Agile BPM versus Adaptive Case Management: Again?

Why is there an ACM versus BPM debate? There is this assumption that ACM is in some way anti-BPM. This is wrong for a number of reasons.

First there is the human interaction between different people, such as various vendors, consultants and analysts who all have their reasons to say what they say, but only when I say it, then it is not a good reason because I sell software. As this is related matter to my business you will find my position on this on my Papyrus blog. I also try to be shorter in my posts as widely suggested. ;-)

Second, it is ME and not ACM who is anti-BPMS (note the S for Suite or Software!) in the sense of using flowcharts to all-out organize a business. I am a BPM believer in terms of outcomes. In a briefing yesterday, Connie Moore of Forrester Research told me again that she agrees with me on the limited benefits of flowcharts. I am not alone. Read the comments on my blogs. ACM is further a very loose concept that I have tried to define better with the help of others. No luck so far. There is no agreed upon definition of what ACM is, so it can’t be anti-anything. It’s only poor me who tells prospects that they can actually get rid of something they are wasting money on. Let’s even throw out some ECM and CRM stuff as we speak about it.

Because I am part of that ACM ‘movement’ there are those who associate my BPMS stance with the stance of ‘ACM’. That is too much honor. However, my preference would be to deliver a BPM solution, because what I propose is clearly and utterly process oriented. I tried to convince the others at the WfMC to call it process and failing that I proposed the name ACM that was agreed upon. I immediately posted afterwards that I am proposing what I call ADAPTIVE PROCESS beyond ACM. Yes, it can be used to deliver the most holistic ACM approach currently available and it also can replace any BPMS (and Social btw). But why would you want to do that if you are happy with BPM anyway? But are you really?

It was once again Adam Deane who tries to bring light to this by restarting the ACM/BPM debate. I just love his weekly quote summaries. While I don’t agree with him on ACM/BPM, I admire his guts and openness to come out and say it. We need this debate. He also has hilarious ideas how to write about it. He also clearly defines himself as a BPM advocate when he says that I am ‘yelling’ at them (the BPM people). I do apologize for my naughty ways! But please be adult about it. Adam was however the one to say that the BPM plane is stuck on the runway. I totally agree, because the BPM airline business class is way overloaded with bureaucrats and system integrators. Adam’s posts show many of the fuzzy areas that people have about ACM. Partially my fault if we haven’t been clear enough about it. Or maybe my posts are too long …

Let me first say that I see ACM filling the huge gap between BPMS and Social Media. ACM uses elements of both and links to both as required. Yes, my kind of ACM can also replace a BPMS in a cinch and provide a customer focused, homogenous Information Workplace.

 

The Spectrum of Corporate Processes from ERP to Social.

Adam first suggested that ACM is really kind of a ECM content transsexual, that also has a ‘bulge’ meaning BPM process features. He is absolutely right. I see it as essential that an Adaptive CASE (not Content, Adam) Management approach is about CONTENT AND PROCESS! Once again, there is no process without content and content without process is irrelevant. Most ACM ‘wannabees’ have however no content features. A first opportunity for confusion, but a clear distinction to BPMS that also have no inbound and outbound content features. Some ECM systems can combine content and workflow features but most need Eclipse to integrate them.

One question is if a BPMS can be used for an ACM infrastructure. There are some overlapping features in ACM and many BPMS, but the approach is utterly different.  The key difference is when and who designs the process, which is not just the flowchart. ACM (and Adaptive Process) move the process design from an analysis or discovery phase into the execution environment. The Forrester analysts watching my demo yesterday said that this is the most holistic approach they have seen so far. But more in a minute.

Another point raised is BPMS Ad-Hoc capabilities. ACM is not just about allowing tasks to be rerouted to anyone else on a whim. That is completely irrelevant. In ACM the authority to reroute tasks can be assigned based on the needs of the business organization. You want to nail it down, do so. In most cases work tasks should anyway be assigned to USER ROLES and not a particular person. Reviewing someone’s work should not happen by rerouting the same work task, but it must be a distinct review task to keep track of it. In other situations should the business user have the authority to pull-in such a review task on the fly without breaking the flow. He might be authorized to skip a task. If goals are fulfilled, other waiting tasks become irrelevant. Business rules might cancel or change task status. A user might also have the authority to create a delegation task for some aspects of his work item. Delegation is not just work reassignment, but asking someone to perform other work in addition to my own. Some BPMS support Ad-Hoc task creation and work delegation, but it is not a key difference anyway.

Adam defines one of the key distinctions: (In BPMS) ‘Process Design and Process Execution are separate entities.’ In ACM however, you DESIGN while you EXECUTE and it is not the same as Social BPM design that is also before execution. The principal concept of ADAPTIVE is that knowledge from execution can be fed back into the templates. In ACM you also optimize during execution as you look at goals and metrics in real-time. AGILE BPM needs the governance bureaucracy to manage design before and optimization after execution.

Adam’s next point is that the execution elements ‘Reporting, Simulation, Forms GUI, Emails, Documents, Roll-back …’ aren’t considered in BPMN during design. Correct! And that is the next ACM distinction to BPMS. In a BPMS, all those elements require substantial additional effort often including programming to make them usable. So when you are done with your wonderful BPMN flowchart, you have at most 20% of the final process functionality. Those  missing 80% need BPM governance to manage. Even Human-BPMS that are very agile and dynamic do not enable users to make all those changes to the process during execution. And why not, for some organizations that may be just what they want! Please, buy a BPMS if you see no other need at all. You will still end up needing adaptive processes as well.

But here now comes a point that does not align with everything that Adam said before. Maybe he can explain it to me in a comment here. He says: ‘The whole IT industry is going SOCIAL …. here is a technology that has implemented the ACM concept. … ACM missed that opportunity.’

Adam, I am sorry. But you are jumping to conclusions. Several ones. First, that Social is not used in ACM. It actually is with chats, blogs, wikis and dynamic user groups. My kind of ACM is a lot more Social already than any BPMS ever can be. I have stated clearly how I see Social Media and how we use it. I am in utter disagreement that simple Social Media are anywhere close to ADAPTIVE, while AGILE BPM bureaucracy is exactly the opposite. Social Media is CHAOTIC and only if there is a learning capability it could be emergent to new knowledge. BPM is not learning because the governance is totalitarian. That is the weird thing about Social BPM. On the one hand they want to control everything, but in the middle of it they are going social during design and execution? If you add Social you break the governance and/or you break the process flow. So it is suddenly not needed? Why then bother with it in first place?

The other assumption Adam makes is that Social Media solve the same problems that ACM solves. Absolutely not. They create new ones! I wrote a blog post on Activity Streams to explain just that.  A social interaction that is not tightly linked into a process (as it is in ACM) does not produce value. Social interaction implies unstructuredness and unpredictable behavior so it would be a lot worse than the email mess we already have. ACM can embed Social because it has no restrictive flow! How would you embed a social interaction into a process flowchart and control it? How do you pass data, content and so on? Yes, as hyperlinks in the message, but then what happens in that link? You open up the process GUI or you loose control. But how is that now related to the predesigned flow, because who will know who answered or changed what in a Tweet response? Maybe Social has a little tweeting birdy in its cage? No, it is again pure hype and social lipstick on the BPM pig. Empowerment is not about Social media, but about authority, goals and means.

Some points Adam didn’t cover: ACM is in difference to BPMS a natural with events, rules and goals. BPMS miss the capability to deal with random events. Therefore, Michael zur Mühlen cuts BPMN back to simple ACM-like task management. Event listeners in BPMN mostly break the flow and it is very difficult to resume normal execution. Flowcharts virutally can’t deal with exceptions or with unknown events, like one more message or document that no one thought of? What if a fraud pattern is detected (i.e. by the UTA*)? How do you deal with all the complex interactions between a large set of processes if not with events? How do you keep business rules and the related data in a BRM engine in sync? How do you trigger rules from events and vice-versa? Can the business users write rules?  How do they map them to business data, content or tasks? How do process owners define goals and link them to objectives and metrics? How are meta-data models linked dynamically to all of the above? Find the answers in your favorite BPMS and you will know why I propose ACM.

Clearly, BPMS or their marketing brochures are being enhanced all the time and soon they will all have ACM features. It is a silly comparison. Many others in the ACM ‘movement’ also don’t offer all of the above. If it makes anyone happy, I am glad to call my solution Adaptive BPM or ECM or CRM (and I have done so) but people will just shrug it off as they are so jaded from the irrational claims in the IT world. Adaptive, dynamic, agile, it all makes no sense to them. Regardless of all that and what we call it, it won’t change my stance towards flowcharting as being destructive to the real-world, human dynamics of a competitive business.

PS: Sorry, that this post is so long again. 8-(

* UTA – User-Trained Agent, the unique Papyrus pattern matching engine that learns and executes in real-time.

2 Comments on “Agile BPM versus Adaptive Case Management: Again?

  1. Excellent response Max, as expected.

    You are a true mover and shaker…

    George Konrad once said:
    You take a number of small steps which you believe are right, thinking maybe tomorrow somebody will treat this as a dangerous provocation. And then you wait. If there is no reaction, you take another step: courage is only an accumulation of small steps.

    Cheers,
    Adam

    Like

    • Thanks Adam, sorry again about the long post. Please do not stop to challenge me or others. It would be a pity and we would all lose out. You are a steadfast and honest rock in an otherwise opportunistic world. All the best, Max

      Like

Leave a Reply

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

WordPress.com Logo

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

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: