. When applied to an entire organization, these metadata picklists must be built into one repository with touchpoints for all paid, owned, and earned media channels.
There will be times you’ll need to use non-standardized, inconsistent values, but these should only be used when there are many non-mutually exclusive values within a classification category. One important instance would be in the description of the touchpoint name or tracking code name — but efforts must be taken to create consistency in the exactness of each name and possible values.
Most of us use less than five classification parameters popularized by the UTM methodologies: source (or publisher), medium (channel), content, campaign, and term (for search). Rarely do we have more than a few metadata categories that are standardized via picklists.
In thinking beyond just simply marketing campaigns (which are typically limited to medium, source, and a few content attributes with UTM), and evolving to tracking and describing experiences with many metadata categories, we begin to drill into actionable touchpoints. The resulting naming convention protocol thereby describes the full spectrum of our interactions with our customers.
We use taxonomy, hierarchy, or schema interchangeably to refer to the organization of metadata into logical groupings; there’s a right way to organize these categories and there’s a wrong way. A system which removes metatdata dependencies (e.g. "Google" and "Paid Search"), we are able to view a report with any dimension we choose. Conversely, when metadata are dependent, there are limitations.
While individual classification is ideal in helping to determine specific attribution, there are times it’s useful to group common elements of metadata. Combining them allows you to see the full picture once filtered down (but of course is reliant on a system using picklists). The following metadata can be appropriately grouped: Channel, Content, Persona.
There is no shortage of data collection, however making it actionable to drive results is the challenge. The key is using a system that can:
To go deeper, download the guide, 5 Tips for Campaign Classification.