Sandy Kemsley gave this presentation on BPM and Case Management hosted by Pega Systems.
Started by saying the Case Management is a hot topic. Lot of talk about structured proceses is for BPM, and unstructured for dynamic BPM or case management. But the truth is that many work situations need both.
F.W. Taylor can be considered the father of modern BPM, and we still see a lot of his influence in BPM today. But not all processes are so well defined. Peter Drucker coined the term “Knowledge Worker” to describe people who think and create the process as they go along. Some definitions from Mastering the Unpredictable:
- Routine Work defined as work that can be analyzed and a common pattern derived.
- Knowledge Work does not have the level of repeatability found in routine work. Not even possible to have a fixed process, and would be too expensive.
When you have a process that involves some part knowledge work, what people are doing today is stepping outside the process system, using email. You want instead to find a way to keep this work inside the process system.
Instant Poll: how routine is your work? 0% all routine, 19% is 3/4 routine, 27% have half and half, 37% is only 1/4 routine, and 17% of the people say they are completely knowledge work.
- BPM is defined as a way of defining a process, and maintaining it over time.
- Case Management is defined as useful for collaborative, dynamic, and information-intensive processes which requires incremental and progressive responses from the case handler to determine actions.
The key thing is that it is dynamic, and you don’t know ahread of time what is going to happen. There is no set of steps in a particular order. Does not mean they can do anything: there are rules and guidelines. But it is the case manager that determines exactly what to do next.
OMG has put out a request for proposal for a case management standard. Case management is really Drucker’s “Management by Objective” supported by technology.
Comparison table: BPM is highly repeatable, while CM is unpredictable. BPM focusses on transactions, while CM focusses on knwoledge. Goal of BPM is to replace people where possible, but CM is there to facilitate the case worker. An example of BPM is straight thru processes, while an example of CM is managing the care of a patient.
There are few times that you have just dynamic process or just structured process. Structured process may spawn a dynamic, collaborative step within the structured process. Some vendors have a “collaboration step” in a structured process. Use case is handling an exception. Other examples are a dynamic process that needs to call a structured process at a point in time, of an otherwise dynamic process.
Non-routine work is today being done most in a manual way. If they are manually handled, there is very little visibility of what is happening, and there is very little control of compliance. Visibility would allow others in the organization to see what is happening. Sandy has seen this, huge thick files physically at one person’s desk, and it is very hard to share this information.
Case Management allows for better visibility, and tracking of artifacts. But still allows easy reassignment of subtasks determined at the time of the work.
Emily Burns is a Senior Marketing Manager leading Pega’s Case Management effort.
Relationship of case to process: one to many and heterogeneous. One challenge of case management is that the scope is so broad that it is hard for people to see the entire subject. Like the parable of the elephant. People are largely unaware of the work that others are doing.
- Case Management is the coordination of multiple tasks, planned or unplanned, and associated with content, towards a concrete objective or goal.
Example is an automobile insurance claim. Consider an accident where two cars collide. There is a lot of context that must be included in the case: for example the prior case history of the parties involved. There is content that will be generated: photos, police report, etc. Context drives this case. If the context was different, the entire case would be different. This case has spun off three sub-cases: bodily injury, and two property damage cases. These sub-cases may be handled by different people. Challenge is to share the right information, without compromising on privacy guidelines.
Imagine what happens if the repair shop comes back and says that the original good-faith estimate was wrong. Originally thought it could be repaired, after looking into it, they find that the vehicle is a total loss. This will cause at that time a complete change in the course of handling for the case.
Note that the process is represented here, but it changes, and begs an essential question: how much of the process really should be captured? Options include complete specification of all tasks, a loose collection of tasks, or even no up-front specification of tasks.
The relationship of case to process: one to many. A single case will have many processes. The key hallmark of case processes is that they are nested and heterogeneous. Different levels of structure for the sub-parts of the overall case. Case processes are dynamic and event driven.
Forrester says there are three overarching case types: investigative, service requests, and incident management. Some cases have aspects of more than one type. Also, a case lives a long time, and may change over time from type to type. Cases must interact with other cases during their life time. A case may re-open another case. Or you may have a number of cases that are duplicate and need consolidation.
Pega Case Management: support all types, all roles, and all processes. Example is British Airport Authority as a case management approach to turning an aircraft around. Covers the span of time that an aircraft enters British airspace, until it leaves again. Interesting because channels are different: e.g. radar. IT does not get any more dynamic than air travel. Which gate will it be assigned to? When will it leave? None of these workers consider themselves “case workers”. Some of them don’t even work for BAA, but still play a critical role in the handling of the case. Can be effected by completely external events, such a fog at Paris airport. (or a volcano?) This coveres the entire range: some processes are completely automated straight-thru processes, while others are completely ad-hoc. Results are an increase in reliability and on-time handling of planes.
Question: What about modeling? What effect do you think this dynamic/structured divide has on modeling. Can you model structured processes in the same way as dynamic cases? Sandy: some discussion of BPMN, and some fragments might be handled, but for the most part BPMN modeling tools are not designed for case management style processes. We do still need some sort of standards for case management modeling which is likely to be different from traditional BPM modeling. Emily: we really need something to allow a process to start as an ad-hoc manner, and then later as you decide it is more structured there is a way to transition to structured, and vice versa.
—————
My take: they got right! Well done! The session is recorded, so I urge everyone to see it. Here is a link to the recording.
Thinking Matters 
Keith, I have seen you put together some diligent stuff in past, but this one – salutes!
I attended the same presentation and I can say you have captured every little bit of detail that was touched upon!
Well, I have a quick thought about this whole process v/s case thing – It’s not about the architecture, but about what organization wants to do and their culture. http://wp.me/pN8i1-5r
Thanks for putting this down, really!
Cheers,
Ashish
Pingback: Tweets that mention Is Case Management and BPM the same? « Thoughts on Collaborative Planning -- Topsy.com
Thanks. Sandy always does such an excellent job documenting talks and presentations, I felt this was a chance to try and do the same for her presentation.
On your blog, I posted this response:
The discussion of case vs process is like the discussion of whether a vehicle is an “off-road” vehicle or not. You start to define the exact qualities of an off-road vehicle, and the definition slips like sand between your fingers. Off-road vehicles have large wheels: but lots of regular vehicles have large wheels. Off-road vehicles have 4 wheel drive, but not all of them, and lots of road-cars have four wheel drive. Off road vehicles tend to be less luxurious, but there are notable examples of luxury off road vehicles. In the end, of course, ANY car can be drive off the road, it is just more risky in some cars.
Does that mean there is no difference? No, if your desire is to drive off road, then there are configurations and capabilities that you will want to make use of. More importantly, there are certain things that it is very important not to have, such as flimsy construction or very low clearance. We can talk about “off-road features” even though this does not necessarily identify a separate category of automobile.
Keith, from what you describe above, I find it surprising that you and I are so often debating case management and BPM – from this post I would think we agree 99%. But some of the statements you make above seem at odds with some of the statements you’ve made in previous discussions in the last week regarding case management – and the idea that if we’re modeling anything ahead of time we’re doing it wrong. Or that it can’t be done in a BPMS… yet Pega is billed as a BPMS, and apparently can do it… so is it also an ACM product or is ACM distinct from case management?
Just trying to reconcile your previous positions – because I read sandy’s post and found nothing to disagree with – and your summary of the points made – nothing to disagree with. So why so accommodating of a “spectrum” now, but in previous discussions the “spectrum” argument was so disagreeable? 🙂
Thx for the thorough write-up-
scott
Scott, I am glad you agree. I have always felt that the discussion was getting hung up on something that was not really there.
Look carefully. It says very clearly that there are two different patterns of work, and they need very different treatment. Heterogeneous means that different approaches are mixed together. The case management is NOT handled like BPM, and the BPM is not handled like case management.
The thing that you and I have been hung up on is when you insist that case management and BPM can be handled the same. Pega does not say that all work can be handled by a BPM approach, Sandy Kemsley does not say that. It does, in fact, require a distinct approach. It is just that their product offers both.
Perhaps when I say it requires a distinct approach, you have been thinking that I am saying that it requires a separate product. I actually have never said you need a different product. Pega, and Fujitsu’s Interstage BPM, offer multiple approaches, and I am quite familiar with the latter. Go back and look: I have never said you need a separate new product, I have only said that knowledge work can not be handled with “BPM” (the method).
We need to talk about the ACM approach because saying that one “has BPM” does not guarantee that one “has ACM”. BPM, as a method, does not include ACM capabilities. We need a label on those capabilities necessary for handling knowledge work.
Does that make sense? Or am I splitting straws?
Keith, no one will be surprised that I am in agreement with your view. 😉
Clearly, an adaptive case must be able to contain structured processes in editable models. Those can be BPM flows, but the typical BPM system cannot handle the higher hierarchy of the case. Pega is not a flow-oriented system but compiles rules into executables that represent process fragments. That allows for the modularity that they speak about and a case handling paradigm can be defined this way too. It is however an encoded paradigm that cannot be changed by the user by adding new entities and new rules. So PEGA is not truly ACM but it is CM with underlying structured processes. Additionally they do not handle content logic and data, except when they hardcode Word into it. I think Interstage is the same. Don’t get me wrong PEGA and Interstage are the best systems out there in the BPM arena in terms of technology.
With the ISIS Papyrus Platform we took however a much broader approach and allow anything to be changed before and during execution controlled by change management authorization. So you can define any combination of adaptiveness on the case or subprocess level, as well as for data entities, rules, content and user GUI presentation. The reason we did this was that neither BPM nor CM or both combined are sufficient to deal with the dynamics of business and human intuition.
What businesses really need is an adaptive process environment that encompasses and consolidates what they do today with ECM, CRM, BPM, BRM, CCM, CM and a few more TLAs …
Keith-
I think actually you have hit the nail on the head for a lot of our seeming disagreement:
1. in some of the discussions (not all) i took you to mean software not just “method”…
2. because i’ve dealt with so many case management solutions under the auspices of bpm programs, my internalized definition of bpm is not as narrowly defined as the framework you set up to explain the difference between bpm and acm. This is not the same thing as saying we tried to turn case / knowledge workers into automatons or replace them with automation – we adapt our approach to the situation. Soooo…. I think that has a lot to do with why were talking past each other.
Thanks for your patience 🙂
Pingback: The Knowledge Between Your Ears « Welcome to the Real (IT) World!
Pingback: links for 2010-05-15 « steinarcarlsen
Keith:
I think the issue here is not the definition: what is BPM what CM is. I consider it’s important to define things, otherwise business people, process people will misinterpret the idea and at the end hurt is done.
The issue here is that BPM continues to be and it will be the philosophy to manage business processes.
Some are structured with more or less flowcharts and documents (some structured processes don’t need documents and charts at all regarding people qualification to perform the tasks!) and some don’t.
Now for those that aren’t structured apply the appropriate definitions like you and others are trying to implement for the execution of such processes types.
You cannot separate BPM from CM, CM process type are manageable like any other process, it applies the same concepts like monitoring cycle time, task time, resource levelling, the only big difference it’s the fact that it can’t be designed like a “predictable” process clearly defined and executed most of the time by the book.
Alberto,
Consider the term “work”. Some will say that all work is a series of sequence of activities. OK? And a process is a sequence of activities. So, therefor, all work is a process, and the management of all work is business process management.
With this world view, one can then conclude, that since CM is way to manage work, and since all work is a process, then CM is (a type of) BPM.
I have heard this argument a lot in recent weeks.
I do not, however, believe that BPM is just a fancy word for management of work. BPM is a philosophical approach to management that says that the “process” is central to doing work, and that the process can be abstracted away from the work situation. That process can be evaluated in terms of how “good” it is, in abstract from the specific situation, and historic information about past situations can be used to “perfect” the process. When this philosophy is followed properly, the process is defined and prefected to a level that requires a specialist to (safely) make changes, and so the process is designed to be used in situations without modification.
CM, however, does not require a process which is abstracted from the situation. It is perfectly reasonable to start a case without ANY process defined at all. Yes, you can (when it makes sense) make use of processes (sequences of tasks) but that is nothing more than saying with CM you “do work”. Doing work, is not the same thing as doing BPM.
As part of handling a case, you might invoke a BPM process, but that does not mean that the original case was being handle by BPM. For example, in the middle of a corporate merger, you might decide to purchase something, by triggering an automated BPM purchase process, but that does not mean that the corporate merger is all just BPM.
It is very easy to say that “all processes can be managed” and therefor the approaches are the same, but saying so is ignoring some very important elements of the approach. (Similar to saying that Chinese Food and French Food are the same because they both involve food that can be eaten.)
On your last paragraph: BPM, remember is about perfecting a process through a continual cycle of improvement. Then you attempt to say that CM is BPM without predictable processes. But think about that a bit: you can’t have a cycle of improvement on a process that can not be predicted. You can not abstract a process and attempt to perfect it, if it can not be predicted. If you remove the continual cycle of improvement from BPM, then BPM is nothing more than managing work.
Both BPM and CM offer ways to manage work, but BPM focuses on a way to optimize routine processes which can be abstracted away from the situation, but CM seeks to make the case manager effective in unpredictable situations where the process can not be abstracted away from the situation. Different approaches, with different outcomes.
Keith, you make valid philosophical points. But Alberto makes valid practical points.
I don’t think the response you’re giving him tells him to do things differently than he described himself, you’re just arguing that the definitions are really important. He makes the practical point and you dive into the philosophical background and definitions.
There are two general schools of thought about definitions (imho). One school says, the definition is whatever is in the dictionary (or whatever the key reference source is), and people should only use it correctly that way. The other school says the dictionary *reflects* the usage, and therefore usage is the source of meaning, the dictionary a reflection of it, and merely speeding the standardization of new forms and usage. The word “epic” means things it never meant when I was little. We used to search for things before we learned how to “google” it. And “googol” used to be a very large number.
To the extent that BPM ever meant that our lives should be reduced to finite automata, I don’t believe that is the current usage.
Let’s not forget, BPM used to be beats per minute…
Scott, I left a note on your blog, but I thought I would mention it here: can you provide some example of work that is NOT BPM? We have seen lots of examples of what you believe is included in BPM, but it would really help us all understand if you can list some example of work that fall outside the realm of BPM.
Pingback: Column 2 : links for 2010-05-25
Yes, it is. I really do not understand why we tend to separate IT from the foundation science and they applied udage sciences that gave birth to it.
I totally support Alberto’s implicit message.
I would also suggest reading -in whichever order-: Mastering the Unpredictable and The Emperor’s New Mind.
Thanks, I like this post very good and informative related to this topic.