AEM Assets Smart Tagging makes it easier for marketers to find the content marketing assets they need, when they need them.

The problem: many assets

Sometimes a search for a piece of marketing content parallels the proverbial search for a needle in a haystack. As previously noted, a 2022 Adobe Commerce study showed that consumers demand personalized content.

  • Because of this, the number of pieces of content created by a company has increased astronomically.
  • A firm may have to create MILLIONS of pieces of content.
  • Each individual asset is necessary to cover the demands of various products, asset types, regions, markets, and audiences/personas.
  • Are you prepared to search for the piece of content that specifically addresses the benefits for Bulgarian small telecom firm finance executives? It may take a while to find it.

One solution: manual tagging in an unstructured way

One solution is to “just tag it,” and ensure that every piece of content is tagged by someone. While this solution is better than no solution at all, there are two drawbacks:

  • Tagging is too slow. It takes a long time for people to manually tag items. Even if you minimize the cost by using low-cost labor, the time factor is still critical. Even if your people can rapidly apply one tag every five seconds, it would take over a month for 10 people to tag 2 million items. Businesses can’t wait a month to access their assets.
  • Tagging is inconsistent. If 10 people are tagging the items, the tags will probably be inconsistent. While one person tags an item as a “car,” another may tag a similar item as an “automobile.” Although the two assets are similar, this is hidden because of inconsistent tag use.

A better solution: auto-tagging based on asset characteristics

Instead of manually applying tags to every asset, many solutions use auto-tagging. Auto-tagging employs artificial intelligence and machine learning to examine the properties of a text, image, or video asset and automatically apply the appropriate tags.

Auto-tagging addresses the two drawback of manual tagging:

  • Auto-tagging is much faster. If you throw enough computing power at the task, a process that requires over a month with manual tagging can complete in minutes with auto-tagging.
  • Auto-tagging is consistent. Because a single algorithm is performing the tagging, rather than 10 separate people, the tags are consistently applied to all assets.

But is auto-tagging good enough? As Adobe notes, “most solutions only apply generic keywords that do little to provide enough detail to quickly find the perfect asset.” When the auto-tagging algorithm is trained on a general model, the resulting tags are somewhat suitable for all businesses, but not targeted for your own business.

The best solution: AEM Assets smart tagging using your taxonomy

But what if the automatic tags DO provide enough detail to meet the needs of YOUR firm?

AEM Assets “smart tagging” creates highly descriptive tags according to a predefined structure tailored to your specific business. The tags used by AEM Assets Smart Tagging are those generally used by your employees and partners, as organized in your tag taxonomy.

In a prior post, I shared the following Adobe example of a well-structured taxonomy:

/subject/car/
/subject/car/sportscar
/subject/car/sportscar/porsche
/subject/car/sportscar/ferrari
…
/subject/car/minivan
/subject/car/minivan/mercedes
/subject/car/minivan/volkswagen
…
/subject/car/limousine

From “Implementing Rapid Omnichannel Messaging With AEM Dynamic Media.”

This is of course a very simple example. Your business may demand that you store more detailed information about each asset, including the horsepower of the car’s engine and the type of fuel the car uses.

Regardless of the complexity of your taxonomy, Adobe Sensei learns it and applies the appropriate tags to new content.

You can automatically run training workflows to make sure the system learns about the assets, folders, and tags already in use. To train your model:

Steps of the training workflow to train tagging model. Adapted from “Add Smart Tags to Assets in AEM.”
  • Create your tag structure.
  • Upload your assets.
  • Select the AEM command “Smart Tag Training.”
  • Create your model.
  • Select the tags to add to the model.
  • Select between 10 and 50 assets for training purposes.
  • Submit the model to initiate training.
  • Afterwards, review the generated report. The “Training Status” column in the report indicates whether the model is trained for a particular tag (flagged green), or whether the service is only partially trained for the tag (flagged yellow).
  • If a tag is insufficiently trained or does not appear in the report, add more assets and re-run the training.

Implementing AEM Assets smart tagging

While AEM Assets Smart Tagging will make a marketer’s life easier as your assets proliferate, your firm will need to perform advance planning to make sure your smart tags run smoothly. Among other things, your firm will need to plan your taxonomy, as well as all the configuration and implementation required to make AEM Assets function properly and increase your business.

You need an Adobe consultant who understands the technology as well as the needs of your business, one who has implemented business-critical solutions for other firms. You need to work with an Adobe Certified Partner such as KBWEB Consult.

Are you ready to deliver a personalized experience to every prospect and customer? If so, schedule a free, confidential 30-minute consultation with KBWEB Consult. You can book a meeting on our Let’s Talk page.