Seven Steps to intelligent Content
February 07, 2010
With increasing frequency, I have been returning to the subject of intelligent content. This time, we will consider by what steps content can be rendered intelligent. In preparing to take these steps, one has presumably established the rationale for the investment and the importance of this rationale will become apparent as we see just how involved this migration can be. If there is a central point to be made about increasing the intelligence of an organization’s content is that it is not only about the content – it is about the system within which that content will exist and the process within which it will evolve. Content can only truly be intelligent when it is an integral part of an intelligent process that is in turn supported by intelligent technology components and business capabilities. In order to achieve this type of alignment and integration, an organization must undertake a series of cyclical iterations, with each cycle progressively elevating and validating the intelligence of the content resources, technology components and business capabilities – as well as the process that unites them.
The Seven Steps to Intelligent Content
The process generally starts with the establishment of a goal – a vision of new capabilities. This vision will frequently emerge from accumulating experience and the quality of this experience improves as the intelligence of the content, components and capabilities increases. The framing of new strategic goals is hard work but it actually gets easier the more intelligent your environment becomes – and this because the quality of the feedback received from operations improves wherever intelligent content, components and capabilities are involved. This is a key way in which these steps are in fact part of a continuous cycle. For the purposes of elevating the intelligence of content, a key outcome from this step is the characterization of the type of content that will be needed to support the future vision.
Analysis entails the studying of the key entities in a domain and their interactions. What this directs us to consider is the fact that the analytical step is much more than what we would typically call “content analysis” although it definitely includes that as well. The analytical effort must look at the system stakeholders (including authors, users, customers, owners), content sources, historical usage metrics, revision and reuse patterns, and technology infrastructure elements. No one piece makes much sense in isolation. Now this is where one of the key challenges lies – analysis can become an overwhelming activity where the sheer volume of details presents an insoluble problem. This is where progressive iteration comes in. By selecting a specific set of content resources and associated processes, a project team can proceed through the following steps so as to achieve concrete results. And by doing so, the team is putting in place intelligent components that will provide enriched feedback from operational activities and this will make future analysis much easier. In a way, each venture into intelligent content is a bit like lowering a sonar buoy into an ocean of hidden processes and, over time, you begin to get a picture of what lies beneath.
Undertaken within a manageable scope boundary, this activity designs the content structures, information products and services, measurement systems, technology components, business capabilities, and integrating processes. Establishing a manageable scope boundary is in fact a critically important preparation – otherwise the scale of work that must be undertaken will be so great that too much time, and too many dollars, will be expended before any results are seen. From a business perspective, this is something to avoid. Even from a purely technical perspective, this is something to avoid because if operational feedback is postponed for too long then all the investments, from design to deployment, can become unhinged from reality and therefore be suspect. It is also important that the design tools, methodologies and notations being used are sufficiently lightweight and accessible to allow the team to work quickly and to continuously engage colleagues from a variety of domains and disciplines.
Of all the steps in this process, the most important is exploration. If analysis and design are advanced quickly, as they should be, then it becomes necessary that they be supplemented by exploratory investments in mock-ups, prototypes, demonstration scenarios, and initial working capabilities. Where intelligent content is the goal, I tend to emphasize this step even more than I would for any other type of technology investment. This is because intelligent content, and the associated components and capabilities, tend to be new for most organizations and there is a learning curve to follow that is best facilitated by hands-on experience. What is also a little different in my treatment of this otherwise familiar step is that I prefer to see it receive a substantial allocation of schedule, expertise and financial resources. I also like to see it taken well beyond prototypes and demos – to the point where the initial capabilities are being trialed by real users and tested by real business stakeholders. This is a topic I have addressed many times in the past including in a whitepaper on XML Business Templates.
Based on the foregoing a series of investments are made to transform content assets, technology components and business capabilities – all with the view to establishing a higher level of intelligence in how the organization as a whole performs. Each of these dimensions represents a challenge with the adaptation of business capabilities being the most daunting – in large part because the types of activities that will be introduced, and the types of new services that become possible, will be completely unfamiliar to the community of business stakeholders. The transformation of content assets will be a challenge all on its own partly because projects often assume that this is the only thing they will need to change in order to realize the benefits associated with an intelligent system. The transformation of content is also a challenge because historically, content has existed in forms, and played roles, that did not require very much intelligence at all. Transforming the technology components in most cases involves introducing new types of technology into organizations and this too can represent a substantial challenge because it will require that those responsible for technology management take the time to learn about a new class of technology tools. The importance of the exploration step becomes more clear as we see how it can be used to progressively tackle some of these challenges by involving the various stakeholders in ways that they can understand and evaluate.
In a comically large number of content management projects, validation is associated with the simple parsing of the transformed content assets against the content model that had been established in the design step. What is even more comical is the fact that many projects do not even perform this simple check. Validation, however, is a much more expansive concept than just parsing content against a given schema, although it includes that. By validation, I mean the complete validation of the target system which means that verification is performed on content structures, on the semantic details of the content, on the functional behaviour of the content within various technology components, and ultimately the performance of the overall system as it conducts various end-to-end processes. It is usually as part of this validation step that a variety of quality control mechanisms are introduced into the content processes so that the performance of the system can be metered and measured so as to provide useful feedback. Again, this clearly entails substantial investment and it should incline projects to tailor the scope of their individual investments so that the undertakings remain manageable in size.
The content, components and capabilities realized through the proceeding steps are at this point deployed into production. As has been coming into focus as we consider each step in turn, these deployments should be small increments that can be monitored, managed and measured individually. Ideally, there would be a regular stream of releases with this keeping an organization’s exposed information services in a state of continuous flux – what I guess we would call a “continuous beta”. The feedback mechanisms architected into deployed content, components and capabilities, as they proliferate, will increasingly come to provoke strategic insights and over time an organization will get better at working through these seven steps so as to act on those insights. In time, and perhaps accidentally, the overall system wherein content becomes useful information will become fundamentally more intelligent.
As things begin to fall into place, and the cyclical process of steps becomes something of an institutional habit, the ascent begins and the results that come into view will be understood as genuinely remarkable. On the more mundane level, the scoping of individual efforts is of paramount importance – the journey towards intelligent content, and with it intelligent components and capabilities, is a long one. The cycle of steps must be repeated often. Not that anyone has the option of sitting out and declaring that the old ways will do just fine. That would be a little like opting out of evolution and becoming some form of Relicosaurus that is oblivious to the fact that it is slated for extinction.
As Richard Dawkins once argued, in support of the compelling logic of the theory of evolution, small steps are the only way to climb “mount improbable”. It is better to succeed beyond all expectations within a limited domain, and thereby to provide an example for other initiatives to emulate, than to partially succeed against a much larger scope. This is why I have sketched out these seven steps as a sequence that is continuously repeated. As I used to tell my children when they were trying to move large volumes of toys from one location to another - many small successes is better than one big disaster.
Having not been involved with content as an *enterprise* concern, I can't tell whether this advice applies equally well to existing content and to new enterprises where the content has yet to be created. Is there anything you could add that's of particular relevance to the latter? Or is that another blog post altogether (perhaps yet to be written, perhaps existing elsewhere already)?
Posted by: Milan Davidovic | February 07, 2010 at 01:00 PM
In the case where content already exists, I have discussed this circumstance in more detail on the Content Wrangler site. In the case where an enterprise is brand new and a green field stretches out before the content designers, I have not addressed that circumstance. To be honest, I had not given it much thought because it is not a circumstance you encounter too often. Or perhaps I should say that when new enterprises appear they sort of just "happen" and when they pause to think about their content they find they have already managed to create a body of legacy documentation that needs to be considered.
If an enterprise did however pause to think about its content before creating a legacy collection, perhaps because it was going to be central to its offering, then that would be a very interesting circumstance. It would be interesting because there would still be challenges, but this time of a different type. The main challenge would be that there would be no current operations against which to measure and evaluate initial investments. I imagine that an incremental release strategy would still be advantageous and there would be potentially extra value in the metering and measurement services being grafted into the content, components and capabilities. So maybe this is a good thought experiment to run through because I suspect that kick-starting the feedback loop would represent a key challenge to overcome.
Very interesting idea, Milan. If I wake up at 03:00am this morning, I know who to blame...or thank...
Posted by: Joe Gollner | February 07, 2010 at 03:11 PM
Thinking about content in advance might be an interesting topic to develop and then take around to places where entrepreneurs get together to talk about starting up new businesses...
Posted by: Milan Davidovic | February 21, 2010 at 02:06 PM
This would indeed be an interesting topic to develop and then to offer to entrepreneurs for consideration. I again suspect that this would be a non-starter for most entrepreneurs but perhaps this should change. Certainly, ventures that are focused on "content" as an integral part of their value proposition should pause to consider how that content might be designed so as to permit as broad an exploitation model as possible. This same strategy would also show its worth if the venture wanted to try a number of exploitation models, with the content strategy being geared to helping keep the cost of each implementation as low as possible. The cost of experimentation and evolution would be thus kept low enough to be sustainble and this may make sense in these circumstances. Of course, the majority of entrepreneurs in my experience (or at least a significant proportion of them) are not really into strategies or even ideas per se - beyond of course the frentic promotion of the "next big thing".
Posted by: Joe Gollner | March 07, 2010 at 12:35 PM