Phil Ayres writes a provocative article on his "Improving IT" blog: "When Business Process Management (BPM) is a waste of time" that BPM projects fall into only two categories:
- BPM automation
- Everything else
He goes on to say that really only the first category should be pursued with BPM.? The basic argument: it makes sense to try to replace people with software programs, rules, algorithms, and database lookups, but it does not make sense to "model" human-to-human interactions and attempt to improve them via BPM.
Well, to the first part:? to the extent that processes CAN be automated, many of them already are.? However, as business innovates faster (generally) than the IT that supports it, there seems to be an endless supply of new processes that have a lot of inefficiencies for the humans that perform them because thing that could be automated aren't (forcing data re-entry for example).
To the extent that something can be reduced to a software program, I'll agree that it should be, if you want your business to run efficiently.
However, I think the reason I fundamentally disagree with Phil is that we are starting with different definitions of what BPM is.? By my way of thinking, a software program is not a business process.? If the program is defined in such a way that its beginning, end, and middle only touch other systems, and touch no human interfaces, then we do not, actually, have a business process. We just have a program (or a non-business process).? The subroutines that my computer runs to keep it humming smoothly (garbage collecting memory for example) are not business processes. Business processes touch people: customers, suppliers. employees.? The job of BPM is to make those interactions between a person and the Business Process more efficient.? Not necessarily making person-to-person communication more efficient, but interaction with the business process.? As a customer, my experience should improve in terms of ease-of-use, or turnaround time, or quality of result, or cost.? As a supplier, my experience should improve in terms of fewer cancellations, more timely notifications or feedback, etc.? As an employee, I should have more information at my fingertips that is relevant to what I'm doing, and I shouldn't need to know the ins and outs of myriad IT systems to do my job, and additionally I shouldn't have to re-key data.
In my opinion, most human-to-human BPM solutions (i.e. those that move work from one human worker to another) end up becoming glorified collaboration tools with a few rules scattered around. Most of the time and effort involved in implementation of new BPM solutions becomes a matter of working out how to make the workflow flexible enough to meet the many interactions that real people in real offices must perform to get work done.
I think what Phil is reacting to is the fact that in many process modeling exercises, there is entirely too much focus on modeling interactions that should not be modeled explicitly. ?? An example?? During an interview process, you might want to conduct a meeting at the end of the day's interviews to review all of the feedback for each candidate and make final decisions (or decide on next steps).? That meeting could be modeled to infinite detail about getting feedback of each type from each interviewer for each candidate and require each person to record approvals... Or it can be modeled as a single activity that the hiring manager coordinates and captures the decision on.? From the process point of view, we only need to know that the "meeting" happens and that a decision is recorded by the hiring manager.? That "meeting" could be physical, virtual, email, whatever. We don't care.
And Phil is right - if you try to capture too much of the silly back-and-forth that real people engage in, you'll waste a lot of time and effort.? But that's why you don't model those parts of the interaction except in the vaguest possible ways, leaving the outcome up to the person most responsible or most appropriate at that stage of the process. It turns out there is still plenty of process improvement meat on the bone - because the inefficiencies are typically when you cross group or department boundaries and the lines of communication are less personal and immediate already - and then BPM gives you a way to steer and control your process (oh, and measure and improve it later).
Phil advocates not using BPM for these processes, but to "Go ahead, remove the waste from that process you want to improve" but then advocates not using BPM to do it:
[...] consider using a tool that is better suited to the type of work that is being done. BPM is probably not it. Find a tool that does not require 3 months of wasted time analyzing how to improve human interactions, before actually delivering anything.
First, I'd not recommend using a BPM tool that requires 3 months of wasted time analyzing how to improve human interactions before delivering anything.? If you're being confronted with that, drop us a line or a comment and I'll help you out with some recommended BPM software and methodology.? I just finished a 3 month project that went from conception to pilot in 3 months, with just a couple of weeks of analysis that happened before I joined the project. (Note: you need less analysis if you're already familiar with your process!)
Second, Phil doesn't explicitly mention which software tools he'd recommend, though he does generally recommend collaborative and case management tools that provide "structure around an otherwise unstructured set of operations."? Well, that's what your BPM software should be doing for you.? And if it isn't, get a new BPM vendor, or get a new BPM consultant to help you see the forest for the trees.
I'm sorry, Phil, but BPM software is not designed to "prevent workers from doing the many things that need to be done," but if that is your experience then you should out the BPM vendor who takes that approach.? Because it isn't true of BPM in general.? I wonder if your experience is with the stack-vendor BPM tools that are, primarily EAI tools with a little diagramming thrown in to look like a BPM tool...
Without BPM, you are going to be hard pressed to track the critical data points you need to measure and improve on your process.? I've yet to see any other software category provide the kind of valuables raw business analytic data that a BPM tool can provide.