Category: Content Management

Enterprise Content Management (ECM), Web Content Management (WCM), Document Management (DM). Whatever you call it this category covers market happenings and lessons learned.

Alfresco recognizes ecmarchitect.com’s Alfresco Developer Series

I’m very pleased to announce to ecmarchitect.com readers that Alfresco has chosen me as their Community Contributor of the Month for December, 2007. This is primarily in recognition of the Alfresco Developer Series articles I’ve posted this year which are aimed at bringing new developers up-to-speed on the platform.

I’m extremely flattered that Alfresco chose me to be the first recipient as part of this program. I think it highlights the fact that in the open source community, there are numerous ways you can get involved that can add value, whether that’s by writing code, helping test a new release, contributing a project to the forge, or writing documentation.

I’m also fortunate that Optaros encourages and expects employees to get involved in the open source community–it’s one of the many reasons I joined the company.

Last, thanks to everyone at Alfresco (John Newton, Matt Asay, Paul Holmes-Higgin, Kevin Cochrane, Luis Sala) and Optaros (Marc Osofsky, Dave Gynn, John Eckman, Brian Doyal) for encouraging, reviewing and promoting the articles.

And a special thanks to those of you that have read the articles and left comments or approached me at conferences over the past year. Knowing you are getting value out of this stuff makes it worthwhile.

Okay, cue the music and cut my mike. I’m off to the after party.

Notes from the Gilbane conference on content management and collaboration

The Gilbane Conference on Content Management and Collaboration wrapped up last week in Boston. This was my first Gilbane conference. The most notable thing about the conference is that all of the sessions are made up of panelists participating in a moderated discussion rather than single speaker, death-by-powerpoint sessions. I found the format refreshing initially, but quickly discovered the downside which is that the panels can easily get way off-topic.

Some rough notes from the conference appear below…

Collaboration Case Studies: Pfizer

  • Pfizer implemented MediaWiki, initially to use as a knowledgebase.
  • Known as Pfizerpedia, the site gets 12,000 unique visitors per month.
  • Key adoption factors were: Seeding the wiki with content, promoting early adoption through key champions, taking advantage of pent-up demand, holding the hands of the users as they learned to use the technology, providing guidelines for acceptable use, integrating the wiki with other content stores (team spaces and formal document management), tracking and reporting on usage and impact
  • Pfizer found that because they lack enterprise search, their wiki evolved into a user-maintained index of sorts. I found it odd that an organization that is so knowledge-centric would lack enterprise search.

Collaboration Case Studies: Mitre

  • This was a great example of Enterprise 2.0 in the real world.
  • Components of their solution: Portal (Oracle), Team spaces (Sharepoint), Search and Expertise Location (Google Search Appliance), Social Bookmarking (Scuttle). If they have wikis or blogs I missed what they are specifically using.
  • Their “Phonebook” app was really compelling. Beyond just being a corporate directory with contact and org info, it allowed users to see what communities everyone belonged to, documents they’ve published, projects they are assigned to, things they’ve bookmarked, and whether or not they are online.

Look at http://www.wikipatterns.com for patterns and anti-patterns around wiki implementations.

According to McKinsey, 40% of the work done in western organizations is Tacit which includes decision making, collaboration, and knowledge management. This is where the focus of IT investments should be.

Mash-ups

Kapow showed a demo of their mash-up maker tool. The simple example was that of being in a spreadsheet and needing to retrieve the stock price for a given symbol. Their point was that not all web sites have an API but with their point-and-click tool you can create REST-based services on top of any web page. In their example, they fired up Kapow, opened the quote.com website within the tool and highlighted the stock symbol field to define it as one of the service’s parameters. They then clicked the stock quote button which returned the price. They highlighted the returned price and defined that as the value the service should return. That’s all they had to do to define the service which they then deployed to a locally running server. They then went into Excel and wrote a formula which invoked the service using the stock symbol in the currently-highlighted cell as the service parameter to return the stock price. Obviously, if quote.com changes their markup, service will have to be redefined, but it was easy to see how business people with little or no technical skills could create their own mash-ups, even when the data sources don’t have an existing API.

IBM showed a demo of their mash-up maker called QEDWiki. They showed how they could build mashups through a web browser. Their tool didn’t provide the service builder–the value of the tool seemed to be bringing together data from existing REST-exposed sources into a single page and being able to do that configuration in the browser. They mentioned a mash-up tool being available at Alphaworks but it wasn’t clear whether or not that was the same package being demo’d.

Opening Keynote

Have you noticed how chummy Adobe and Alfresco are these days? John Newton, Alfresco CTO, and David Mendels, SVP from Adobe, were both on the opening keynote panel. The two were definitely in sync on where they thought content management was going. John said he thinks social computing will drive ECM from being used by 10% of the people in an organization today to being used by 80% or 90% in the near future. He mentioned the Facebook integration that’s been getting so much press lately. David said that content must be service-enabled so that it can be assembled in new ways which plays right into Alfresco’s recent addition of the REST framework.

Mendels also let it slip that Adobe has two hosted content management solutions, both of which run on Alfresco. One is Buzzword, which Adobe recently acquired. The other wasn’t named.

Alfresco says it’s all about connections. Adobe says it’s all about interaction. Seems pretty in-step to me.

WCM Keynote

This was a disappointing mix of closed-source WCM vendors. None of the vendors differentiated themselves at all or offered up anything new or interesting with regard to where WCM is headed.

WCM Analyst Panel

As a general rule, you shouldn’t miss an opportunity to hear Tony Byrne speak. His honesty and straightforwardness is always refreshing at these events. He gave the audience a piece of advice regarding evaluating CMS vendors which was to insist on a bakeoff. He said, “You wouldn’t buy a ferrari by watching the sales guy drive the car around the lot, you’d insist on getting behind the wheel. Why should it be different with a CMS?” I’d add a bit to that. When you do the test drive, you should take your mechanic.

I see many customers making CMS decisions before thinking about who’s going to do the implementation and the customization. Or they wait too long to get a professional services firm involved in the process. Obviously, I’m biased–my ECM practice at Optaros is in the business of helping clients with CMS evaluations and customizations–but the point is to seek advice from subject matter experts. Even if you do a bakeoff, there’s still a lot to learn from the people that have been there that you might not uncover during the bakeoff.

The Future of Collaboration/Enterprise 2.0

I was extremely frustrated with this session. I attended thinking the panel would stick to the topic–Enterprise 2.0. Unfortunately, the discussion was around everything but that. The moderator and the panel seemed to confuse “Web 2.0” with “Enterprise 2.0”. Rather than talk about how Web 2.0 technologies can be applied within an organization to boost collaboration, leverage the power of the social network across the org, and reap the benefits of a less-structured, self-forming, self-regulated approach to Knowledge Management (this is McAfee’s and generally everybody else’s definition of Enterprise 2.0), the entire session was devoted to old ideas around customer engagement, customer-driven product development, and online communities. It was a very extranet/internet-centric discussion which entirely misses the point.

I wasn’t the only one who was frustrated–after asking the panel a question which essentially boiled down to “Is it you or me? Which one of us is confused?” several people approached me to share their disappointment.

Andrew McAfee

This was a panel composed of Frank Gilbane and Andrew McAfee. McAfee has done a lot of research around Enterprise 2.0 at Harvard and is always an entertaining speaker. Unfortunately, the format and the length of the slot didn’t really give him much room to stretch his legs. I did get a chance to ask him if he had done any research into the size of an organization that’s required to get the full network effect inherent in Enterprise 2.0 solutions. He said no one really knows yet what the minimum size is but anecdotal evidence suggests it’s “surprisingly small”. If you are looking for examples of real-world Enterprise 2.0 implementations, you should check out the site he started for capturing Enterprise 2.0 case studies.

Making RFP’s more effective

In the WCM Analyst Panel at Gilbane last week there was a bit of discussion about how to write RFP’s. The Gilbane analyst gave advice like, “Use open-ended questions rather than yes/no questions” and then went on to complain about vendor (un-)responsiveness to his 150-question RFP’s.

I dislike RFP’s. Even if you eliminate from consideration RFP’s that are obviously rigged toward one vendor or RFP’s sent to a vendor for the sole purpose of the vendor serving as column fodder, the vast majority of RFP’s are too often misused. RFP’s may be an acceptable way to buy commodities like office supplies and construction materials, but they are a terrible way to by complex, often heavily-customized software. It’s highly unlikely that you are going to be able to come up with a set of questions that either (a) distinguishes one vendor from another in the areas that matter or (b) gets to the heart of whether or not that solution will ultimately be a good fit for your very specific requirements.

I realize that in some companies, the purchasing department can be pretty hardcore and that the RFP process may be as certain as death and taxes. If it’s unavoidable, at least try to use RFP’s in a way that makes the most of your time and the vendor’s. One thing the analyst said that I totally agree with is that you should think of your CMS implementation as you would a custom application implementation. If you accept that a significant amount of customization will be required, why would you then go on to ask detailed questions about functionality that in all probability won’t exist until it is developed as part of your project?

The point of the RFP, then, should be to figure out if the CMS fits your environment and your world view. RFP’s should be focused on weeding out solutions that won’t work for you based on things like architecture (platform, language, and other “enterprise footprint” dependencies), licensing model and cost, company viability/stability, support options, and ease of customization. You ought to be able to do that in 10 questions or less. And, really, you ought to be able to answer those questions on your own by looking at the vendor’s web site. Once you’ve used those to create a short list, then it’s time to have real conversations between yourself, the vendor, and your integrator and start working out bake-off logistics.

Anything more complicated than this is a colossal waste of everyone’s time. I recently saw a company issue an RFP to about 15 different CMS vendors running the gamut from the usual “leading” closed-source vendors to a couple of open source players to mid-market players to services firms pitching custom solutions. Such a diverse field is a huge red flag and an indicator that either the client really doesn’t know what they are looking for or they don’t understand the market. Analysts, services firms, and online communities can help in either case, but only if the client (and the client’s purchasing department) lets them.

The other painful thing about this particular RFP was that it was well over 200 questions with the majority of those being around unimaginable minutiae. I’ll bet the first 10 to 20 questions could have been used to eliminate 2/3 of the field from consideration.

Rather than reacting to lackluster RFP responses by adding more questions, diving into microscopic detail, or tweaking the format, consider whether a shorter RFP focused on narrowing the field based only on the most critical fundamental requirements followed by a bake-off with two or three vendors wouldn’t be more effective. My hunch is that in most cases, it will lead to a higher quality decision in a shorter amount of time.

Get your Alfresco ‘flow on

With the busy holiday season approaching, it’s hard to find time to streamline your content-centric business processes. Here’s a tip: Buy everyone gift cards. Use the time saved to learn how to implement advanced workflows using Alfresco‘s embedded JBoss jBPM engine. This article and the accompanying source code should be everything you need to get started and then some.

The article starts by describing jBPM concepts and the high-level steps for implementing advanced workflows in Alfresco and then dives into the details by walking you through an example.

The example extends the “SomeCo Whitepapers” story from earlier articles in the Alfresco Developer Series by implementing a business process to enable SomeCo’s engineering team, marketing team, and third-party partners to review whitepapers before being published to the SomeCo web site.

The integration of third-party partners is handled through email–recipients simply click a link to approve or reject the workflow task.

As in prior articles, the source code bundle is cumulative–it contains all of the “SomeCo” code we’ve worked on thus far.

About the “Alfresco Developer” series of articles

The Alfresco Developer Series of articles is a collection of technical tutorials aimed at getting you up-to-speed quickly on the key aspects of Alfresco. The series covers extensions and customizations performed during a typical Alfresco implementation by walking through a realistic example that is expanded upon in each successive article. The content is based on real-world Alfresco projects executed by the Optaros ECM practice for clients around the globe.

Past articles include:

Flex and Java presentation framework discussions

Given Alfresco’s “serious commitment to Flex”, James Ward’s response to Matt Raible’s Java presentation frameworks post is interesting. We’ve been having similar discussions internally at Optaros around this as well.

It’ll be interesting to see how well Alfresco addresses some of the challenges in Flex both authors mention, such as automated testing, printing, and accessibility, all of which will be critical if Alfresco really is going to re-build their web client on the Flex SDK.

Who’s buying what Forrester’s measuring?

It’s hard for me to lend much credence to Forrester’s Vendor Scorecard for Alfresco, recently released as part of its Forrester Wave report on ECM Suites. The reason is that I don’t believe too many people are looking to buy an “ECM Suite” at least as Forrester defines it.

Forrester doesn’t seem to care that most of us have moved past the legacy definition of ECM as an all-encompassing, single-vendor software suite. But as the “Suite” term implies, Forrester’s report is slanted toward gigantic vendors who offer every piece of technology even remotely related to ECM.

This is in direct contrast to what Optaros clients ask us for. They want agility, speed of implementation, product innovation, and freedom of choice. Many are implementing Web 2.0 and Enterprise 2.0 solutions which work better with the concept of “ECM as a service” rather than “ECM as giant, complex software offered by a single vendor who has cobbled together their offering from acquired and poorly integrated products”.

Given that, it is odd that Alfresco would be measured with the same scorecard used to evaluate an “ECM suite”, but that’s what they did. They judged Alfresco unfairly in a couple of places and generously in others. Really, though, the report made me realize Forrester needs to come up with a better way to segment and evaluate the market because I’m not sure people are in the market for the product that would score well against their scorecard.

Let’s take a look at the report and you’ll see what I mean. Specifically, I’m looking at the Vendor Scorecard spreadsheet for Alfresco. I’ll start by calling out what they got right.

Alfresco scored a perfect 5 for “Platform readiness”. Unfortunately for Alfresco, it’s only worth 2.5% of the total score. This set of criteria favors suites that accomplish more with fewer disparate sub-systems or products so it makes sense that Alfresco would do well here.

Alfresco took a ding on E-forms and Email archiving scoring 1.0’s for both criteria, which together count toward 2.5% of the total points possible. Alfresco can bump up their score by extending current XForms functionality to the DM side of the house which is long overdue. Email archiving should get a better rating when the new “email addressable objects” functionality is released.

This starts a theme that repeats itself throughout the report: In keeping with the “Suite” approach, Forrester expects an ECM vendor to offer better E-forms and Email archiving, for example, than pure-play vendors. This automatically favors the largest of the ECM vendors because they simply acquire and integrate pure-play offerings. So don’t expect a “5” in these areas for Alfresco any time soon using the “Suite” standard.

“Core capability architecture” scored well, which is a measure of Alfresco’s ability to work in an “enterprise” IT environment, but, strangely, this score is only worth 5% of the total. That’s not a typo. The ability for the product to leverage a client’s application and security infrastructure as well as its extensibility is not as important, in Forrester’s view, as the ability of the ECM vendor to offer Business Intelligence software (see “Differentiation Strategy”).

Forrester’s expectations for the breadth of an ECM Suite become clear when you look at what a vendor would have to do to get a perfect score. And this is where I start to question the value of their definition of the ECM Suite, let a lone the scorecard itself. Let’s look at a few examples…

Under “Core ECM capability breadth”, the “Content Integration” category is interesting. On the face of it, I would think products that support open standards and have many API’s available for integrating the repository with other systems would do well here. On the contrary, Forrester gave Alfresco a 1.0. Here’s what a vendor must offer to get a 5:

“Comprehensive, independent content integration offering supporting central, federated bi-directional access and management of content in heterogeneous, disparate repositories and content information sources.”

It’s like the “E” in ECM stands for “Excessive”. Seriously, I’m sure certain clients will have a legitimate need for this functionality but there are more common, more practical integration-related concerns vendors should be striving for.

Let’s turn our attention to “Extended capabilities” which is worth 10% of the total points possible. Alfresco got an overall score of 1.05. Alfresco’s lack of an Enterprise Search offering (Score of 0 with a weight of 15% for the category) and, in my opinion, a mis-rated BPM capability (Score of 1.0 with a weight of 20% for the category) were the primary culprits here.

The inclusion of Enterprise Seach as part of an ECM Suite evaluation floors me. Does anyone really want to buy an enterprise search engine from their ECM vendor? I suppose if you are IBM you are happy it is included. It is about this point in the report I start to form a mental image of Forrester’s Dream ECM Suite as the giant, lumbering Imperial Star Destroyer shown in the opening sequence of Star Wars (or, better yet, the parody of the same in Spaceballs in which the enormous ship has a bumper sticker that reads, “We Brake for Nobody”).

The BPM score is simply a bad rating. Either the criteria for BPM is so overblown as to not be relevant to the majority of clients who need workflow within an ECM suite or the analysts simply did not do enough research into the capabilities of the embedded JBoss jBPM engine. Forrester doesn’t offer any details as to why it considers the functionality to be “basic” so we’re forced to guess.

Collaboration is another example where Forrester’s threshold is set way too high. Alfresco scored a 2. To get a “3” in this criteria–an average score–you have to offer real-time web conferencing. Huh?

When you get to “Strategy” and “Market Presence” you realize who Forrester values: Huge vendors that offer everything to everybody and have more than $10 billion in annual revenue (that’s what it takes to earn a 5). And to be successful, the vendor must not focus exclusively on ECM. Take a look at “Differentiation strategy”. This one’s important to Forrester. At 12.5% of the total score it represents the single most heavily weighted item in the scorecard. To score the max, here’s what a vendor has to do:

“Vendor differentiates by offering ECM as part of a broader information management portfolio (data, master data management, BI, content, BPM, portal, and collaboration technologies) with a promise of lower TCO.”

Hmm. Who could offer broader information management, Business Intelligence apps, Portal, AND collaboration technologies? Wait, don’t answer yet. The vendor musn’t stop there. It’s not quite broad enough. To score max points on “Vendor’s business”, another 5% of the total, the vendor must provide “…IT application infrastructure solutions including application infrastructure, developer tools, middleware, and information management.” Basically, if your name is IBM, Oracle, or Microsoft, give yourself a 5.

So Alfresco took a couple of knocks here and there it probably shouldn’t have. The bigger issue is that Forrester’s vision of ECM is out of step with what clients want and their scorecard reflects it. No one’s in the market for Imperial Star Destroyers anymore.

Notes from the Alfresco Community Conference

I’m back from the Alfresco Community Conference which was held Wednesday in New York City. It was a day well spent with excellent presentations given by John Powell, John Newton, Kevin Cochrane, Paul Holmes-Higgin, and Helen Dann. Christian Science Monitor’s Russ Danner gave an update on regional Alfresco user groups. And three Alfresco clients–Kaplan, Harvard Business School Publishing, and Film Solutions–all gave brief updates on solutions they’ve built with Alfresco. It sounds like Kaplan has a significant Alfresco WCM-based site going live this week, BTW.

The conference also gave me the chance to meet several of you in person. Thanks for the kind words and your continued interest in the blog. I made some progress on the Advanced Workflow article on the plane home so stay tuned for that.

Here are some (rough) notes from the conference…

Alfresco has released a new independent benchmark

  • 100 million objects
  • Loaded the objects at 140 objects/second (using an RMI interface on a different node than the repository)
  • Unlike some competitors–ahem, Microsoft–there are no artificial limits in terms of the number of objects in a collection or the number of collections.
  • Linear scalability–there was no degradation as volume increased.
  • Sub-second response times.
  • Next stop is 1 billion objects. An impressive side note is to compare how long it took Documentum to get to a billion objects (~10 years?) with Alfresco’s efforts (~3 years?).

Alfresco’s architectural principles (para-phrased):

  • Content is a service
  • Everything must be modular
  • Lightweight scripting is important
  • Enterprise scale
  • Must fit into the user’s environment (multiple platforms, multiple browsers, multiple databases, etc.)
  • The web is the interface
  • Continuous innovation

Problems with Sharepoint (I’m considering doing a Sharepoint vs. Alfresco smackdown article so we can get a real independent comparison going, but for now, here’s Alfresco’s opinion):

  • Slow, doesn’t scale. (Limited to 5 million docs per library and about 10 libraries total).
  • Hard to program and extend.
  • Poor WCM.
  • 100% Microsoft stack from browser to back-end and everything in-between.
  • John Newton said, “IBM should be competing directly with Sharepoint but they aren’t providing any guidance.”

A comment John made that is begging to be drilled into is that he considers Alfresco’s Office integration to be better than Microsoft’s. This is a bold claim. Most people would cite “deep MS Office integration” as one of the things that would immediately take you down the Sharepoint road. So I guess I’ve got a to-do to figure out how much truth there is to that claim.

Exciting things just released or coming soon:

  • Major commitment to Flex
  • Flex SDK
  • New Flex-based client for Knowledge Workers called “Networks”
  • Joomla, WordPress, and MediaWiki integration
  • Facebook and iGoogle integration. Essentially run facebook apps/gadgets that expose the Alfresco repository. It sounds like Alfresco will also be an OpenSocial container when OpenSocial is ready.
  • JSR-283 which provides a SQL-like query capability to the repository (Recovering Documentum developers, think “DQL”).
  • Support for ATOM Publishing

What’s going on with various releases (None of these dates are necessarily set in stone. Check the Alfresco wiki for the official roadmap):

  • 1.4 is nearing end of life. Everyone should be upgrading to 2.1.1.
  • 2.1.1 is now available.
  • 2.2 is on its way and should be out before the end of the year.
  • 2.9 Community (Q4 of 2007)
  • 3.0
  • Community in February of 2008, Enterprise in Q2
  • WYSIWYG for mash-up compositions
  • Flex-based RIA client for ECM

More info on 2.2:

  • This release is the foundation for the new collaboration features that will start to find their way into the product.
  • Email addressable spaces. Allows every object in the repository to have an email address. For example, mail a piece of content to a folder in the repository. Respond to an email about a piece of content and start a discussion thread that gets persisted to the repository.
  • Branching. Allows web projects to start as copies of an existing web project.
  • Web Content ACL’s. Exposes object level permissions in WCM to the UI. (Previously, only the DM store could be used to assign permissions at the object level. The WCM store was limited to the site level).
  • In-context editing. Allows someone to click an icon while previewing a web site that immediately launches the web form used to create that content.
  • Find content created with a specific web form. Adds a link to the sandbox page next to the “Create Content” link for each web form that takes you to the content created with that web form. MUCH NEEDED.

    More info on 2.9 community:

    • Like a preview of some significant features that are coming in 3.0
    • Dynamic data dictionary. Allows you to implement new models or override existing models without a restart. YES!
    • Multi-tenancy. Good for internal or third-party hosting situations in which you want to essentially partition the repository so that two or more “tenants” would never know the others existed. Segregates workflows, repositories, models, etc.
    • iGoogle/Facebook integration. Exposes things like search, workflow tasks, and “document library”.
    • Collaboration AMP. Includes ability to post content to WordPress, use Alfresco for the back-end to MediaWiki, IM presence integration.

      All in all a good opportunity to hear from and interact with Alfresco leadership and others in the community. I’m looking forward to the next one.