Smart, Scalable, and Structured: Building a Future-Ready Content Framework

Transcript

Welcome everyone and thank you for joining today’s webinar. In this session, we will explore how to scale content operations using the powerful features available in AI assets. From leveraging metadata and modular content to harnessing smart automation tools, whether you are managing content at scale or planning for future growth, this framework will help you streamline operations and deliver consistently across channels. So let’s dive in.

My name is Ritesh Mittal and I’m thrilled to be your host for today. I have over 17 years of experience in industry working with Adobe solutions, especially Adobe Experience Manager, Adobe Analytics, Adobe Campaign and various other tech cloud technologies. I’m also an Adobe Champion and Adobe Community Advisor Captain. And I also host a YouTube channel, Tech Talk with Ritesh, where I share insights on various technology aspects.

Here is a quick look at what we will be covering in today’s session. We will dive into metadata and taxonomy concepts, where I will share best practices to help you structure your content for better discovery, reuse and governance. Following that, I’ll walk you through two various metadata strategies concepts, including demo, so that we can cover various concepts within AM assets. Next, we will move into the modular content using content fragments, how they will help derive scalability and consistency across experiences. Then we will explore the smart content management, highlighting features like AI powered automation, content insights that can streamline your operations. And finally, we will wrap up with a Q&A session. So feel free to note down any questions throughout the presentation. Let me set the stage with the very first generic example, that is the content chaos problem.

Let’s start with something we can all relate to, that is a bookshelf. We all read books. Imagine walking into a library where books are scattered everywhere. You see multiple copies of the same book. Some are outdated editions and others are personalized for different readers, but they are all mixed up. You are unsure where to find and what you need. You are more unsure what to put, where to put, or if it is something new. Now think of your organization’s digital content in the same way. You have got duplicate assets, multiple versions of the same content, and no clear process for updating or replacing them. When you need a specific version, say one tailored for a particular persona, you are not even sure where to look or even whether it even exists.

As more content gets created, the storage grows. Of course, it will grow and so do the silos, and that will be more challenging. Content gets harder to find, reuse becomes a challenge, and soon you are spending more time managing content than actually using it. And here’s the takeaway. More content does not automatically mean better content operations. Without structured metadata and intelligent systems in place, what you get is, yep, a chaos. And that is the problem we are here to solve.

Now what we really need is we need a content that is smart, scalable, and structured. And let’s see how can we achieve that. Okay, so to bring order to content chaos, we need a framework which is grounded on at least these three foundation pillars. Structure, scalability, and smartness. Let’s walk through what each of these means in the context of Adobe Experience Manager assets. First, we have structure. The backbone of organized content, and under this pillar, we focus on two critical components, metadata and taxonomy.

Metadata gives your content a meaning, a context, searchability. It is what makes content discoverable. Taxonomy brings consistency and classification. So your team are speaking the same language across regions, campaigns, platforms, whatever you think of. The second pillar is scalable, and here we bring in the power of content fragments, reusable modular content units that can scale across channels and audiences.

Now within this, we see two major benefits we can reuse, and which I will explain in further slides. We can, which will of course reduce the redundant effort. We also have had less delivery capability within content fragment.

It can enable seamless integration with apps, websites, and of course, emerging platforms. And finally, I’ll walk you through the smart features that we have. That is one of the pillars that gives content intelligence with building capabilities and smart features in your message. We can automate tagging, improve recommendations, and even generate insights that drive faster decisions and content creations. So together, these three pillars form the foundation for a content framework that is not just ready for today, but built for tomorrow.

Now let’s explore the first pillar, that is the structure piece, and we’ll cover the metadata and taxonomy. And there is often some confusion around this term, what really is a metadata. In simple terms, metadata is data about your digital assets. It describes, categorizes, and provides context to make your content searchable, manageable, and reusable.

Now think of metadata as the label on a file folder, or the description tag on a video. Now without it, content becomes just another file lost in the system. It answers the question, what is this asset? Who is it for? Where should it be used? And when does it expire? We all have that expiration date as well, right, with the assets. And in AEM, metadata is not just informational, it’s operational, because it derives search, automation, personalization, various aspects, even workflow efficiency. So when we say metadata is the backbone of structured content, we mean it quite literally. It’s what gives content purpose and power in a scalable ecosystem.

Okay, yeah, right, I mean, who needs metadata anyway? I mean, this little guy makes a good point, his mom didn’t tag her, lullabies.

And somehow she still knew them by heart, right? Moms are lovely and most intelligent, God creation anyways.

Imagine this, you’re walking into a library that looks like this. What will happen? I mean, the shelves upon shelves of books, but none of them have titles, no authors, no categories, no labels, just this books stuff.

Now someone says, hey, can you grab me a comprehensive guide on AEM content fragments? I just want to search this book. Good luck with that. Because this is exactly what is like working with the digital assets that don’t have proper metadata or taxonomy in place. Even the most valuable content becomes invisible. If you can’t find it, understand it or trust it, it’s the right version, right? So that’s why structured through metadata and taxonomy isn’t just a best practice. It is a mandatory, it is essential to scaling content operation. Without it, we cannot survive in the ecosystem.

Now, let’s look at this. Well, someone actually found the content fragment book. Now why? Because now we are in a structured library. Every book is labeled, categorized, and it is easy to locate. Exactly like how your content should be inside AEM assets.

Now let’s dive into taxonomy and why it is important.

Okay. Yeah. Love that. I mean, well, that’s one of the ways we can remember this word taxonomy. In your taxes on time. Okay. Keep moving. No, it’s not true. So don’t get, go to this, with this funny guy.

But taxonomy is essentially a systematic classification, a structured way to organize digital content based on clearly defined categories, attributes, or metadata.

It ensures that everyone on your team is literally speaking the same language, like I said earlier, and when it comes to content organization. So this shared understanding enables automated workflows to function smoothly, saving time and reducing manual errors.

Of course, and most importantly, taxonomy improves content in discovery by making it easier to find exactly what you need. So while also reducing duplication and various others, for example, if it is cluttered. So it will reduce that. Now, this is a typical example of a, what it takes. So we have a sitz and we have various, aspects of, for example, metadata is one of the taxonomy. Then we have tags, folders, workflow. Then we have renditions, et cetera. Right. We can further, classify them. So for example, within metadata, we might, we can have author detail, title, description, keywords, within folders, we can have project folders. And then for permissions, we can have various permissions, read access, edit access. So this is this, this whole example is showing that what a typical taxonomy will look like.

Now let’s see how do we view and update the metadata within AM asset.

So we go to dam, right? We click on any of the asset. We go to the properties and there we see all the metadata associated. We have various tabs. And in this tab, we can see all the metadata associated. We also see the smart tags here.

Now, metadata schemas are like that. You can define interface. What are the properties you want to define for your assets? Consider a common scenario in AM dam. We’ll see this, demo as well, where different asset folders require different metadata schemas. So for example, you have, maybe more than one folder, right? And, for these two folders, there will be assets and those assets should have different, metadata. Like one should have one property, but maybe another image or another asset will have another property. How can we do that? So we have option available. Let’s see this demo.

So what we are going to do first, I will, let me create two folders.

I’ll create asset folder A.

Then I’ll create asset folder B.

Right? So I’ve created two folders. Now I’ll create two different schemas, metadata schemas for it. So click on tools, assets. And then click on metadata schemas.

And these are the default schemas we have. I’m just picking on this and then image one, and I’m selecting this JPEG one. What I will do, I’ll just create a copy of it. So first let me edit it, see what are the different metadata properties we have in this schema. Now you want to create, just select the JPEG version and I can just copy it to create a new version. So let me create first metadata schema, which is for folder A. I’ll name it as, metadata schema A.

Just copy it.

And I will create another schema for folder B. So I’ll just rename it metadata schema B.

So now I have those schemas. What I will do, I’ll just edit them. Let me first edit the schema A and what I will do, I’ll just add a new property just for demo perspective. I’ll change it, the field label as property for folder A.

Yeah. So this property I have created only for assets within folder A and now I will edit the metadata schema for folder B.

And here I will add the label as property for folder B.

Now I need to assign these metadata schema for those folders. So I will go to assets and files and I’ll select one of the folder. Again, I’ll go to the properties and here, if we scroll down, we can see that there is a property. We can specify which metadata schema will be applicable for this one. So I’ll just select this one, save and close.

And I’ll do the same thing for asset folder B.

Yeah, I’ll assign metadata schema B for this folder. Now what will happen if we just, let’s save it first.

Now, if I go to folder A and I’ll just, I need to, sorry, I need to import a file. So I’ll just import a file.

And if you see the property of this image, you can see that this, yes, this property for folder A, which is metadata schema A, that property is adding here. Similarly for folder B also, I’ll just drag and drop a file.

Yes.

Now we can see that property.

Okay. So this is how we can achieve different schemas for different folders.

Okay. So now let me explain the scalable aspect of asset management.

Let’s begin by considering a scenario where content fragments have not been implemented.

So, okay. So we can see three key experiences here, the email campaign, the web portal and the search or FAQ section. Each of these requires individual attention and updates because we are not using content fragments though the content is, might be same. But we need to, because we are not using content fragments and it is different places. So we need that. And which leads to inefficiencies, inconsistency and a significant amount of manual work. In this case, the same content, whether it’s a product description, a promotional message or customer support information, it can be anything, needs to be manually upgraded and managed separately across multiple touch points. So what did we observe in this previous scenario? Content, despite being the same or similar, it is handled in multiple variations or versions across different channels.

And whether it’s on this website or in an email or in a support page, each instance of content is siloed and managed separately. Now think about this, what if a change is needed, but there is no time.

In a situation where update needs to be made quickly, this manual approach can become a And the result could be a delayed response, inconsistent messaging and the risk of publishing outdated or incorrect content.

To summarize the challenges, what we have just discussed, here are the key issues with using outdated content strategies. When content is not centrally managed, different channels may present conflicting messages. Without a centralized system, updates become tedious and of course time consuming.

And as the content is manually updated at several places, the risk of errors or missed updates increases.

With content scattered across different systems, tracking and managing becomes a challenge.

The effort to maintain separate versions of the same content drains valuable time and resources. And trust me, it really happens. The longer it takes to make updates across multiple channels, the longer it takes to bring content to market. And sometimes what happens in regulated industries, inconsistent content management can lead to compliance issues, particularly when content must meet legal standards across all channels. So these are the issues which we can face with outdated content management strategies. Now being from the implementation background, I can imagine a frustrated team member who realizes he needs to make the same update, maybe five different places. And this is exactly the frustration that comes with managing content manually across multiple channels and repositories. The lack of a centralized content management solution creates unnecessary work and confusion.

So now the question is how do we fix it? And the answer is, of course, content fragments.

With a solution like content fragment in AEM, we can centralize and manage content efficiently because it will ensure reusable and easily updated content across all channels.

So the question is what are content fragments? Content fragments are modular and structured content pieces that are without presentation layer. This means they contain only the raw content, like, for example, product description or any marketing text, and it can be reused across various channels and touch points. The key benefit is that the content is separated from layout and presentation, making it easier to manage and update.

You can create content once and deploy it consistently. This is my favorite line, actually. We create once and we can use it multiple times across different channels. That is the beauty of using content fragments.

Let’s take a look at the difference when content fragments can make on the same scenario. The same scenario, but this time we have implemented content fragments.

As you can see, the same content can now be reused across various touch points without the need to recreate it each time. All of these content is stored centrally and can easily be updated just by one update, and it’s sufficient.

Any change made to the content will automatically reflect across all touch points and channels. So, I mean, life is good.

So this is one of the benefits, like you create once and use multiple times. What are the other benefits? How can we boost collaboration and workflow with content fragments? This is one of the favorite topics within the organization when we manage the team, whether it’s a centralized team or a decentralized team.

Effective collaboration and streamlined workflows are crucial in any content-driven organization. With content fragments, teams can work more efficiently and deliver content faster.

Let’s imagine this. One team member creates the content and another uses it seamlessly across all touch points. We have that content authors, they create the content for us, or business authors, they create the content for us. This approach allows for faster content delivery without the need to recreate content for each platform, and you must be noticing the happy faces.

It really happens when we as a team get more time for other productive work. For example, some synergies, coffee breaks, because now we have more time for team building activities.

Let’s see the benefits of this workflow efficiency. Reusing content means teams can publish faster and of course it will reduce the delays. No more waiting for content to be recreated for different platforms. Teams can focus on their strengths, leveraging shared content for more consistent and faster outcomes.

Okay, another one. Let’s see this. How can we enhance the global consistency using the same content fragments? As organizations expand globally, maintaining consistent content across regions becomes increasingly important. So for example, your website or your web portal is not only expanded to one region, it’s maybe globally. So content fragments make this easier by providing a centralized way to manage region-specific variation. We have one concept called variation within content fragment. We can use that. Imagine you are running a free health checkup promotion for different regions, like for example, for Asia, Australia, Europe or North America. Instead of creating separate content for each region, now you can create a single content fragment and customize it as a region. So I mean, we will just create one content fragment and we’ll create various regions for that. This ensures that each geography gets a tailored message while the master content remains the same. So I mean, there will be various variations, but there will be master content. So when you will change that, you will get to know, okay, you will need to update the variation as well. And what will be the benefits of this global consistency? First thing is it will ensure a unified brand message across all regions and languages.

Easily manage regional and promotional content without redundancy and centralized content management reduces the need to create duplicate versions for different regions.

Okay, now let’s explore the smart content management aspect.

So we have discussed the structured, scalable, and now let’s see the smart aspect of it. Now in this section, let’s look at how ADM enables smart, fast, and context-aware content management through some powerful out-of-the-box features. So first we have smart tagging. This uses Adobe’s AI engine to automatically apply relevant tags to your assets.

So this not only saves time for your content team, but also enhances searchability and discoverability, especially when managing thousands of assets.

We just need to enable this service and if the service is enabled at the time of import, the smart tags will be created. Next, find similar image feature. It allows you to quickly locate visually similar images across your asset repository and we will see the demo for this. This is particularly useful for maintaining brand consistency or repurposing assets while reducing duplication and manual search effort. And finally, this is a new feature, variation generation. This feature helps generate multiple renditions of an image or asset. Automatically, it will generate automatically, of course, whether you need different crop sizes or aspect ratios for different channels or devices.

This automation makes it significantly faster and more consistent. Together, these capabilities, streamlined content operations, reduce manual effort and ensure your team can focus more on creativity and strategy rather than repetitive tasks.

So let’s see the demo of find similar image feature.

So what we will do? Yeah, select the image.

And here we have that option. Click on that and we can see that the image what we selected is in the search bar.

Right. And we are getting the visually similar images we are getting in the search result.

Right. We also have the filter option. So, for example, you are looking for any particular file type, for example, images. We can just filter it in the file type and it will give me the result.

Now, let’s see those.

Select an image and click on the properties. We also see the smart text, which we have seen earlier. These are the smart text created once the smart text service is enabled. They are automatically generated.

All right. So let’s bring it all together.

I have tried to explain it in three pillars, but they are actually sharing the pillars. These concepts are sharing the pillars. We explored structure, scalability and smart features in AMS and they are linked with each other. For example, with content fragments, modular design and strong governance, we ensure consistency, reusability across every experience we deliver and by tagging assets properly and building an adaptable taxonomy, we make our content discoverable, reusable, ready for omnichannel delivery. With Adobe’s AI capabilities from smart tagging and cropping to variation generation and usage insights, we are not just managing content anymore. We are managing it intelligently.

So whether it’s scaling campaigns globally or automating repetitive tasks, EEM assets help us work faster, smarter, and of course, with greater impact. That is what we want eventually. We want that the work what we are doing or the delivery, what we are finally delivering, that should have some impact, a greater impact.

Okay. So yeah, that is what I had for this session. And I encourage all of you to connect with me for further discussions on asset management or even any other topic related to Adobe Experience Manager or any other Adobe tool. Feel free to reach out through my YouTube channel or through LinkedIn.

Thank you so much for your time and I look forward to engaging with you further.

Amazing. That was packed with insights. Thank you so much for starting us off, Vitesh.

Thank you so much for having me.

Perfect. Of course. Well, let’s jump right into a Q&A. We’ll get started with questions that came in during the session, but please feel free to add more as we go.

First question I saw come in is, is creating multiple metadata schema a best practice based on your damn use cases? Okay. Very nice question. Though it depends, it depends on the use case, the business use case we have. So when we want to create the multiple metadata schema, so for example, you have various assets.

Assets could be your images or videos and you have different, you know, metadata for those assets. So for example, within your organization, which is an enterprise and you must be having various domains or various lines, right? And for those lines or businesses, you need a different searchable metadata. There is, that is one of the use cases. For example, there is one organization which is having IT department, HR department, and we want to keep those assets at one place as whole enterprise. So what we can do is maybe one of the folders is having metadata schema related to HR department, whereas the IT department could have another metadata schema. So that is one of the use case. It totally depends on what kind of solution we want to implement.

Gotcha. Thank you. That, yeah, that clarification is super helpful.

This next one, great question. I see this one frequently, but someone in the audience has seen experience fragments in addition to content fragments and they’re wondering what is the difference between content and experience fragments? Okay. I have heard this multiple times. So the main difference is, okay, we have this, something called a layout or presentation, right? Whenever we talk about any of the digital media, that is a combination of your data and your presentation layer.

Now content from fragment, you consider it is a raw data, just plain data without any presentation layer at all, without any layout. So if you have the requirement that you want to deliver this data for omnichannel delivery, the same raw structured data can be utilized for various use cases, maybe for mobile or desktop or anything, right? Then you will go for content fragments. It’s just text property and text value. Whereas if you want to have with layout, with presentation, with proper theme HTML, then you would go with experience fragment. And let me give you an example. So for example, for experience fragment, you can consider we have a banner, which I want to target maybe to websites and this banner or another example could be header or footer. We should be reused throughout your web portal. Then you should go for experience fragment. Whereas if you have, you are pulling some data, you have this data in your repository, just raw format data and you want to deliver it or headless is one good example. The same data should be replicated to mobile experience as well. Then I will go with content.

Gotcha. Thank you so much for sharing that distinction. Actually, and you answered the next question as well that was asking about different differences and benefits of using one versus the other. So two birds, one stone. Thank you so much.

Interesting question. Next, someone’s asking if can a folder have multiple metadata attached to it? Oh, that’s a nice question. I believe though I need to check because when we select it’s a drop down value. So I believe we cannot, but I need to check it. Well, not sure right away. Yeah, that’s a good question. Maybe a good candidate for the follow-up. Ritesh is beyond the Skill King session in September. So that is a great call out.

But that’s a nice question. Yeah.

Here’s one by using content fragments does repeated content across the site hurt SEO? Okay. So I mean, one thing I want to just point out content fragments will not impact your SEO at all because it’s just considered data that you are going to expose for various use cases. Now when we talk about SEO, SEO means like on your web portal, you have meta keywords through which your data is discoverable on Google, right? It’s nothing, nothing. Your content fragments will not impact SEO at all. So consider that you have same content fragment like being displayed at multiple places, but maybe on those pages you have different keywords, meta keywords. So of course the SEO will not hamper at all. I mean, content fragments is not going to hamper.

SEO at all.

Great. Thank you for confirming.

What about this one asking what are there, is it possible or best practice for removing duplicate images across a dam? Okay. Yep. Nice question. So in cloud as a service solution, we have a one service out of the box available, which is deduplication. And the behind logic is there is one property, which is SHA1, which is added to your asset. So what happens whenever you upload any image or this SHA1 property, which is very unique is added to your asset. Now, even if the name of your image is different, but the content is same, because AI is there. I mean, Adobe is using AI. So the same unique SHA1 property is added to your asset. Now what happens is when we enable this service, we check against each of those assets that if there are more images with same SHA1 properties, and those are tagged as duplicate images. Now you can go to the, I mean, author admin UI console and see if all those images, duplicate images are listed and we can take action accordingly. But here is there is a catch. If the asset count is high, then we need to make sure that it won’t impact your server, because what happens, I mean, if there are 100 images, each images will be validated against the rest of the 99 one. So it might impact server load as well. So we need to consider that thing in our mind.

Gotcha. Okay. Thank you. That is helpful. Yeah.

I’m not sure if you’re leveraging both of these tools, but someone’s asking if variation is the same as dynamic media.

So, so variation generation is a new feature. Dynamic media is there since long. Those are two different features. Dynamic media is that previously it was, you can host your images assets at a dynamic media. So it is just like offloading your asset load from your application, from your EM server. So the load will go to dynamic media. So that is one thing, but sorry, I just missed it. The other one was find similar images, right? Well, dynamic media. Yeah. So find similar images is the feature that you can search the same look and feel image in your system. So for example, you, I actually have given the demo as well for this. You select that image. And if there are other images which are looking like same, look, for example, you have selected the beach image and there are other images as well containing beach photo, et cetera. Then it will just search the images. Those are looking visually same.

Gotcha. Gotcha. Okay.

And now we have a couple of great questions referencing the smart tagging portion of your presentation. So up first, this person is asking is can smart tags be trained to identify specific products or faces? So the short answer is yes, smart tag can be trained and that there is actually a feature called enhanced smart tag. So what we need to do that first, I mean, the smart tag feature should be enabled on your folder. You should run that workflow that to enable, to add the smart tags. And after that, we can create our own smart tags and then we can add them to the existing small smart tags. What will happen? Once these newly created smart tags are enabled, those will be trained to exist with the existing smart tags. So yes, short answer is yes. So that is possible. For example, your use cases related to maybe some e-commerce related data or some banking related data, some retail related data, we can train the smart tags, the existing smart tags. And then once it is done, the model is trained. Then next upload will have your newly created smart tags to your assets added. Gotcha. Thank you, Ritesh. And this is kind of like a follow up question to that. Maybe the same flavor, but a different industry. But for smart tagging can be trained well enough to tag a manufacturing product, let’s say, like to tell the difference between having a shooting equipment between a boom lift and a bulldozer. Yes, yes, it can be.

Question bouncing back to content fragment variations. One of our attendees is still unclear on the benefits of making variations versus just making a new content fragment.

In that case, is it best used if the variations share a large percentage of the same material? So the example given is 80% shared content versus 20% unique to variation. Do you have any thoughts on that? Maybe we need to ask this question.

Yeah, maybe that’s a good candidate for having questions as well.

Awesome.

One looks like one more smart tags question just on its capabilities and how smart can it get. But this person is asking if they can identify people and tag their names specifically.

Can it be trained to recognize maybe the same people? Sorry, can you please come again? Yeah, so for smart tagging, can you train your smart tags to identify people and tag their names? Well, maybe we’ll recognize them. Okay, sure.

Okay, so maybe I will rephrase the question. The question is, we need to tag the assets with the person who has uploaded those images, right? So this is not out of the box available. But this is doable. I mean, we just need to write a new service and then use the asset. There are asset APIs available. We can do that. Gotcha. We just need to grab the author detail and just need to create a new namespace and then add it.

Okay, thank you.

Another one about do variations apply to regular pieces of content like copy, for example? Okay, yeah. Again, we need to park this question. I mean, this variation is a new feature. So maybe we can park the variation related question. So yeah, we’ll have this in September Connect. Yeah, that sounds perfect.

Another question here about, are you able to find similar images based off of file names or other data? So not specifically what the images look like themselves? So this find similar images is for visual check. But if the question is that, for example, if there are more than one image containing the same text, so it is just out of the box available. We just have the search feature available in AM console. Even if there is, we can just start typing in. And if the word is contained in any of the file name, those will anyways appear in the search result.

Thank you. Gotcha. Yeah.

We have time for maybe one or two more. Keep moving along. Question about metadata schemas. Can a metadata schema inherit from another base level schema, like the page template, for example? Yes, it can be done. Yes. So for example, you have a folder and then you have subfolders. So subfolders will, can you edit the metadata schemas of the parent one? Yes.

Gotcha. And actually, that is the whole beauty of using the metadata schemas. I mean, you have those schemas set on the parent one, and then your subfolder children will also inherit the same.

And same topic, kind of following up on that, with different metadata schemas, can you also modify the search criteria based on the type of asset? Sorry. I mean, again, can you please repeat, Will? Yeah. So with different metadata schemas, can you modify the search criteria fields based on what type of asset it is? Of course we can do that, right? Because I mean, whenever we implement search functionality, we need to pass the query parameter, the key parameter, like what keyword or what parameter we want to search. Even if we are using any enterprise search or within AEM, we are doing that right. We just need to, you know, for those properties, we need to create the indexes. And yes, I mean, then it’s fairly simple. I mean, the straightforward one. So for example, you have two metadata schema. One is for maybe property A and another is having property B. We just need to have those indexed, those two properties indexed, and then it will work.

Perfect. Super, super helpful. Thank you so much, Ritesh. And that brings us to all the time we have for Q&A for our first session. So Ritesh, thank you so much for getting us started with our first session today. Thank you so much, Will, for having me.