It’s much harder to find a nonfiction book in a library than a fiction book. Why? Because the author is the main search element in fiction, books are usually arranged alphabetically by author last name, sometimes by title. This makes finding what you’re looking for relatively simple if you know some basic information.
Sometimes, fiction is organized by genre, which is a less effective primary classification method. Knowing a book’s genre is science fiction doesn’t make it easier to find without an author name or title. Nonfiction, on the other hand, is classified in libraries by genre, using a system named after inventor Thomas Dewey. This is because the author name and title is not usually the primary classification criteria when looking for nonfiction; it’s the topic. Different types of books need different types of classification in order to be findable.
Comparing digital marketing data to book categorization can be instructive. Data is, after all, about context, findability, prioritization and usability. Digital marketing data is often difficult to use because there is so much of it, it lacks prioritization and context, and it sits in silos. How to extract and prioritize it, add context and make it usable, much less actionable?
The easiest solution is not to pull and manipulate data out of silos into unifying systems, which is how most data unification is being done today. Instead, the most efficient way to get the data you need out of digital marketing systems is to leave it where it is but add a data layer on top of it and classify it: apply a taxonomy system to all of your digital data, regardless of how or where it is generated. By using a classification system that sits on top of the silos where the data lives, you add critical context to your data and make it universally applicable, giving analysts the ability to see patterns and take action, all without having to touch the original datasets.
This makes it more practical and easier to get at the answers needed in any digital marketing measurement program: what actions are being generated by campaigns? What is driving the action? What is the result of the action? What are the attributes of the digital assets that drove the action? Answering these questions is the first step toward a digital marketing data taxonomy: unifying data from email, marketing automation, ad campaigns, sentiment analytics, social metrics, employee advocacy programs, PR impact. All can be better understood in context by universal application of a classification system to see patterns.
The element common to all programs is content. Content is also one of the most expensive, most effective, and most poorly used of all marketing assets, mainly because content has been so difficult to measure. You can’t measure content – which crosses all silos – with web metrics, or social data, or email data, alone. This is because most of those systems can’t answer the right question about the content: which assets are most successful on which platforms? With which audiences? In which formats? If you can’t answer these questions, I can almost guarantee you are spending too much or too little in the wrong places on content or distribution.
The answer? Classification, applied at the point of distribution, to track engagement and impact.
How? You can start with spreadsheets. Put each content asset in a vertical list. Across the top, add categories for your content: topic, format, channel, target audience, type of CTA, asset CTA. Then add the number of clicks and conversions under each column. Put a five-week testing plan in place to isolate what works, where.
Taxonomy and classification is the answer to the challenges of findability and goes beyond digital marketing data. It’s the answer to the challenge of depth of access that currently hamper the 99.9999% of content not on the first page of Google and the basis on which effective AI will be built in marketing programs. It’s a slow revolution, because it seems complicated. But like the Dewey decimal system, it’s the only answer to bring clarity out of the cacophony that currently exists with digital marketing data.
This article originally appeared on the ACA web site.
Latest posts by Jennifer Evans (see all)
- The most important content question (and how to answer it) - November 6, 2017
- Just Launched: Interbrand’s Best Global Brands 2017 - October 2, 2017
- 8 Key Questions About Community Dynamics - August 9, 2017