Dec 06, 2004
On the last day of the Gilbane conference, Erik Hartman demonstrated his new site Content Management Overview. The initiative is based on CMSML and has information on 92 content management systems. The site will allow vendors to submit information about their product (moderated by Erik's team) and is designed to be used to narrow down the field of products considered in a CMS selection. I think the tool has the potential to be very useful, especially if the vendors and OS projects participate and there is reason to believe they will. Erik said that he frequently is asked by vendors why their product is not included. He showed an email from a Vignette executive as example.
During his presentation, preceding Erik's, Bob Doyle handed over the management of the CMSML repository over to Erik. CMS Review will continue to support the CMS comparison tool using data from Hartman's site.
I would encourage people to use the Content Management Overview site. If you disagree with any product claims, you can add comments directly on the site.
Dec 06, 2004
I just heard that Artemis Software is distributing an out-of-the-box integration of Microsoft Content Management Server called MCMS.RAPID that is to be installed on top of MCMS 2002. The interesting thing is that, MCMS.RAPID is an open source project. Read the license here. I was not able to find an announcement on Microsoft's website.
The email announcement that I received made it seem like this was an Open Source CMS. But in reality it is just a set of customizations that provide a higher starting point for an MCMS development project. You still need to purchase the full MCMS product from Microsoft. I have heard criticism of MCMS for being very expensive to license. This set of extensions might reduce the total cost of ownership and make MCMS a more attractive option. Still, as I noted in another post, Microsoft does not seem to be paying much attention to this product so I would carefully consider purchasing it.
Dec 01, 2004
The first day of the Gilbane Conference had a very interesting Keynote panel of industry analysts who focus on Content Management technologies. The panelists were:
-
Steven Ashley, Senior VP, Research from Baird & Co.
-
Joshua Duhl, Research Director of Content Management from IDC
-
Hadley Reynolds, Vice President & Research Director from Delphi
-
Kyle McNabb, Senior analyst from Forrester Research
-
Alan Pelz-Sharpe, Vice President, North America from Ovum
The panelists were asked to speak for 5 minutes on something that they think the audience should know about trends and direction of ECM. Here were the highlights....
Joshua Duhl kicked it off with the bold statement that ECM is a myth. There has been a big consolidation of companies (and their capabilities) and there is more the come. Vendors will say that they have it all working together but they don't. This idea was echoed by Alan Pelz-Sharpe who said that companies have so many forms of content and processes that it is unrealistic and impractical to manage it all within one system. On the other side, Kyle McNabb said that it is very real and focused on the idea of "one throat to choke." In the Q&A there was a similar topic about best of breed vs. ECM. Kyle McNabb's recommended big ECM vendors who have complete solutions with best of breed components. It think the general consensus was that ECM is a vision (not a product) worth pursuing for some companies if the cost/benefit is right.
Steven Ashley, who focuses on financial analysis of CM companies, made the point that all the money has been, and continues to be, in Document Management. He pointed to three huge deals that FileNet made ($10MM, $9MM, and $8MM). He said that Documentum is doing equally well. Oracle and Micorsoft are preparing to compete heavily in this area. The interesting thing is that an informal audience poll said that most of the audience was focused on Web Content Management.
Most companies under budget their CM initiatives. They should expect to spend 2 to 3 times the license cost on services. The established ECM players will have an advantage in integration costs over the platform players (Oracle and Microsoft)
Other interesting points
-
Pretty much everyone agreed on Alan Pelz-Sharpe's point that no one wants to do compliance. Customers just see it as a cost with no upside.
-
The analysts on the panel had no interest in Open Source CMS. The topic did not come up until one of the audience (not me) asked the question. Alan Pelz-Sharpe said that he did a little research into the market and found that some of the Web Content Management products were pretty good and may present a good deal for small to mid size companies. Generally the analysts didn't follow them and avoided recommending them.
-
I am not sure if I heard him right but I think that Joshua Duhl made the point that Microsoft did not offer a content management system. I was surprised because it was such a huge topic last year (remember, Microsoft bought NCompass and started selling it as Microsoft Content Management Server). I talked to a couple friends in the vendor community and learned that MS has basically done nothing with the product since purchasing it and all the buzz is around SharePoint and their document management solutions. Is Microsoft stepping away from their Content Management Server product?
Nov 30, 2004
I just heard Frank Gilbane define Content Management as the "management of unstructured information" at the Gilbane Content Management conference this morning. The definition is growing on me and here is why... I think there needs to be some limits on what we include when we say "content." For example, I would not call transactions in a General Ledger system "content." I think that the most structured we get with content is semi-structured where certain information (such as publish date, author name, and maybe even title) is structured and can be fielded or tagged. However, the body of the content (as in the beautiful prose I am writing here) is most certainly unstructured. Is the degree of structure how I decide what to call content and what to call data?
Thoughts?
Nov 30, 2004
I just just read Jeffrey Veen's essay Making A Better CMS and was struck by his intentionally provocative claim: “Most open source content management software is useless. The only thing worse is every commercial CMS I’ve used.†In the essay, Veen admits that his sample is limited to the out of the box installs that he was able to try on OpenSourceCMS which, although he does not say, is limited to a few LAMP based community CMS products.
I will refrain from my impulse to Fisk this article because I feel like people who make a living by complaining about software do have something to add to the dialog. I do have issue with calling a whole category of software “Useless†because a couple applications do not meet his vision of how software should work. CMS is a hard domain because these systems are used in such different ways by such different users. There are constant trade offs between "ease of use" and complexity, "out of box" and flexibility. I can see why Jeffrey is frustrated by everything he has not written himself. On the bright side, with Open Source, we have the ability to change what we don't like.
Here are some valid points that Jeffrey raises:
-
"Write task-based documentation first." A couple of products do this really well. For example, Plone's How Tos are excellent. However, some products have almost no documentation at all. I think this is a great area where business users can contribute to the OSS movement. We have all these excellent writers using content management systems. Rather than complain, ask questions and write documentation.
-
"Why do you insist on Web sites that have 'columns'?" Layout and positioning with CSS is relatively new. Most websites don't do it. There are some incompatibility issues with older browsers. Still, I think this technology makes leaner and easier to manage pages and newer software has a great opportunity to use it.
-
"Make it easy to install." One of the great things about OSS is that it allows people to freely download and demo it (without even filling out a form and getting hassled by a salesperson). Creating a simple and bullet proof install routine is critical to driving adoption. Some products are better than others. However, my experience with most commercial systems (CMS, and other server based software) is that they don't even try to be easy to install. In fact, most big software vendors have installation engineers that go on-site and install the software for their paying customers (not for free either). They rationalize this with the argument installing is a one-time thing and varies from environment to environment.
-
"Users of a public Web site should never, never, be presented with a way to log into the CMS." Good point, I see that Jeffrey can write his “own templates, and even dip into object-oriented Perl and Python.†He should be able to take care of that.
Nov 29, 2004
I just saw a post on the CM Pros mailing list describing a project to compare CMS using a markup language called CMSML which is like Description of a Project for CMS with special emphasis on the features that CMS should have. For those of you that are enrolled in the CM Professionals Summit, you can see a presentation by Bob Doyle and Erik Hartman including a demo of a site that includes ratings for 93 CMS systems (how about that for a teaser!).
The tool uses a star-based scoring system for the categories Editing, Content Management, Document Management, Information Retrieval, and Records Management. Each of these categories has from 10 to 30 criteria that sum up to an overall score. The categories sound a little like "uses" similar to the approach that I advocated in an earlier post. For example, looking for a Document Management System vs. a Content Management System. I could see this information being used to get to a short list of products that excel in a few key areas.
Nov 26, 2004
I just noticed this article on CMS watch. What I find interesting is that when I used to work at divine (then purveyor of Content Server (CS) as well as a very capable lower market solution called Participant Server (PS)) the strategy was to sell CS as an industrial strenth enterprise class CMS, and PS as a solution to departments or small companies where CS was overkill. Then corporate politics started to distort the strategy. The sales organization came largely from Open Market (makers of CS) and the services organization came from ePrise (makers of PS). Sales all but forgot about PS because they didn't know the product and the high license price of CS produced higher commissions. Services didn't like CS as much because they found it harder to use making it difficult to deliver against the overblown expectations and underestimated services costs set by sales.
I wonder if CS has gotten much easier to use in the last couple of years. I remember seeing a press release announcing First Site which contains a bunch of starter templates and reusable components (kudos to FatWire for getting that done, I know that such a project was on the queue at divine for ages). Also, I wonder if FatWire is experiencing organizational issues similar to divine. FatWire had its own CMS which seems to have taken a second seat to CS.
Nov 24, 2004
Over the past few months, my attention has been on Open Source CMS products. My experience so far is that a lot of the energy in the space is focused on collaborative content management: where the line between readers and contributors is blurred. This makes sense because collaborative content is an "itch" that OSS developers like to scratch. In the OS community, everyone wants to communicate and participate. There is less emphasis on centralized editorial control because the OSS community also has a culture capable of self policing and members genuinely want to add to constructively to the dialog (Don't get me wrong, there are some excellent traditional Open Source CMS products out there). You look at the success of projects like Wikipedia and you see evidence of this trend. Also, much of the online documentation for Open Source products allows viewers to correct and annotate content (see Zope and PHP for examples).
Inside companies, I am seeing a transition from tightly controlled internal informational resources like knowledge bases to Wiki based systems where anyone to can contribute. By the way, I think this is a huge challenge to the Enterprise Content Management (ECM) trend promoted by commercial CMS where all content is centralized under a single system and taxonomy. Here is an InfoWorld article describing how Google uses internal blogs.
But will it stop at the firewall? If you read the Cluetrain Manifesto you might think that this open dialog way may be used for external communications too.
So what are the implications? It sounds like chaos. Thousands of mini sites with dissonant ideas and information. Sounds like the Internet!
No control to enforce company policy, messaging, and branding and protect intellectual property. Companies are going to have to come up with new policies for dealing with their empowered populations.
Decentralization would make all these enterprise taxonomy efforts a lot harder. With more authors publishing in less formal ways, it is more difficult to enforce tagging and other attempts at organization. If so, search is going to get a lot more important (hmmm... Google.... blogs.... Interesting!). Editors, responsible for maintaining the quality of communications, may feel overwhelmed with all the channels they need to monitor. By the way, I wonder if anyone at ThoughtWorks edits Martin Fowler's blog when he talks about ThoughtWorks.
Vignette's purchase of IntraSpect and their Collaboration Services offering shows commercial ECM reacting. Licensing schemes for some products might need to change for their customers to be able to open up these systems to a wider user base. I wonder if any vendors are or will soon start to offer a Wiki.
Nov 24, 2004
A few days ago a colleague of mine referred me to this Slashdot posting about the state of Open Source Content Management Systems. One thing I was reminded of is that the term CMS, means such different things to different people: blog, wiki, message boards, file repository, what I would call a "traditional" web CMS (clear distinction between readers and authors, editorial process, workflow, etc.) and everything in between. "Content" is a very vague term - people to take it to mean whatever is inside the system. So that makes a CMS anything that manages the stuff that it manages. Now it sounds like we are talking about information systems in general. The wikipedia definition does little to narrow the field: "A. content management system (or CMS) is a system used to organize and facilitate collaborative content creation." Still, I like it better than "Knowledge Management" but that is another post ;).
So I started thinking about categorizing the broad spectrum of CMS and was thinking that a good way to break down the space is along three axes related to the kinds of business problems they are trying to solve:
1) The proportion of contributors to readers. Or, similarly, the level of differentiation between contributors and readers
2) The level of editorial control. This could be measured by the number of stakeholders that need to review content before publishing or the amount of refining that needs to be done before the asset is approved.
3) What type of content. On one extreme, there are documents such as MS Word documents, video files, static PDFs. On the other side is textual content that can be rendered into web
pages.
If this model works for you, you might start thinking of CMS implementations that you are familiar with and mentally plot them along these three axes. For example, a company marketing site might have few authors to the number of readers, high degree of editorial control, and mostly textual content.
Where I think this model adds value is that it would allow an organization trying to select a CMS to quickly winnow down that vast universe of products to a subset designed to solve the specific set of problems.
I know that a lot of effort has been made to compare features of CMS (CMS Matrix does a really nice job on a subset of CMS products). But maintaining comprehensive feature lists is very difficult and the binary scoring of features (has it/does not have it) may not be helpful because the way each feature is implemented determines the utility to the user. Stepping back to the business problem might lead to a more direct path to the solution than to focus on individual features.
Once the field is narrowed down the features can be compared in a more qualitative fashion through demonstrations and prototypes.