Before taking some holidays, Julian Wraith asked the “CMS gurus” (nice to be considered as a “content guru” ;-) ) some answers to one simple question:

What is the future of Content Management?

(Sorry, long blog post in front of you: time and aspirin – should I say Tamiflu – needed ;-) )

Omnipresent, Unified, Pervasive Content (courtesy of Laurence Hart)

I will directly reforwarded you to the excellent blog post of Laurence Hart. There is nothing more to say about it. Get, Link, Merge ECM Systems into one single omnipresent Content Cloud. This is a nice long term vision of the CM world. But finally doesn’t it sound a bit like the Linked Data paradigm and quite close to the Semantic Web vision of Tim Barnes Lee?

A Common Content Information Infrastructure (CCII) will become the new standardized middleware stack

There is currently an unclear separation between applications frameworks and content infrastructure. But at the end of the day everything is content and every application has first to deal with content items rather than with processes, states, UI components or other application oriented paradigms. We recently saw the emergence of new content enabling application development frameworks such as Apache Sling in favor or more “old-styled” web application framework such as Struts, Turbine and other kind of Web 1.0 application development frameworks.

Such content enabled development frameworks have just started to emerge. Their limits and their scope is quite unclear: Excepted the core common library services (e.g: some versioning, check-in/out; querying;…interfaces which are today considered as a commodity) lots of questions are still open. Should they also natively integrate and support some business processes engines or some ruling system? Certain transformation services? A common search API? A standardized reverse cache proxying system? What is exactly the scope of services concerned by such a content infrastructure?

This Information Infrastructure stack then looks like an aggregate of several existing pieces of technology (e.g: concretely speaking it looks like a combination of JCR, OSGi, BPEL, CMIS,…and all others now more and more common bells and whistles).

Such a Common Content Information Infrastructure (I will call it a CCII) will certainly become sooner or later the underlying core technology for any new CEVA (Content Enabled Vertical Applications). But I agree there is still a long road to go.

We can then easily forecast two types of solutions:

- Some free and commoditized Open Source CCII stacks
- Some commercial, proprietary and high-end solutions with strong value added offerings

Ideally speaking all of them should be cloud-ready and be able to work either in a private or in public cloud environment. We could also imagine that an actor such as Google might want to enter and disrupt such a market by providing its own new cloud focused content infrastructure.

In the open source world, in order to be able to reach a critical mass of contributors and to be able to really leverage and commoditize such a large and complex stack, only most important projects will succeed. I could see two of them: Apache and perhaps RH JBoss (if they get enough leadership and vision to evolve their current middleware offering into that direction). It will be then interesting to know how will evolve actors such as Nuxeo or Alfresco in regards of their own CCII stack. Should they continue to offer both a CCII stack and some CEVA apps or should they rather contribute and outsource their CCII stack to a 3^rd party (e.g: recent Exo Platform JCR and Portal contribution to JBoss).

On the high end side this trend began with the acquisition of Documentum by EMC a few years ago. We now see some proposals to suggest, for example, using Documentum as a possible high-end CCII substitute for MS SharePoint at the place of the default underlying MS SQL Server storage mechanism. But Oracle, the information company, or IBM, if they succeed to leverage their FileNet acquisition and to merge all their content initiatives under one single umbrella, should also be able to rapidly provide such a new global CCII platform. I of course do not include Microsoft which already provides such a whole integrated CCII stack and which is rapidly reinforcing it as part of its .Net and other kind of middleware extensions.

Ideally speaking such a Common Content Information Infrastructure should then be as standardized and as switchable as possible for all the future CEVA applications running on top of them. It would be a shame to re-start some strong vendor lock-in scenarios with some IBM CEVA apps, some Oracle CEVA apps or some Microsoft CEVA apps. This is however currently the case and the risk to see such a scenario happen in a near future is quite high.

There was a strong trend among all actors to standardize the J2EE middleware stack a few years ago. The next challenge will be now to standardize such a Content Information Infrastructure stack. Call it J3EE, CMIS 2.0 or CCII 1.0. In all the cases lots of questions are still open today. Who should do it: The JCP (hum I mean this is more now the OCP: the Oracle Community Process) such as it was done for the java application server? The OASIS Group; AIIM? Another vendor and technology-neutral content-focus consortium? Will CMIS 1.0 be enough to standardize a first version of this CCII stack. Certainly not. It looks like more similar to the EJB 1.0 spec: a set of core standardized API but which are not really sufficient to really fully develop any vendor-neutral CEVA apps. Will there be an interest for a CMIS 2.0 to pursue the content interoperability effort or will this process stop in favor of a more conservative vendor lock-in position after the 1.0 release? Let’s see.

In all the cases, once the CCII in place, and if the content industry can reach such a milestone, CM actors will then be able to outsource their entire core content services to a standardized content middleware stack and only focus on working on value added content applications. How much energy is today spent reinventing the wheel and redeveloping yet another versioning system and yet another content definition mechanism?

Every CM vendor should then question itself about how much he wants to continue to invest into its own proprietary or standardized but homegrown CCII stack versus leveraging an existing or in progress CCII stack and to focus only on the atop CEVA solutions.

Time for composites

Boundaries between a WCM, a DAM, a Portal, a Social platform or an ECM tend to become more and more vague. We recently saw a trend to re-focus and re-categorize products around solutions rather than by technological compliance (e.g: new pillars for SharePoint 2010).

For example for years there was a battle in order to know if it was better to have an underlying Portal and an embedded WCM on top or a core WCM architecture empowered with some portlets. Such a debate is now obsolete.

Any web page or piece of information should now contain an aggregate of several sub-systems:

• Some texts coming from your WCM
• Some pictures coming from your DAM
• Some gadgets apps hosted by Google
• Some attachments coming for your ECM
• Some UGC content areas empowered by your Social Collaboration tools
• Some shared email content you want to republish
• Etc…

You only have to consider the enthusiasm raised recently by the announcement of the new Google Wave service to understand this need of simplifying and easing aggregation of pieces of heterogeneous information and other kind of micro-applications into one single page (or “wave”).

WCM first mission statement was originally about easing or even fully replacing the job of WebMasters and letting companies delegate page and content creation to non technical authors. In a Web 2.0 world such a business practice is today a commodity and anyone is able to publish content on the web in a few couple of clicks. WCM first mission tends then to focus on derivative needs such as better accessibility enforcement, improved web analytics figures, stronger integration with existing legacy systems, etc… But these are only secondary pillars.

However it is still quite hard to easily create and manage “composites” (composite pages; composite spaces; personalized composite areas; composite apps…). This still currently requires lots of integration time. Composites also require a better underlying architecture focused on standards, content reuse and other kind of content relationship engines and consistency rules. Last but not least, in order to be useful, composites should be easy to use.

The apparition of some “Mashup Servers” (e.g: JackBe) were precursors of this new generation of composites but were still too difficult to use. Today combined with WCM features (multi-devices management), some Web 2.0 technology (e.g: inline editing), the strength of ECM back-end (core library services) and new interoperability standards such as JCR or CMIS API, the development of a universal content composite aggregator will certainly be a turnkey feature for any CM solution.

Enough technology; time to turn them into effective solutions

Users now wants solutions to their problems, no more a state-of-the-art technological platform composed of hundreds of (most of the time useless) checkbox features and API.

The CM world was driven by techies for the last 10 years. XHTML, RSS, AJAX, JCR, CMIS, XML, RDFa, BPEL or even at a higher level RM, ECM, WCM, PIM, DAM, … are the terms and paradigms that a basic knowledge worker has today to understand in order to start working with any Content Management solution. If it goes on, one will soon need a Phd in CM in order to start understanding how to deal with his documents and other piece of information he has to deal with on a daily basis.

So it is now time for solutions to take precedence on technology. For instance an analyst in a large bank may want to ease the collection and sharing of information about his portfolio of watched companies. He may then start looking for an application to help him better do his job. But he would often rapidly be conducted nowadays to start a new ECM selection project, associate it to a Portal framework and finally empower the whole by some Web Scraping utility. The whole solution will of course require a huge quantity of integration and customization time and brings lots of maintenance and support issues. Meanwhile such a knowledge worker was just looking for a ready-to-use solution which was gathering pieces of content available out there, share them more easily with his colleagues, add some value to them and push the most interesting ones to his boss.

A better separation between CCII and CEVA types of vendors will certainly help boost the emergence of such ready to use content enabled vertical applications to the benefit of everybody. On one hand IT departments will then be able to focus on selecting the best Open Source or proprietary horizontal CCII platforms with all the key underlying technological criterias they need (robustness, scalability, volumetry, performance, programming language…). On the other hand functional users could focus on selecting the right CEVA solution which best fit with their business needs knowing the fact that their data will be securely managed from a state of the art manner in the underlying core CCII platform.

Currently selecting a CCII and a CEVA apps is still tightly coupled in the CM world (and especially the WCM one). It would be so simpler to get a better segregation of duties between CCII infrastructure and CEVA application vendors.

From Content to Information Management (courtesy of Lisa Welchman and Tony Byrne)

Until now we were mainly speaking about content management (content objects storage; aggregation of content items into a composite; content in the cloud; interoperability of content items;…).

But Content is not Information. And Information is not Knowledge. So there is still a long road in front of all the CM vendors to first become some IM vendors (Information Management Vendors) and then perhaps some KM vendors (Knowledge Management Vendors - in the noble sense and not the already obsolete and old-fashioned KM paradigms).

It sounds a bit like Enterprise 2.0. And E2.0 sounds like below the Web 3.0 which is basically the Semantic Web which was initially already planned 10 years ago to become the Web 2.0. So the already old-fashioned Web 2.0 world of AJAX technologies looks like indeed a version 1.5 of the Web and the new E2.0 trends a kind of intermediate 1.6 version adjusted to the needs of organizations. The real Web 2.0 (now the Web 3.0 and probably the Web 4.0 or even more) will still have to emerge. This one will have to turn content into knowledge. Lost in the numbers? No worries. What is important to remember is that CM vendors will still have lots of work in front of them to find ways to turn their content items into valuable, shareable, collaborative, moderated, indexed, categorized and secured information before even thinking about turning them into effective knowledge items.

Lisa Welchman recently also blog about it and so do Tony Byrne from CMSWatch a few months ago.

So what’s missing first in order to move from raw content to valuable information: mainly a better unified content streaming flow.

Such a content flow should basically:

• ease the collection of raw content spread-out on various sub-systems or on the web
• better store, classify, moderate, share, comment and retrieve such content items in a standardized manner in order to turn them into valuable information
• and finally find ways to push or pull such valuable information to the right person on the right pages, on the right workspaces or on the right device (be it a mobile, a RSS feed or a rich client à la Seesmic).

Combined with the aforementioned facilities to rapidly and instantly develop composite pages available on multiple devices and for different channels and you get the next CM milestone.

This is of course on a short term basis. On the long run a “Universal Secured Content Information Cloud” will perhaps be available to anyone at anytime (cf: first point). Such a need to duplicate or federate content will then vanish in profit of improved rights and identity management to let you access to your content according to your current authorizations when you need it on the device you are currently using.

Your Personal Digital Filing Cabinet for your entire Lifecycle

One of the next key challenge for a CM system is to improve the notion of “personal digital filing cabinet” (vs some “Corporate and Shared Knowledge Spaces”) and to let knowledge workers decide which data or information belongs to them and which information is shared with others or directly belongs to the company.

Currently most users are still in the first phase of capturing new digital assets. Be it by using their newly acquired digital camera, by scraping or clipping web pages, by creating some new documents or by adding some comments to a discussion forum online.

But how could an employee keep track of all his “records”? Where is his universal federated all-life-long content activity stream aggregator? Which information belongs to him and which one belongs to his current or previous employers? What if an employee is leaving the organization he is currently working for? Which information can he take with him as part of his personal information wallet and which one are considered as being some collective intelligence belonging only to the organisation.

A few years ago everyone had access to hardcopy photocopies, post-its or notes on pages stored in some filing cabinets. Your personal knowledge was secured and safe. This is not the case any more in a digital world.

One of our customer, a major European university, was for example asking to us about how a student would be able to live the campus after 3, 4, 5 or even more years of presences and to take with him all his digital assets. Not so long ago, students were used to make kilometers of photocopies of all their courses, researches, homeworks, etc… Currently more and more thinks are done from an electronic manner. And once the student usually leaves the campus he is also suddenly loosing most of his “knowledge” by loosing access to his university login.

Of course the same is true for lots of occupations: lawyers are used to annotate their core law bibles and could not live without it whatever their current employer is. Same is true for a biologist, a physician or nearly all professions.

Sounds a bit like Surfulater, Evernote & similar offerings? Yes and no. Such solutions look like more unofficially ripping web content which could cause severe confidential issues within the context of a company. And they mainly focus on unstructured data and not on more precisely defined “content objects”. But yes, such a notion of personal filing cabinet which one could keep all his life long will certainly be a hot topic for the CM industry and is of course closely related to some authentification and authorizations rules coupled to better cross-vendors content platform interoperability.

 How easy it would be to take with you all your latest web researches, notes, documents, emails, import them in your new job and start working with your new colleague. Briefly speaking one single but interoperable digital cabinet which could be easily (un)mounted to and from any CM system?

Of course if access to your entire spectrum of information is transparent, available in the web cloud and every piece of such information automatically benefit from standard library services (versioning, file plans…) of course this could drastically helps the development of everyone personal digital filing cabinet.

For the moment let’s go back to the photocopier room!

Start solutioning the Information Overload issue

“The right information to the right person at the right time”. I remember one of my university teacher who was already speaking about that back in the 90ies and who was telling to me that search engines, taxonomies and other information pushing technologies would have solved such an issue in a few couple of years. We are now reaching 2010 and we are still far, far away form such an utopia.

For instance, how much time do you now spend (loose?) per week on Social Networks in order to try to parse valuable information from spam and to get updated? Which quantity of activity streams do you have to read to get one single valuable entry? Multiply now this number by the quantity of your colleagues which are probably doing exactly the same thing than you and following more and less the same “friends”? Certainly not the most efficient process we can imagine.

The problem is symptomatic of new social networks but is certainly not new. Enterprise Information Overload (internal email Overload; saturated shared network drives; uncategorized records;…) are today common practices in most companies.

Moreover information overload is now not only impacting top executives but is rapidly reaching standard employee. And such a content flood will certainly not stop in the near future. In a digital world there is no more this kind of automate paper based archiving process which consisted to archive in the cellar at the end of the year all documents related to the previous year. Digital content (we can not speak about information any more) is now visible for years by anyone. And RM solutions are far from being intuitive enough or widely distributed to be the next year ultimate solution. Probably 99.9% of the digital content available today is not regulated by any file plan and will stay in such a state for the next 5 years.

Is it time for Content Management vendors to better address such an information overload issue? Yes, certainly. How? Perhaps by trying to better integrate some semantic web technologies. But such a process will also take several more years.

Thanks for all the ones which had the patience to read this blog until here. I certainly have a lot of other points to mention, but I currently lack some time to expose all of them. Stay tuned and you will be able to read them soon in future posts.