Tuesday, February 28, 2006
It's confession time - I got it wrong. Big time. In the late 90's Chris Harris-Jones and I wrote the book Web Content Management - Strategies, Technologies & Markets together. A central assertion of it was that WCM should be a part of ECM (or IDM as it was often called then). That assertion was largelly due to me arguing till I was blue in the face with Chris, who was far from convinced!
Yet if we look around us today - WCM and ECM are not managed together. Though some ECM vendors have built perfectly good WCM tools (Documentum being a good example), they still tend to sell them quite separately, and they are almost always deployed standalone from each other.
At the end of the day Web Content must be subject to the same retention/compliancy/record management rules and procedures as any other content. Currently it is seldom managed this way, if ever. In additon Web Content should not duplicate content - if content sits in repositories within the enterprise, these should be accessed as seamlessly as possible. But there, for all practical purposes the connection ends.
Document management, is for the most part either a tight process driven activity (such as claims processing) and/or a collaborative activity. WCM is primarily however a publishing and/or trading activity. The skill sets and methods for managing both differ widely.
What is clearly happening (to state it for the 100th time) is that DM and Workflow functionality is being absorbed into the infrastructure. But WCM activity is not. Rather it can be (most portals have some WCM capability), but in general is not. It is almost as if WCM is seeing something of a rebirth after its asscendency in the dot.com - and so it should. Problem is so many of the WCM solutions out their are unwieldy, difficult to use and incredibly proprietory in nature....in addition we have the added confusion that WCM is such a catch all term. Though there are lots of WCM tools available there are so many ways to manage a website that it's almost too broad. Compare an Interwoven with a Plone, a Fatwire, an IBM Websphere Portal or a SharePoint....all can 'manage' websites.....
Clearly a post that needs more thought, but a new thought train for sure :-)
Posted by alan pelz-sharpe at Tuesday, February 28, 2006
Friday, February 24, 2006
My colleague Apoorv has pointed out in his blog that Alfresco only supports level 1 JSR170 compliance. In my gushing endorsement below I had not noted this and think it well worth bringing to note.
JSR 170 is turning out to be one of the industries best kept secrets. I had a meeting with a client today, one who is very well informed and up to speed on all things CM, but had not realized the significance of this standard. And I find this is not uncommon - would love to hear from others on this.
JSR 170 is not a silver bullet, but it is probably the first meaningful example of a cross integration standard for CM. Previous attempts such as DMA or even the various workflow protocols from the WFMC fell well short of the mark. But the fact that there have been well tried attempts in the past opens up the question why, and if the JSR spec has any more chance of success.
Firstly a bit of history - the DMA (Document Management Alliance) now under the control of AIIM, tried to build specs for a model API for cross repository integrations. The WFMC tried to do something similar for workflow applications. These were worthy attempts but failed due largely (or entirely) to lack of enthusiasm from DM vendors. Why the heck would they want anyone else to access their repositories? In short a good idea for users, a rotten one for vendors.
But then came the portal technology and web applications that often if not always needed to interact with some kind of content repository. JSR 170 came from this need for a standard Java based API to access disparate repositories.
Most ECM and WCM vendors still think its a pretty rotten idea for others to access their repositories - why they cry isn't all of an enterprises content in a single instance of my (insert as neccessary - Hummingbird/OpenText/Interwoven/Documentum etc) repository?
But it just doesn't add up any more particularly in a web services and SOA world - so reluctantly they are coming round to it. More because it can be a disqualifier to RFP's than anything else I suspect. And on that note I would like to recommend here that EVERY RFP ask vendors if they are JSR 170 compliant and to what degree (not just a check box).
Database and Portal vendors are of course leading the way in JSR 170 compliance - but the ECM and WCM vendors are coming along at a steady pace. But all are not.
In my opinion and I am sure Apoorv would agree with me (and James Governor too I suspect) any software that has a content repository or interfaces with one should be JSR 170 compliant, if its not it shouldn't be on your list. By the way 'membership' of the JSR Expert Committee does not automatically mean the companies products are compliant :-)
What a week! Sick in bed with Gastric Flu and also desparately trying to finish a project, and sort out a ton of other things. Hence no postings.
Couple of things I have wanted to find time to write about for a while but haven't quite gotten round to are:
- Which analysts cover this space well?
- Which books are worth reading/buying?
Reason being I get asked both these questions often! I'll deal with the analyst one first, as that's easiest but also a potential minefield as I know so many of them in the ECM/RM space.
The way I see Industry Analysts providing value is in two ways - direct interaction with the analyst or reading the reports that their firms put out (asking them to get involved in consulting is usually a big mistake - take if from me). For most mortals it is extremely difficult to gain access to an analyst for any period of time unless you are a major client of the analyst firm. But if you are one of the fortunate - then all the leading firms have some good analysts that are worth talking to:
- Toby Bell & Karen Shegda at Gartner
- Sue Feldman (particularly on search and taxonomy) at IDC
- Connie Moore & Bob Markham at Forrester
- Chris Harris-Jones, Sarah Kitmer and Angela Ashenden at Ovum
- Mike Gotta (particularly for collaboration) at Burton Research
- Jim Murphy at AMR
For written output though its a different list - as for whatever reason the level of detail, accuracy and value of the written output of the various firms differs widely. Leading the bunch (quite frankly by a fairly good margin) is:
- CMS Watch led by Tony Byrne (yes I am a guest contributor there!)
- Forrester - overall excellent coverage
- Ovum - particularly for the major ECM vendor evaluations and ECM market forecasts
- Burton - for collaboration and portals
Tuesday, February 14, 2006
I held off commenting on Alfresco (Italian for outdoors...?) till the buzz died down, and I had a chance to look a little closer at what they are offering. I will state here that I have not tested it - and am only looking at this from a high level. What I am impressed with is the range of platforms, databases and languages (french, german) etc that the product supports.
For an Open source and early release product it bodes well (and well in advance) of many established non Open Source alternatives. And a couple of things that really made me sit up and take notice were the support for JSR 170, and the fact that advanced functionality such as automatic tagging is included as standard. Just reading the specs makes one realise that this appears to be a pretty serious product, and the service fees are not astronomical. So for mid sized businesses it may well be a good option.
Yet the success of Alfresco and other free downloads (open source or not) such as Nuxeo, SharePoint or 80:20 depends on the amount of services these deployments generate. And ECM and RM are highly service intensive, so low cost or free software appears to be a perfectly rational way to go - my concern as always with Open Source is future upgrades and modifications. There are so very many 'dead or dormant' Open Source CM projects that its a legitimate concern. But then I for one don't be believe in the concept of a free lunch - I do though believe in a bargain, and from what little I can tell at this time, it's a bargain.
I am in truth always loathe to recomend Open Source alternatives, partly as I once got flamed for having the temerity to suggest in the press that Open Source may actually not be as free as it looks (so I clearly hold grudges). But also that I suspect those who are genuinely altruistic spend there times at Soup Kitchens and raising money for the PTA, rather than coding Open Source for the general good. Even if altruism was the origin of Open Source activity - products like Alfresco are there to make good living for the firms owners. There is nothing wrong with that, and the move to low/no cost ECM software is someting I applaud. What I do have a bit of problem with is the positioning of Alfresco the first open source ECM tool? Actually I have two problems, both are arguably trivial but anyone who reads reads blogs on RM and ECM will at least grasp what I am bothered about.
Firstly, to me the word enterprise means just that - but Alfresco appears to be focused on SMB's - Small to Medium Size Businesses, not Enterprises. So how come its an ECM tool?
Secondly, if we accept Alfresco's loose definition of ECM, then they are not the first Open source ECM player. Nuxeo claim that status, and are probably mightily miffed as a French firm that an UK one is claiming the title. You can't blame Alfresco for grabbing what publicity they can, and though I have met a couple of the founders in the past and respect them. I am a pedant and get thoroughly annoyed over pedantic issues like the E in ECM. INterestingly If one looks at Documentum, they long heralded themselves as the largest and leading ECM player, but by market share or revenue never were (historically has been Filenet or IBM), though they are probably actually approaching leader status now after a very solid last two years. I guess the founders have taken their marketing experience at Documentum and applied it at Alfresco, and good luck to them.
From what limited data I have, I like Alfresco, and having met on a couple of occasions some of the founders, I trust them, I wish them luck. I just don't buy into the politics or religion of Open Source. No flaming please.
Posted by alan pelz-sharpe at Tuesday, February 14, 2006
Monday, February 13, 2006
Thursday was an odd but eventful day last week - I was the guest speak at the last of my round table events. This one was in Beverly Hills, at the famous Spago - it was well attended, well run, the food and wine was terrific, and Wolfgang Puck himself put in a personal appearence to say hello to everyone! The same day I heard that another keynote is joining us in Denmark later this year - Nick Carr of HBR fame...I don't remember Content Management ever having this kind of profile - long may it continue :-)
The Beverley Hills discussion focused mainly around privacy and security issues, and the need or desire, to deploy some ECM functionality to the masses. As a result of this discussion I revised my now lengthy draft article for CMS Watch and have resubmitted it. Not sure when it will be published or what it will look like when it is, but hopefully it will sumarize more thoroughly the thoughts and discussions on this blog around the evolution of ECM.
Posted by alan pelz-sharpe at Monday, February 13, 2006
Wednesday, February 08, 2006
A topic I bleat about constantly is the serve lack of people who understand the art and science of managing information. There are lots of people who can manage the technical depolyment of ECM and ERM solutions - but few who really understand the business and human dynamics that underpin the work these technologies are designed to support.
Hence I noticed this today, that according to the Bureau of Labor Statistics' Monthly Labor Review, five career paths in particular will face worker shortages over the coming years:
Featured prominantly is that of Librarian
Reasons for shortages are a combination of increased demand, fewer young people entering the professions, unattractive wages and working conditions.....
'Librarian', is all too close a skill set to 'Record Manager', and I suspect the growing crisis in the lack of skilled information/document/records managers is only going to grow in the short term. It just isn't an attractive, well respected or well paid job at present - but it is from the minds of those who know how to catalogue and track vast swathes of information that the next generation of ECM software should come.
Smart vendors should begin to employ librarians and records managers to guide, validate and structure future approaches to ECM. Smart enterprises should be searching for these talents and rewarding them well to run their information management strategies.
Posted by alan pelz-sharpe at Wednesday, February 08, 2006
Friday, February 03, 2006
In the past week I have been meeting with a whole host of CIO's and IT Directors across the states. All from either large Enteprises or large Government Departments and currently focused on ECM initiatives.
It may not be terribly scientific but its a pretty good cross section of what is going on here in the US at present - and key themes ran throughout the week. Note that what I have listed here are the topics raised by them not me, and quite frankly had you asked me to guess at the start of the week what these themes would be, I would have been wrong.
Here they are (in no particular order)
- Dealing with multi vendor multi repository environments
- federated cross repository search
- Concern over the complexity (but importance) of internationalization
- Sharepoint Portal server instances running out of control
There were other topics discussed, but throughout the week these all came up again and again. But the single biggest topic?
How to avoid ECM failure in future, based on a chronic history of ECM/DM/CM failure in the past.
It is clear that ECM is being implemented in many organizations simply due to the 'stick' of compliancy rather than the 'carrots' of added benefits. But the scars from past attempts based on the carrot approach has left many with scars and serious reservations.
When I touched on why past projects had fallen so far short of expectations - the following three themes were repeated consistently:
- Insufficient time spent on process and business consulting in advance of implementations
- Insufficient time spent on building, understanding and maintaining taxonomies
- Buying overly complex and over engineered ECM systems
- Duped or oversold by ECM vendors on technologies relenvance and capabilities
The need for ECM appears to be paramount due to compliancy pressures, and the seemingly literal pressure of the increased and ever increasing volumes of content hitting us. But in truth I suspect the need is not so much for ECM, and more for tools, processes and procedures that help us organize and navigate the content volumes, securely and simply. Traditional ECM tools may not always be the right choice.
Also note that beyond retention - and the need to be 'compliant' records management hardly came up in any conversation, and from this sampling of CIO level USA its not really on the agenda.
Well a different night (this time Atlanta) and a different audience, and equally glamorous (venue Joel in Deerpark). Quote of the night for me came when somebody explained the need to move to electronic retrieval of records, due to the speed vs manually pulling records from filing cabinets. I bit my tongue and was polite (unusual for me), gosh if only it were that simple! For those of us steeped in RM, its this kind of thinking that makes our blood boil!
Of course logically the line of thoughts is right. It is much quicker to retrieve something electronically than it is manually, seconds rather than minutes or hours. But this line of thinking needs to be considered in context (RM joke there). For typically in any business process analysis it is the exceptions that we look for and study the most. The regular activity 80-98% of stuff can be automated, but its the 2% that can floor us.
In a well run office with paper filing cabinets things are never lost. They are retrieved with machine precision every time. For they are filed with extreme precision every time. Things occasionally get misplaced (one or two file folders back etc) but short of somebody doing something outragous such as deliberately pulling files out and destroying them, they are always found. I prided myself on my projects in Oil & Gas that I always could find the documents when needed (occasionaly took me a while - but always found them).
However electronic data is a whole different issue - firstly electronic data is seldom filed with extreme precision. Often it is not only filed incorrectly but also physically in the wrong area of a repository (electronic filing cabinet), sometimes in the wrong repository entirely. When something is lost electronically it can be difficult if not near impossible to find. I remember a terrible row on a particular oil platform project when an Electrical lead engineer accused me of losing a document he was supposed to have worked on. After hours of looking for it and unable to find it, I figured out where it was - on HIS desk where it had been put two weeks earlier. When on later projects we more fully utilised EDM systems, we on occasion moved things into virtual black holes and never found them again.
Hopefully the reader can see where I am going with this, and as such am just using it to blow off some steam and will not labour the point further. Other than to say that information management technology is awesome, but without proper procedures, skill sets and structures it is of very limited value. The exceptions in electronic RM are often far more profound, difficult to manage and potentially damaging than those related to manual records.
Thursday, February 02, 2006
Last night I was the guest speaker at a CIO Roundtable event in New York. It was an interesting and quite glamorous occasion (venue Le Bernadin), but with some very solid and down to earth discussions taking place.
At the end of the evening I did wonder if the days of big number licence deals for ECM are numbered. Big Dollar deals though are clearly in vogue, and will continue to be so for some time....
To put it another way, the E in ECM seems to be taking on two different (though compatible) meanings:-
- Enterprise definition A - High level across the breadth of the Enterprise
- Enterprise definition B - Involving everyone in the enteprise
In the current market we are seeing the likes of IBM, Filenet and Documentum negotiating some of the biggest deals of their history.With the aim of taking their intense, heavy duty document and content management capabilities across the whole enterprise.
Yet a slightly closer look at these deals reveals something a little different. In most cases the big deals are in fact consolidation exercise, with the bonus of an upgrade in technology. Most of the big deals are focused on knitting together various reposititories (mainly second tier vendors such as Hummingbird, OpenText, Vignette etc) into a standard architecture or centralized repository. This is ECM that uses definition of Enterprise number 1.
The second thing we are starting to see emerge is organic growth in content management capabilities, outside the control of IT, driven by local users own needs to bring some affordable organization to their content. This is most clearly illustrated by the viral growth in Sharepoint deployments. But also in the success of the likes of Xerox Docushare and 80:20. In one instance I know of, there are upward of 1500 Sharepoint instances within a single enterprise!
In the industry the current buzzword is co-existance, and its a buzzword that is probably right on the money. Because from what I can see now, the likelihood is that we will start to see more hybrid deals being done with one vendor selected for the high end needs of ECM, and a standard (Microsoft or Oracle for example) being imposed for the rest of the needs of the organisation (the bulk, but not the mission critical).
That seems like a good model to adopt, but before we really get there I think we are going to see some difficult times ahead. For there is no doubt we are at in inflection point. Enterprises are now opening their wallets and prioritizing ECM projects. These are ambitious projects and many vendors are going to have their technology tested to breaking point for the first time. I suspect some will regret their boastful claims and end up licking their wounds. Whatever happens the landscape will change. Basically we are coming to the realization that not all content is equal, but it all at least needs to be organized and accessible.
What I think we will start to see is the emergence of vendors focusing on basic but widescale content organization vs lumping themselves with heavy duty ECM activities....
Posted by alan pelz-sharpe at Thursday, February 02, 2006