My good friend and colleague Bruce Silver made a post recently on BPMN and Case Management positioning ACM as a kind of BPM. This is a mistake that many people make, and I thought it worth highlighting how this line of reasoning confuses rather than explains. If you define a case management using BPM terms and positioning it as a special case of BPM and then concluding that is it a kind of BPM system is not surprising. The logic is persuasive, but it starts from a misunderstanding of who uses an ACM system.
To understand ACM (or case management), one needs to remember that it is about non-programmers getting work done. That leads to a different conclusion. Lets start over from the beginning, but this time remember who it is that is “driving” the system.
- You are a doctor who just read about a new kind of treatment for a disease in a medical journal. You happen to have a patient with that particular illness, and since other things have not worked well, you decide to try this new treatment. You sit down to edit the BPMN diagram describing that patient’s care to include steps for the new treatment.
- You are a judge presiding over a court case. The defense attorney just made a huge procedural blunder. You and everyone in the court have invested a lot of time and effort into this case. You want to see justice done. You would like to avoid having the case thrown out for procedural issues, probably causing the whole thing to start again. You come up with an idea on how to satisfy everyone and keep the case going forward, so you quickly verify this with everyone, and then sit down to edit your BPMN diagram that describes the progress of the court case.
- You are an executive of a large oil company. You have just learned that there was an explosion on your oil platform. You are going to have assess the situation and come up with some pretty creative responses as this situation is not like any you have been in before. You get an idea, so you sit down to edit the BPMN diagram that describes how you manage oil drilling operations.
- You are a director sitting on the board of a consumer product corporation. A large industrial giant has just announced that they are launching a new product in a space your company has been dominant in for many years. Furthermore, their offering has a creative new feature that ties the product back to their very successful and stylish other product lines. You need to come up with a strategy. You have people who can access the real value/quality of the product. You have contacts who might be able to tell you if they are selling the product below cost and destroying the market. There might be patent issues. There will be wave of press interest around this product launch, and if you can get some halo leverage if you can get the marketing program together quickly enough. So you sit down to edit your BPMN diagram on how you run marketing campaigns.
The part where the people edit BPMN is unrealistic. The point should be clear: none of these people will ever use BPMN to describe the work that they need done. The doctor is too busy keeping up with changes in medicine to learn BPMN. The judge will not have a BPMN digram that has been kept up to date about the court case. The oil executive has no time to draw up a careful and fully detailed diagram of what is to be done. The executive has no need to go to the trouble to make a process suitable for reuse.
Nor will these case managers have BPMN experts sitting at their elbow to make such changes for them. Without the process expert instantly at hand to modify the diagram, the time delay will eliminate any benefit of having the diagram in the first place. To expect the judge to send off instructions to someone else in the department to modify the course case will take too long, even if it takes only a few minutes — which would be uncharacteristically fast for any government institution anyway. These are knowledge work cases, and they will never be described using BPMN. In every one of these cases, the value in making a clear documentation of what should be done is wasted because you are not likely to need to use that diagram ever again.
BPMN is a language that is used by process professionals. It is used to create a clear and unambiguous diagram of a process for repeated use. The only way to recuperate the cost of drawing the diagram is to use it many times, which is not what a knowledge worker is doing.
The stock and trade of a knowledge worker is use their expertise and knowledge of the specific situation to figure out the best path. Doctors incorporate 11 new facts into their practice every day. If a doctor had BPMN diagrams to describe their treatment plans, they would be spending nearly full time just keeping them up to date. A board of directors is presented with new and unexpected situations every time they meet. None of these professional will use BPMN directly: executives will never draw up action plans in BPMN, stock brokers will never use BPMN to describe their investment plans, police detectives will never use BPMN to manage their investigations, salespeople will never use BPMN to manage a sale, social workers will never use BPMN to manage social cases, nurses will never use BPMN to define care schedules, marketing directors will never use BPMN to design campaigns, …
Saying the all we need is an extension to BPMN in order to support case management completely misses the point of what case management is about: Case management allows the case manager to be in control, to leverage their expertise, to do what needs to be done. For the case manager (i.e. the doctor, the nurse, the lawyer, the executive, etc) to be in control, they need to be able to directly manipulate case. If the case depends upon BPMN to work, then the case manager must know how to edit and modify the BPMN, and in most professions they will NOT.
BPMN was designed by and for IT folks as a kind of computer language that hopefully business people can validate. But there is no evidence that business people will ever draw or modify those diagrams. That is a job for a process professional. BPM is for process professionals to use, and it is only useful in repeatable, predictable processes.
Case management is what the rest of us use: the process non-professional. ACM is not competition for BPM, it is competition for Email. Most case management today is done with documents sent by email.
Return to the post from Bruce Silver. It states as an axiom that “A case is a set of activities related to each other in some way.” Even at this first step the description avoids the fundamental aspect that a case is a place where an professional does work as an intelligent human being. What has been described is not case management, but instead a kind of BPM, with some extra flexibility. After that, we are not surprised to reach the conclusion that BPM can handle it. But case management is about real professionals doing real work in real time. Such professionals have no need to model anything. They are doing the job, not modeling it.
Case management is not something that the IT department prepares for the case managers. Case management is something that is done by the case managers themselves.
Thinking Matters 
Keith, while I am in utter agreement with you what ACM is about and who will use and that BPMN WILL NOT be used by the performers, I still would agree with Bruce the ACM is a kind of process management. But yes, ACM can’t be done through an orthodox flowcharted BPM solution. But that does not mean that there is no need for BPMN at all.
We have actually extended BPMN (which is permissible) so that it can be used for ACM, but the most important part was to define what must be left out to make it understandable for those who want to look at the progress of a case. A priori it will only be used by those who want to define sub-processes as parts of a case, for orchestration of user or system interactions (wizards). We enable the perfomers to add and adapt the progression, tasks and process resources without having to enter the BPMN editor. But people who then want to verify what happened and turn it into a template for future use find it practical. Plus it helps to get those BPM pundits on board who can not imagine it any other way.
It is important to make it clear that ACM provides suport for high-value knowledge activity and that the people who use it are neither technical not interested in dealing with the issues. Additionally I find BPMN in its original form utterly limited and too ambiguous. Plus it lacks support for process resources such as user interaction details, data models, content and rules.
Thanks, Max
Max, thanks for the response! I think the best way to draw a distinction between ACM and BPM is in the way that people use them. I know that all implementations include a little of both: BPM systems have some ad-hoc capability, and ACM systems have some pre-programming capability. That line is very fuzzy unless we talk about how it is used.
Similarly, all knowledge workers will perform a combination: some routine (automatable) work, and some unique (unpredictable) work. When routine work becomes automated, we should call that technology that automates it “BPM”, even if it is part of a system that claims to support ACM. A doctor, working completely in ACM mode will order “lab blood analysis” which is likely to be highly automated by BPM. Thus case managers will use BPM all the time, just like they use word-processors and spreadsheets.
There is still a very clear distinction between the ways that a person uses ACM and BPM. ACM is completely under the control of the case manager (in every aspect) for themselves to use. BPM is under the control of the process expert, and it provided primarily for others to use. It is the difference between these modes of use, that drive the different required feature set.
If we look only at capability, then one can argue (falsely) that a single unified language makes the most sense. It is that argumentation (based on the invalid assumption that one approach is better than two) that leads us to a dead end. Once you understand that the audience for ACM is different than the audience for BPM, only then can you realize that there is no real need for a unified approach.
Hi Keith, you seem to be suggesting that ACM should connect with a BPMS to drive the rigidly defined processes. Is that correct? That means that for those we still need the complexity of process design and optimization? What about processes that evolve from a case-user-driven to a flow- or rule-driven form and vice versa? Are you saying that it is better that there is s systemic break and the process has to be recoded to be supported in another tool? This is the only reason that I propose that the ACM environment must inherently support well-defined processes also. They could still be created by non-technical performers, but the rules and definitions makes their execution rigid. A doctor might also want to describe a regimen for a treatment and does not want to hand it over to a BPM expert. He might not have the time to do so to save the patient. He still wants to ensure that it is executed as defined. Later another expert might want to add or change something in that procedure and also does not want to refer to BPM designers to do so.
Therefore I do not see a clear distinction between ACM and BPM, except that a BPMS cannot perform the kind of ACM user-driven processes that you describe so well. My recommendation of a homogenous system that does both really well is not only driven by my product orientation. Remember? I don’t have to sell what I have – I develop what empowers people!
I see it the opposite way: The only reason people have to keep ACM and BPM as independent functions are sales perspectives or BPM design or consulting skills that might become obsolete. From a business benefit perspective, a homogenous solution that also encompasses architecture, content and rules is the only thing that makes sense. Agree? Whether this is easy to sell or argument from an existing BPM mindset is a completely different story.
Great post. I think what execs need is something light like email… and something that connects with email… since that’s where they live… but they also need to the ability (by them or their surrogates) to assign tasks, track progress, get alerts, etc. I’m not sure what that magic tools is… e.g. sharepoint/outlook are too generic… web based tools like Asana and Basecamp are a bit closer. Perhaps it’s not a tool at all but a people driven service.
Dave, good points. I think you have hit upon the key element: it is like email, but it has planning and task tracking capability. My overview of social tools with/without task tracking is here: https://social-biz.org/2010/12/11/social-has-no-future-yet/
Another key is that like email and telephones, the ACM needs to be deployed as an infrastructure and not as focused applications. It is a basic capability that knowledge workers live in. They need to have the assurance that all others in their organization have access to it as well. If some people in your organization did not have access, then you would have to remember to never assign tasks to them, and you could never trust that task assignment would work when assigning to new, unfamiliar people. For it all to work, it needs to be ubiquitous.
wow – great post. thanks so much for this link!
Keith, I think it is utterly confusing to recommend that we refer to processes that are done within an ACM system once as ACM and others that are rigid as BPM. Yes, one could say that ACM does cases and BPM does processes. But then I would say that ACM does both cases AND processes. I would say that ACM is an evolution of BPM concepts. An ACM system is an implementation of a goal-oriented BPM concept. I have been writing on a blog post on this subject for some time, but I felt I was going to open the ACM/BPM discussion again and kept out of it.
You do say in your last comment that ACM has to be ubiquitous and an infrastructure. I totally agree. But that means that it also does BPM-like processes (with flowcharts), it adds Social and it adds the goal-oriented architecture. Also a doctor adds case-fragments that have clear and sometimes independent goals. A patients blod test might show that a treatment is necessary and it has to be controlled by the frequent goal-verification of repeated blood-tests. All the data elements of the blood tests must be architectured to ensure that there is no ambiguity.
Not sure why we are having this definition problem. It seems utterly clear to me.
Max, I would disagree with you that “An ACM system is an implementation of a goal-oriented BPM concept” is missing the point.
ACM has many services needed in order to function. A BPMS is one of those services since we don’t have a better way today of executing a process/task/process fragment/activity/emergent process/knowledge driven process/etc. If you have a tool that works and you can make a tasking model which runs in a BPMS then great.
However, I think that a tasking model is the key here. No matter what it runs in BPMS or other from a runtime perspective. There is a true need (and many vendors are getting there) to have a flow of work not a workflow defined. The goal of a flow of work may seem the same as a workflow, but the driver of a flow of work is the case/knowledge worker not the system as in a workflow.
By the way, I have found this posting and the discussion thread very interesting here.
Pingback: Lamenting Definitions » Process for the Enterprise
Pingback: BPM and Case Management Quotes « Adam Deane