Case Management à la Appian

A four part series (maybe more???) from from Malcolm Ross of Appian on how they support Case Management.  Appian is a product I know well, and can recommend because it is based on my ideas of Collaborative Planning.

Part 1

This post states that case management is data centric. This is correct, because when the process is not predictable, how can you use the process as the organizing principle?  Since you can’t, what else is there: data.  The example: a patient, and this is a good example;  in a case management system you organize all the actions around the patient and the patient record, which *must* exist in advance, and through the entire treatment of the patient.

Part 2 – Ad-Hoc Activities

The term “ad-hoc” is correct but problematic.  Being latin, it actually means “customized”, so literally an ad-hoc process means a customized process, which is more or less correct for case management.  The problem is that most people believe that “ad-hoc” means “random” or in no particular order.  This is not the case in case management.

Appian’s support for “ad-hoc” is fairly good, but you should be aware that of a slight bias in the example.  It states that there is a designer of the process, and then later a case manager who can choose to activate a particular activity.  This excludes the possibility that truly unexpected activity occurs.  Because of the separation of the process designer and the case manager, if the case manager needs to do a “third thing” which is not in the list, that case manager is completely out of luck.  True, Adaptive Case Management would allow the case manager to add a task without having to learn BPMN notation, and without having to have activities in advance.

This is not a critique of the Appian product, but you simply need to be aware of the “pre-planning bias” in the example given.  It seems that many people think that an a “ad-hoc process” is one in which a finite set of options is presented in advance, and the case manager needs to pick one of the pre-defined activities.  This will handle many cases, but not all cases that a knowledge worker needs to handle.

Part 3 – Real-Time Events and Business Intelligence

I *love* the example of Dr. House.  Not only am I a fan of the show, but it is a great example that illustrated the kinds of emergent processes that a knowledge worker can run into.  When you consider a combination of a few diseases, a patient can literally be 1-in-a-billion, and the doctor may be handling a case that has literally never been handled before.  There will be no pre-built process for this case.  Medical science advances at such a pace that doctors will be trying techniques which literally have no precedence.  A technique such as BPM, which is based upon predefined processes can never apply in these situations.

I like the description of Dr. House as the knowledge worker, and Dr. Cuddy as the “management”.  This is indeed the dichotomy in many workplaces, and the show makes evident the wisdom that Dr. Cuddy must employ in order to succeed.

But again, I must caution the reader.  The example supplied is one of a pre-planned event, and a pre-planned response.   In every episode of “House” the team is subjected to an event that could never have been anticipated in any imagination.  Can you imagine Dr. House sitting down and drawing a BPMN diagram at that point?  Can you imagine Dr. Cuddy drawing that diagram?  Of course not.  The event would be gond before the diagram was drawn.  The “real-time” events that occur in these situations are not predictable, and would need to be handled without process modeling.  It is ironic to show a pre-planned model as if it could be used in response to an unanticipated event.

Part 4 – Enterprise Content Management

Here he talks about content management, and this is simply part of the “data orientation” that is critical to case management.  Data can be structured or unstructured, and ECM is usually used for the unstructured.  While we would like all information to be represented in the structured form, case managers must live in the real world, and that means handling documents from elsewhere.

Strangely, Ross mentions how BPMN is adding content management features. That is all fine, but can you imagine Dr. House sitting down to draw a BPMN diagram?  Not me.  Dr. House is going to exhibit what we call “direct manipulation”.  That is, he is going to “do” something instead of creating a model to do it.  A true ACM system will allow a user to “do” something directly, and then capture this action and preserve what was done in history, so you can learn from it next time.  Dr. House does not want to model the world, instead he lives in the world.  It is a difference.

To truly understand case management, we need to escape from the pre-planned application development mind set.  We need to see the world as something that a knowledge worker swims in, and we need to design systems that help knowledge workers, without placing artificial constraints on how they do it.


Malcolm Ross: this is good stuff. You are on a roll.  I can’t wait to see how Enterprise 2.0 fits in the mix.

5 thoughts on “Case Management à la Appian

  1. Keith,
    As usual, great post.

    I don’t agree with your statement “It seems that many people think that an a “ad-hoc process” is one in which a finite set of options is presented in advance, and the case manager needs to pick one of the pre-defined activities. This will handle many cases, but not all cases that a knowledge worker needs to handle.”

    Only BPM vendors and analysts feel this way, not most real users. For them ad-hoc is what you do in email – no preplanning (or programming) required – the participants themselves own the process flow.

    I agree that using a BPMN-like modeling language for ACM isn’t viable in the real world of unpredictable processes – and certainly making BPMN more complex (i.e. adding new content mangement constructs) won’t help.

    That leads me to another problem. Documents play two roles in most real world case management scenarios – as content used and generated by the process, and as content that drives the process. Content and flow need to be linked in both directions – in other words the document should be aware of context of its actual usage within processes.

    This has deep implications for BPM suites, implications that most vendors ignore. They are hoping to use the buzz for their own purposes, but don’t really support adaptive case management

    Jacob Ukelson – CTO ActionBase

  2. Pingback: links for 2010-05-19 « steinarcarlsen

  3. Pingback: Column 2 : links for 2010-05-25

  4. Pingback: Library clips :: Have we been doing Enterprise 2.0 in reverse : Socialising processes and Adaptive Case Management :: July :: 2010

  5. Pingback: Barely a Year Old, and ACM is Dead » Process for the Enterprise

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s