We’ve managed to make significant progress in developing technologies to deliver content to different devices and to track it. However, there’s another level of endeavor and opportunity that we’re missing, and I’d like to make the case for investigating what I’m calling content systems.
What am I talking about and why should you care? Simple, because the next-generation opportunities are going to be based upon this work, and you want to be aware of what the possibilities are so you’re prepared to fit content systems into your overall performance technology strategy.
Let me distinguish between two concepts: content models and content architectures. Together, these constitute content systems. Let’s start with content models, as they’re more closely related to what you already do.
If you’ve developed a course, you should have developed an introduction, a presentation of the concept, one or more examples, and some practice elements. You’ve likely had them on separate screens, and stored the components as separate files. But you may not have consciously separated them into their constituent components of a learning experience. Even SCORM, an interoperability standard for content, doesn’t care about that level of description. But you should.
In my definition, a content model is just a tight level of description around content. You probably do separate out video from audio from photos from text. But you probably do not separate out content in terms of other levels of description, and that’s what a content model is about. So, for example, you could describe a particular piece of content not just in terms of its format (again, video, document, etc.), but also its learning role (intro, concept, example, job aid, etc.), domain (sales, trouble-shooting), product or service offering (router 5/3XJ, homeowner without children, savings account), and more. Moreover, you don’t do this after the fact, but instead you create the model beforehand and develop to it. Figure 1 may help you see what I mean.
Figure 1: A content model provides multiple levels of description
For example, for a publisher, I articulated the learning elements (with greater definition around them, including sub-sections for introductions, for instance), and levels starting with independent learning modules, and then with others, and with an instructor, and as a course aggregate, etc. The point was not to tighten up design of those elements (though that was a benevolent side effect), but instead to facilitate offering different elements as different business models, and, of course, not to preclude the ability to create adaptive learning paths when their systems were ready. There’s more, but you should get the idea. It was quite an effort, with at least 68 different potential components, to indicate the level of specificity. And it is the basis of their existing product.
A content model is more than a template, although that’s one way to look at it. A content model is a template, but it’s for more than a course, and it’s more about detailing the semantic roles of elements rather than how to design each element. The tight definitions support accessing the content by definition, not by link, and that’s a technical advantage.
Content models go further, of course; performance support tools can and should be part of your content model. Your organization might not need as rigorous a model as a content publisher (unless, of course, you are one), but when you add in the possible forms, your model may be similarly faceted.
And content models are very much about separating content from format. It’s about stipulating content in ways that allow the content to appear flexibly in different channels, as well as articulating what short versus long versions are, what can be stripped out programmatically, and more.
Let me make a diversion here. From the point of view of content systems, we don’t distinguish passive content such as documents, pictures, audio, and video files from interactive content like simulations and games, interactive checklists, etc. Although they might report back to a system, they’re developed, reside as files, and can be distributed. However, from a learner or performer perspective, perhaps more appropriately from a cognitive perspective, they differ hugely. Today, and just for now, we’re not going to care much, but don’t let that allow you to think that a knowledge dump is a worthwhile component of a learning experience. (Editorial soapbox off.)
So why, I’m sure you’re asking, should you go through this trouble? To answer that, we need to talk about content architectures.
Imagine that you have two different individuals, with different amounts of experience, facing the same problem but equipped with different devices: one has a tablet and one has a pocketable device. Frankly, I think they’d likely want different information presented. Say they’re facing a malfunctioning piece of equipment out in the field, and need to troubleshoot the problem. For the one who has some experience with troubleshooting on a similar device, but no experience in this device’s component systems, I’d probably err more on the side of explaining the underlying technology, whereas for the person who knows the component systems but doesn’t have a lot of troubleshooting experience, I’d probably focus more on the diagnostic process. And for the person with the tablet, I might show a video of the process, whereas on the pocketable I might just choose to provide a schematic or a checklist.
Similarly, the person taking a refresher at home on their desktop might have a different learning experience than the novice taking a course for the first time on a tablet. This is the customization and personalization that we now have the technological capability to do, and have the strong arguments for the need to do so. How do we make that happen?
In one instance, we took an existing process for developing manuals to accompany complex medical devices and modified it to populate both the manual and the onboard help. This required writing into a particular model and it also required writing rules pulled from experts to populate a context-sensitive help system. When someone asked for help on this system, we could offer direct links for the things users were likely to need, while still backing it up with a full search and browse capability.
We can, or already do, have rich descriptions of context. We have models of the learner’s competencies from an LMS. We can tell their devices by the way they access the system. We can know what they’re doing from their calendar or location or both. So we have the information to customize the presentation, but we can only do this if we have a similarly rich description of the content.
Well, we could hardwire the respective solutions by hand, but this doesn’t scale well. When you look at Amazon and Netflix recommendations, there’s no one looking individually at your behavior and personally crafting your recommendations. Instead, we’re in the age of system-generated content: Web 3.0. (Web 2.0 is user-generated content, and Web 1.0 was producer-generated content).
This is what a content architecture does: it pulls out content by description, on the basis of rules that stipulate what information to use to do so based upon the context, and it creates custom content on the fly. You can be delivering on the dream of “the right stuff” (the right information at the right time to the right person at the right place on the right device) to start optimizing your employees’ performance.
Over a decade ago, I led the development of the Intellectricity™ project that precisely focused on pulling together content on the fly. In that case, we were doing it on the basis of who the learner was and how they’d recently performed (leading to my persistent problem with existing learning style instruments, but that’s another story), but the same model can be used here. We had rules that looked at the learner and context, and recommended the best next learning object for them. (Figure 2)
Figure 2: The Intellectricity™ project pulled content together based on both the learner’s recent performance and on context
So, content architectures are the systems that pull this content out as needed. Note that the content does not have to be pre-developed; systems now exist that can parse your content from the existing formats and make it available at a more granular level. These constitute content systems as well, where the content model is the descriptions resulting from the parsing and the architecture is the search engine on top of this.
Why care now?
This may be well beyond where you’re ready to go now. I get it. And that’s OK. However, you need to know this for two reasons: where you’re going, and what’s coming. Oh, and the technology is already here; we’re capable of delivering on this now.
What’s coming: more and more people are going to want the results of finer-granulated searches. The evidence is already clear that knowledge workers spend a significant portion of their day looking for information, and too often unsuccessfully. And as folks start having experience with smarter systems, they’re going to be increasingly demanding such capabilities.
Where you’re going: toward thinking more strategically about your use of technology to support organizational performance. Given that we’re not seeing less competition in the market space, you’re going to need to. And you can. There’s a lot more: e-community, performance support, and more, but content systems are an underpinning that will grow increasingly important.
There’s more yet. If you’ve heard about big data, mining, or analytics, this is a key ingredient. Simply stated, by tracking what people use and the outcomes, we can start finding emergent patterns that indicate what works, and start using them to advance our approaches. To do that, we need to have that tight definition to support making valid inferences.
And if you’re ready to think about mobile, you really ought to think about content systems. The effort invested even in limited content models pays off hugely in being able to deliver in ways that may not be affected by the device you’re delivering for, and that makes a huge difference in the opportunities available to you.
You may not be ready, but you need to have the concept on your radar, as the opportunity and the need are increasingly converging. If you’re looking for a competitive edge, don’t be content, and start thinking about being systematic with your content.