One of the barriers to successfully rolling out ECM systems is the steep learning curve for users. Before ECM turns up they are happily saving their documents onto a mapped drive or to their own machine. Then ECM arrives and they are having software installed on their desktop or more likely given a URL for a complicated web client that they need training on and makes saving their documents much more complicated than it was. The interface also does 50 other things they have no interest in.

When saving a document they now have to fill in metadata, but they soon work out how to get round that and the minimum number of characters they need to add to get it to save. They keep local copies of the document whilst they are working on it and upload their latest version when they have finished. If they check it out for days other people working on it moan they have it checked out so they stop checking documents out and things get out of sync. All these problems just for creating and editing documents, surely it should not be so difficult?

This revolution on how documents are saved can look like a process designed to make peoples’ jobs harder, not the initial impact you want to make on your users. But how can you get round this? The answer may be to start the functionality small and grow ECM in an evolutionary way.

A great way of getting everyone saving documents into your ECM system is Alfresco CIFS. This allows Alfresco to act as a shared drive, for a demo of CIFS in action there is a presentation on the Alfresco website www.alfresco.com/products/presentations/1.4/AlfrescoCIFS_viewlet_swf.html. CIFS allows you to copy all of your documents from the existing shared drive into Alfresco, then expose Alfresco as a shared drive so that for users there is no change in the way they work with their documents. They can save files direct from all the applications they have on the machine just by using “Save As” and “Save”. This means no software installation is required on the client machines. If the mapped drive is created through a login script you just change the login script to point to the Alfresco server.

So if the users are just doing the same things as before what have we achieved? Well now we can have auditing turned on, so for compliance reasons we can now see everyone who has edited a document and everyone who has looked at a document. We can have versioning happening in the background so when someone accidently deletes a section they want to get back there is no need to go looking through the backups to see if we have that version. We also now have full text search available this can be exposed as an Open Search through a browser.

One of the criticisms of this approach is that it is seen as a backdoor, that users can get their documents into the system without having to fill in any metadata. However we can extract metadata from the file itself, for Office type documents and PDFs we can extract the properties like Subject and Author and add to the document metadata. The question can also be asked do we want to have metadata for every document that gets added or just identifiable sets of documents that will be of interest to the whole organisation. If so we can set up workflows to pick up documents that we want classified with metadata and email the user asking for the required metadata. This workflow could also include an approval process for corporate documents.

But now there seems to be an even better solution for this. Integrated Semantics have built an integration with the Open Calais project which was demonstrated during an Alfresco Tech Talk http://wiki.alfresco.com/wiki/Live, look for episode 18. This allows automated extraction of semantic tags and categorisation of those tags. We threw a few documents at it here and were pretty impressed. It is well worth having a look at.

So how does the project evolve? Well, all organisations have different levels of users. There will be some users who work with Alfresco through Windows Explorer and that is all they will ever do. There are also the more curious users who with a push in the right direction will use the __CheckInOut.exe to check documents out and more importantly __ShowDetails.exe to have a look at a document in the web client. Users who want an advanced search can be pointed in the direction of the advanced search. A department may have a workflow requirement for their documents so they can have a workflow developed and get trained on how to complete workflows. This allows users to evolve to the level and interface they require rather than having one way of working imposed on all users.