Content Intelligence at risk from chaotic tagging - The importance of having an official taxonomy

THRON Team
THRON.com

Every day, entire Marketing&Sales teams wake up worrying about how to optimize how they reach out to digitally mature consumers, namely consumers who are more aware and more demanding, not to mention readier to complain when their expectations are not handled with the utmost care and expertise.   

In truth, there is no need to harbor such deep-rooted fears of our consumers’ potential reactions because the very same consumers have already given us the solution – people tell brands what they want in the data they provide. We know that when users interact with online content, data is generated about their interests. Analyzing and extrapolating from this data (extracting value) requires the intervention of Content Intelligence (CI) which is none other than Artificial Intelligence (AI) applied to content.

Content Intelligence is “created” from a suite of Artificial Intelligence (AI) technologies that “x-ray contents in three steps:

- semantic analysis which uses neural networks and natural language, etc., to “read”, “listen”, “look” and subsequently classify any type of content, from HTML pages to multimedia files (photos, videos, etc.), based on a selected topic indicated via the association of a tag.

- behavioral analysis in which analytical techniques are applied to the content activated by the tags, to return the interests of individuals.

- predictive analysis in which the content-generated data is processed using machine learning approaches and algorithms to forecast a user’s next move.

At this stage, it’s like the brand’s journey to reach its consumers has turned onto the motorway and on this must faster road, the connection with the customer is more direct given that a unique, personalized customer experience can be offered.

It’s important not to forget, though, that even the best motorways need maintenance and we should never overlook the risks of chaotically tagged content.  To evolve internal processes so that the use of Content Management can be optimized, preliminary steps are required to organize and rationalize content already held. Duplicates and inconsistencies must be removed, and a univocal taxonomy introduced to facilitate the reading of strategic data collected through CI. 

 

When does tagging become chaotic?

 

Let’s play an imaginary game for a second. We are wine-growers and own a large estate on the Euganei hills.  Our winery is well-known for its excellent vintages and is so famous that tasters and connoisseurs come from outside Italy. We take them on a tour of the basement where we keep all our “treasure”.   The customer tastes a few wines and tells you which ones he likes but the assistant makes the silly mistake of not taking a note of the names. What do you do now? What was it the customer wanted? If only you knew, you could try to sell the same vintage of a similar bodied wine. Your winery missed out on the opportunity to make a sale.

Tags are just the same - they are labels which classify topics within content - and can be associated to users expressing a preference for a particular type of content. Different rules may be applied to the tagging process within the organization, depending on who assigns the tags (internal or external editors, etc.) with the result that uniformity of terminology can be lost

Unfortunately, marketing teams sometimes lack the time to set up an official taxonomy which calls spade a spade and a shovel a shovel.  Any time spent on developing the tools needed to do optimal CI would be very quickly recovered through the benefits gained from having updated and real-time information on user interests. 

Content Intelligence for Dummies  by Wiley lays out the various ways tagging can get chaotic. There may be duplicates, which are when identical pieces of content are used across multiple channels but are tagged differently. For example, this would be like saving the same file with two different names, in two different folders, making it difficult for you to find.

Another problem is format. The same content in different formats may be given different tags. I am still the same person whether I am fat or thin. Not forgetting, either, that editors often use different terms to indicate the same thing; for example, someone in Tuscany may assign the tag “babbo” to talk about a father, which another editor in Veneto tags as “dad”.

Things can get messy!  Chaotic tagging makes aggregate user information (single customer view) foggy and often contradictory. Content Intelligence for Dummies suggests the following rule: “The more accurate the tagging, the more powerful the insight Content Intelligence can provide.” Powerful insights will give you essential information on which to launch automatic, personalized marketing campaigns and send user-centric content and CTAs.

We can see this in the figure below, which presents the single customer view created by a travel agency’s CI software.  The “tag cloud” on the right of the screen provides a useful ap of user Kate Parker's current interests. The topics and contents the CI software suggests she might be interested in are listed below, as a basis for recommendations that accurately reflect her interests.

tagcloud

The task of helping organizations build an official taxonomy is noted as falling within the sphere of the Content Intelligence Manager (CIM), the new professional role for the individual specializing in Content Intelligence. 

Optimizing content classification, which means selecting the official tags with which to label content given a single taxonomy and shared across the many teams in the organization, brings consistency to content organization and enables you to extract valuable data about customer interests. In turn, this feeds into the benefits of Content Intelligence, namely to increase productivity and efficiency, protect brand consistency, and increase engagement and conversation rates

The chapter in Content Intelligence for Dummies that looks at the danger of inconsistent tagging ends with a tip, namely don’t stress about tagging - better to use fewer but more significant tags, because overly general tags impede efficiency. 

 

All that remains now is to wish you all “Happy tagging!