Digital Metadata [and why it’s cooler than you think]
Metadata – ‘data about data’ – is an often misunderstood topic that glazes the eyes of the best of us. It interested me as much as a dusty library or the sensible pair of shoes my mom made me buy in elementary school when I had my eye on a pair of light-up jelly flats. While I recognized that metadata’s functional role describing the who, what, where, when, and why of the data an organization produced, I thought metadata was a static ‘nice to have’ set of labels that should sit quietly in a data warehouse. I couldn’t relate.
Companies that share my previously held beliefs and fail to manage their digital metadata put themselves at a serious disadvantage. Data is wasted and digital experiences suffer. This has real consequences, given that 52% of consumers say they’ll switch brands without customized experiences.
Whether it’s during content creation, marketing campaign promotion or customer interactions, the creation and management of metadata is relevant to the business and can unlock the full value of your data.
This article will describe a couple of types of metadata that interact in the business, and discuss how metadata can be used to improve branded experiences and outcomes.
What is metadata?
Very simply, metadata is the context for your data. Applied correctly, metadata will describe each important aspect of your data. When you transfer data across marketing solutions or into large data warehouses to be combined, manipulated and analyzed, metadata will tell you what that data is, where it comes from, and why it matters. As Gartner aptly describes, metadata “provides the understanding that unlocks the value of data”, turns your information into assets and improves their use.
Imagine a pharmacy with hundreds of bottles filled with thousands of unlabeled, nondescript pills and ointments. Without the metadata – the item numbers, dosage instructions, batch numbers, warnings, brand and chemical names – the pharmacy would be chaotic and dangerous. No one would know how to locate the medicine a patient needed or how to use it. The wait time on getting a prescription would be massive, and customers might walk away from that poor experience with the wrong material. Digital metadata is the same; you need context to deliver the right experiences and insights.
There are specific types of metadata, for example, descriptive metadata, that we won’t get into here (you can learn more about them in this article). However, I’d like to describe a couple of categories that interact to deliver contextualized experiences for customers.
Customer: A customer has historical metadata over time from activities like form fills, data import, or appended behavioral data. Customer data extends beyond a single touchpoint and includes the history of a user’s engagement with your brand. This category will also include descriptors for CRM data.
Campaign: You also create and interact with metadata during the creation and promotion of a campaign. Metadata here will describe the data collected when a customer engages with a campaign link. Metadata can be appended to campaigns through dimensions (Google Analytics) or SAINT (Adobe Analytics).
Content: Every time a customer or user engages with a campaign they’re engaging with content (URL, images, videos, etc.). Metadata attributes are appended here as well. Metadata enables organizations to conduct multivariate and split tests. For example, when a user clicks on a tracking link to a page, the page experience might be completely different because of the collected browsing metadata.
How can I make the most of metadata?
The degree to which you create and govern metadata will impact the effectiveness of your organization’s personalization efforts. Context is an essential element to best-practice experience models. Machine learning and AI require cross-touchpoint experience metadata to run. When metadata is consistent and managed correctly in an organization, customer, campaign, and content metadata can be used to serve experiences that are relevant to use behaviors and preferences. Ultimately, metadata management will help address the challenge many organizations have unifying customer insights across channels and devices.
Data enrichment is the process of adding meta-data to the data you’ve already collected in your analytics solution. Metadata can be used to enrich the data you collect in your analytics solution and allows you to collect data beyond a limited number of dimensions. It’s up to you to decide if there’s a different way to define the data you collect and add metadata to it. The reason you would do this is if you decide if there’s data you can add to a dimension that will give you additional or granular views into your experiences, websites, etc.
Data enrichment can help you:
- Aggregate data like an excel pivot table; group things together to view performance over time
- Create additional reports in your analytics solution by the metadata attribute
- Use any metadata attribute to segment your data
You can find a detailed walkthrough of analytics metadata and data enrichment use cases in our Analytics Nexus presentation by Adam Greco.
Properly managing your metadata across an organization can help you make the most of the data your organization is collecting, but there are a few barriers to overcome.
Metadata can be used to enhance digital experiences, but in practice, it becomes more complicated with the creation of more channels, devices, technologies, etc. through which users interact with your brand. In order to have a personalized experience, you need robust and complete cross-channel customer reports. But how can you build those when data about customer habits and behaviors are being created and reported from multiple marketing tech solutions that don’t communicate? The customer profile is fragmented, siloed in separate systems with different formats. It results in experiences that may not be relevant to customers.
Although users might interact with different segments of an organization (for example, those departments that deliver content vs those that interact with and create CRM data), those segments often create metadata definitions that differ widely from each other, which obscures the connection or significance of data when you go to combine and analyze it. This becomes especially troubling given the importance of machine learning in delivering competitive (and personalized) experiences. Machine learning algorithms need well-defined data to produce insights. Building a unified marketing taxonomy is part of creating metadata descriptions that the whole company can use.
Unfortunately, organizations are usually unfamiliar with the true scope of the data their creating, and where to find it. Unfortunately, it’s easy to become overwhelmed with poorly-defined data, and organizations spend most of their time locating, cleaning, and reconfiguring data in warehouses so they can analyze it (the symptom) instead of creating processes to create well-defined data in the business groups themselves (the disease).
Don’t get overwhelmed with creating a metadata plan for your entire organization – start small with one or two teams. Despite challenges, taking charge of your digital metadata will produce the context you need to personalize your experiences, gain valuable insights into your campaigns, and find and leverage content effectively.