Content management made strategic

DEFINING WHAT a content management vendor is can be a perplexing task. The relative immaturity of the technology combined with fast-changing business requirements have prompted document management and e-commerce platform vendors to tout different feature sets, ranging from workflow to personalization.

Despite these fluid boundaries, enterprises are increasingly looking to purchase software packages, rather than build their own applications, to handle the explosion of content they face. InfoWorld's Executive Editor Martin LaMonica spoke to B.C. Krishna, the CTO of Burlington, Mass.-based content management vendor OpenMarket, about how the effective use of content is in and of itself a strategic application.

InfoWorld: Many content management applications first appeared on public Web sites. How has that evolved in the enterprise?

Krishna: What we find is that people are building applications derived from the need to manage content. In other words, content management is not the end goal. It's the first step in the process of putting together any business application.

There are two approaches. One is, "I've got bunches of content to manage and I need to go home early." The other approach to managing content is, "Information is central and strategic to me."

[For example,] if I were a publisher, my business would depend upon the proper management of information. As a business, my end goal of managing that information is to ensure that I have my brand represented -- the proper content represented. My business depends upon advertising, so I need to make sure that the content works for my audience and the advertisers go there.

We follow the second approach to content management, selling to companies for which the management of information is a strategic value proposition. We do not fall into the category of companies for which the immediate pain point is managing Web site pages so that people can go home early.

InfoWorld: Isn't it typical that companies have several systems to manage their content?

Krishna: The problem is that a company such as Chase has thirty, forty, fifty lines of business. Organically what used to happen is every line of business had its own approach to following the content management problem. So, basically, they said, "OK, well, I've got a problem. I'm going to get this system to manage it. I'm going to get that designer to build templates for me. I'm going to get this design house to help me do that stuff. Well, I need to have a work flow, so I'm going to put together this set of steps in my work flow, and I'm off and running." And so, what happened was, during the course of a year or two, every line of business had its own approach. They had their own designs, their own infrastructures, their own workflows, and it was quite a big mess. It was not a situation with a common look and feel, a common workflow, a common infrastructure, an efficient use of resources and skill sets, in order to fundamentally manage the process, not on a departmental basis, but on a global companywide basis. That class of content management problem is something that involves multiple divisions, size and scale, and in many, many cases, it involves global reach as well. In the case of Chase, they're using this approach to content management to also now run a Brazilian Web site, a Spanish language Web site, and a Japanese Web site. So, it's really a sort of content management in the large. If you were to take the departmental approach [in a corporate merger], then you'll have an even more severe problem in terms of being able to organize content across all these entities. So I think that it's our approach, and it's our feeling, and we feel quite strongly that it's important to take much more of a strategic approach to the way in which they think about the content management problem. We elevate [content management] up the organization so that it can be addressed at a strategic and corporate level rather than at the departmental level."

InfoWorld: Are companies addressing the issue at a high level and corporatewide?

Krishna: The good news is that everybody recognizes that they need to have a content management system in place. The bad news is that content management has not permeated through the organization -- up the organization -- and has in general happened only in a relatively small number of the leading-edge enterprises. I'm afraid that the industry has not gone far enough to recognize this as a strategic issue.

InfoWorld: How do you define the process of managing content?

Krishna: [We define it] in terms of content management and delivery. In other words, it is the whole end-to-end process of acquisition, collaboration within the enterprise to be able to manage the content, and then the dissemination of that content to the Web site, to partners, to channels, to syndication, and to other organizations. It's a large, end-to-end problem.

[You need to] disseminate information out to the Web site so that it can be read by the audience. But very often it's the case that that's only one place where the content goes. It's sometimes the case that you need to distribute it out to a PDA or to an e-mail blast or to a wireless device. That is the scale and scope of the problem.

InfoWorld: What are the core technology components that people need?

Krishna: The core components must have an open, flexible mechanism to manage information architecture so that businesses can describe lots of different classes of content that can still be fed through a common operation, such as version control, editing, workflow. But even more importantly this [should] be located on top of enterprise-standard infrastructure, such as [Java] application servers.

InfoWorld: What are some of those areas for which you see real business needs arising over the next year or two?

Krishna: The early adopters in this marketplace are clearly companies such as financial services firms and publishers and so on. But we haven't seen for instance the transportation marketplace needs. We haven't seen, for instance, what the health care factor needs. Their requirements are quite unique and quite different; they move at a different pace. What does an airline industry need in the area of information management? Well, certainly they need to put up Web sites, but that's not the class of problems that we're talking about. How do they, for instance, make sure that they deliver a greater service to everyday passengers who have basic problems.

InfoWorld: Will XML help portability of content between different companies?

Krishna: Absolutely. There's no question about it: XXML is important to the portability of content. But let's not kid ourselves. XML by itself is not sufficient. It needs additional standards. The best way is to use an example. If I were a [financial services] firm sending information to you \#209 how am I going to describe the information? It is not sufficient to simply say that I'm going to put some tags around it. What tags are they? And what do they mean to you? If I say "price," what does that mean to you? Do you expect price in dollars? Do you expect price in cents? Do you expect price to be padded out to six decimal places? Those are the kind of standards to be agreed on.

InfoWorld: How are technologies such as XML and Java changing things?

Krishna: I think Java and XML are technologies that co-exist. If you think about the problem space, there are two things that we have to really pay attention to.

One is, how is information structured? Any need to managing content requires the content to be properly structured. Before XML...the only way that people really thought about structuring information was to think about the capabilities of a relational database or a large database. So XML offers a great, great, great opportunity to really structure information properly. This is not a new idea. SGML was designed to serve that same purpose, right? But unfortunately it was too complicated and didn't get widespread use.

At the same time, this information always has to be put to some use. There is some application behind that information, and that application requires business logic. That's where Java comes in.

So XML is part of the structure, and Java is part of the application.

InfoWorld: What is the appropriate architecture for a content management system? Is it a centralized model with all of the content in one place, or does it make more sense to have a more distributed architecture?

Krishna: I think that centralization is required, but not because centralization is good. Centralization is required because it forces standardization. Once you standardize on an approach, you can then replicate that out to different divisions within the organization, for instance.

As we all know, centralization comes with its problems, and now you're going to one central place for all your requests, all your requirements. But essentially, as the organization grows, you run into stress points within that centrally organized and managed infrastructure.

InfoWorld: Do you see an opportunity for peer-to-peer technology here?

Krishna: I think it's early, but the p-to-p has promise. It has promise indeed in the content management area. We've got our own sort of brainstorming efforts in that area. P-to-P is, unfortunately, as the name suggests, decentralized in the extreme. And it is a great way for individuals to share information. But I just wonder about the lack of control it suggests. In an enterprise business environment, people do need to have some level of regulation and control. It doesn't need to be Communist-like control, but certainly some level of control is required and desired.

InfoWorld: What's the best way to think about content strategically?

Krishna: Do you really know where your information is today for your e-business? That is the core of the question. Do you know where your information is for your e-business across the enterprise? Do you have a strategy to do that? Do you have a strategy to manage that? And the answer is not to say, "Well, yes, I've got databases" or "I've got data warehouses." That is not the answer.