Well it is time to say good bye to 2008 and hello to 2009. Will it be a happy new year in IT for all? Will the global economic crisis play scrooge? Here is what some of the experts have to say:

Gartner
http://blogs.zdnet.com/BTL/?p=10403

Forrester
http://www.networkworld.com/includes/ads-pre.html

IDC
http://www.russoft.org/docs/?doc=1581

All of them see a slowing of growth but not a contraction. As for my prediction well I am a little bit hopeful that BPM which is viewed by many purchasers of BPM products as a cost cutting technology/exercise will be spared the wrong of the economic turmoil. Well thats it for me for 08. I hope everyone has a happy holiday season.

Every now and then I do a little playing around with Google Trends to find out how the BPM industry is tracking.  I typed in the following search terms:

workflow

Business Process Management

BPM

bpel

bpmn

software (used software just to get a base line) 

If looked at only by themselves workflow, BPM, and Business Process Management would seem to be contracting. Google does not create these charts against total searches but as a percentage of total searches. If we use the term software as a baseline it would seem that these terms are relatively stable. The interest in the BPEL and BPMN languages seems to be increasing and fairly dramatically. I guess this may be expected due to the relative youthfulness of these terms as compared with the others. Additionally BPM may not be a very valid search term as many people might be searching on “Beats per minute”

I went to a BPM Trends seminar a few months back. It was filled with a lot of interesting information and if you want to find out more about how the BPM industry is segmented I highly recommend you drop by for a session. One of the most interesting and somewhat comical things that they said at the session was there was a list of the top concerns/issues from CEO’s and CTO’s. The lists were both very different with only one commonality. Both groups felt that the IT and Finance sides of the business did not work very well together. It got me thinking as to why that could be. Two possible reasons stood out.

One reason could be that there are simply not enough people in the world who have a good understanding of both skill sets. There is some much terminology today that either group could listen to a conversation the other group was having and not make any sense out of it. It is also very difficult to become and expert in both as many people would not seek to attend university for a period of between 10 and 15 years.

The other reason is perhaps the concept of what a business is differs greatly between the two camps and thus they have difficulty communicating. The IT person might view the business as a series of multi-directional interconnected people, systems and processes. To them the business is conceptually a network.

The CEO might view the business as a people and systems belonging to business units which are part of divisions which make up the company. In their mind reporting paths and org charts most clearly define what the business is. To them the business is conceptually a tree.

Please keep in mind that I am for the most part and IT person and I am approaching this from an IT bias.

To be honest I really don’t know the effects of the credit crunch on BPM but I am just going to state some reasons why it may be a little less susceptible to the credit crunch and some reasons why it may be a little more so.  Again these are just my musings.

Less susceptible

BPM is often purchased because it is a cost savings measure.

BPM may be seen as a saviour to so many issues and may be looked at out of a sense of economic desperation.

BPM offers greater business agility which could be very important to respond to businesses that are in a great state of flux.

More susceptible

Usually a visionary is responsible for getting the momentum going for BPM.  I believe that in times of economic turmoil their ideas may be overlooked as Management is more concerned with key business fundamentals.

For most BPM vendors there is a large initial capital outlay.  This is where BPMSaaS vendors may have an advantage.

My father once told me that historically the best thing to invest in during an economic crisis is waste removal.  Everyone has to have their garbage taken away.  While you could argue that BPM eliminates waste it is not a necessity to the same degree.

I came across the following blog article today:

http://blogs.wsj.com/biztech/2008/09/29/how-to-make-business-more-interesting-video-games/?mod=googlenews_wsj

According to the article a San Francisco State University professor is intending to teach students about BPM by immersing then into a virtual world where they can study tasks and use technology to improve them. 

I think anything that gets more people involved with BPM is a good thing.  Currently there does not seem to be enough human resources capable to deliver all the BPM work that is required and I feel we are still in the early days of BPM adoption.  BPM spending is suppose to double over the few years depending on who you believe (Gartner, Forrester, Butler etc) and this will mean we need a lot more BPM practicioners in a very short amount of time.

I still don’t think the guys at Blizzard have much to worry about.

Here are some of the more common questions I ask clients when trying to find out where they are in terms of their business process management maturity

Do you know what all your critical business processes are in your department or organisation?

Do you know what all your non-critical processes are in your department or organisation?

Is there a document that lists all the critical (and non-critical) business processes?

For each of these processes is there a process owner?

For each process is there a high level design document that outlines which steps are performed and by whom or what system?

For each process is there a low level design document that details what is required to perform each step?

For each process is there training material?

For each process are there defined and documented KPI’s?

For each process how often are the KPI’s measured?

For each process are they as automated as they can be?

For each process are they as integrated with other systems, and processes as they can be?

For each process what is its value and its cost to the business?

For each process how agile is it to change?

For each process how aligned is it with the needs or direction of the organisation?

For each process are management or users of the system given a oppurtunity to contribute to the increased efficiency of the process?

For each process are there scheduled periods to review the process to search for inefficiencies?

For each process is there significant reporting to highlight errors and exceptions, inefficient tasks, people and flows (preferably in real-time)?

Is there a methodology put in place for when changes are made to processes?

There are a lot more questions out there but these ones I find are usually a good place to start and obviously if they answer no to some of the early questions it means the later ones are usually not applicable.

It has been a long time since I have written anything but I have been really busy.  If anyone is interested I was just in CIO magazine in an article about workflow.  If you would like to read it you can visit: http://www.cio.com/article/148309/Making_Workflow_Work_and_Flow_for_You/1

Enjoy and happy 2008

In the following article a couple of guys discuss their reasons behind their view that Workflow and BPM are different.

http://www.transformmag.com/showArticle.jhtml?articleID=16400140

[Sorry the above link is no longer valid..I will continue to search for the original material]

I disagree with many of there arguments. 

 Nathaniel Palmer writes:

Workflow is concerned with the application-specific sequencing of activities via predefined instruction sets, involving either or both automated procedures (software-based) and manual activities (people work)…BPM is concerned with the definition, execution and management of business processes defined independently of any single application.  BPM is a superset of workflow, further differentiated by the ability to coordinate activities across multiple applications with fine grain control.

First off I have not heard of a modern workflow system that did not control the definition, execution and management of business processes defined independently of any single application.

Second workflow systems coordinate activities across multiple applications.

Third the phrase fine grain control does not differentiate between anything.  A bicycle is a bicycle whatever the size.

 Palmer then writes:  As workflow processes are tied to single applications, process flow is hardwired and does accommodate alternative means for reaching the same task or goal.

 Again I have not come across any modern workflow system where this is the case, perhaps he is confusing a system with workflow and a workflow system.

 Finally Palmer writes:  Distilled into single-word definitions, workflow is about repetition and BPM is about coordination (also automation and orchestration, respectively)

 So if workflow is not about coordination, automation and orchestration what is workflow about?

 Next in the article Jim Sinur gives his definition:

BPM is supercharged workflow that has sophisticated flow design through process modeling and analysis. BPM supports hyper volumes of work and numbers of users with sophisticated event-state engines for long-running business events and transactions.   It is user friendly, “insanely open,” has smart agility features supported by rules engines and utilizes commodity integration technologies.

OK I have highlighted all the rediculous sales adjectives here.  What he is saying essentially is BPM is workflow with more adjectives.  According to Sinur BPM is more supercharged, sophisticated, hyper, long-running, friendly, open, and smart workflow.  So he is either saying BPM is workflow with more adjectives in front of it.  A bicycle is a bicycle whatever the hyper-superchargedness it may have.

In summing Sinur declares:

Finally, BPM links to legacy composite components and applications, orchestrates Web services, measures business activity and optimizes processes for better business result and work throughput.

Again workflow systems have done all of the following for a long time now, except for web services which is a relatively new technology.  In addition Business Activity Monitoring in someform has always been a part of some workflow systems it just didn’t have such a fancy title.

Now this may be an almost religious belief but to me it seems that workflow + marketing = BPM, and that is the only real difference. 

PS let me apologize for the long delay between posts.  I have been away alot.

Hi I am back from a very interesting holiday to thailand.  Alot of fun but very touristy unlike when I was there 20 years ago.   I just am doing a quick post about the new workflow patterns website http://www.workflowpatterns.com .  For those of you who dont kow about this site they are effectively organising and naming all the different types of workflow patterns that they encounter.   Sort of like a botanist would the first time he or she enters a jungle.  And it is a jungle out there.  There are so many types of patterns out there and even more names for those types of patterns.  That is why this site is so important.  It is definately giving order where there was chaos.

PS you can even check out how Kontinuum satisfies the patterns (or doesn’t in some cases) since many, many new names for patterns have arised since our control patterns document was created by clicking http://www.workflowpatterns.com/vendors/kontinuum.php 

This is a history of workflow not as it actually is but as I understand it. 

The early years: 

A hominid grabs a stick.  With said stick he or she makes decisions on how to best maximize the usefulness of the stick.  After picking us several sticks they start to realize that not all sticks are created equal and eventually the start to carry the really good sticks with them so they can re-use the later.  Continually over time the stick is improved by sharpening tying rocks to it.  The adoption of the stick and other tools and there continuous improvement causes much success within the hominid community. 

After a while certain hominids became good at doing different things with sticks and rocks.  Some became good a making things of sticks into boats, others became good at turning rocks into swords; still others became good making food.  In short the hominids got jobs.  These jobs meant that in order for the hominids to achieve maximum success different people had to become skilled at different things but they still had to communicate with each other to ensure that the processes needed to keep their families and communities alive were followed. 

The middle years/ages 

At some point around 2000 B.C. in
India people came up with the idea of a guild.  A guild being a group of dudes who decide that they really know a lot about something and can deliver whatever incarnation of food, sticks and rocks that they work.  The also can deliver their products at a high quality.  Humankind adopts standards.  Not only do they adopt standards for quality but they standardize the methods in which they make their products. 

The Modern Age 

Some guy in France creates a new kind of standard, a specification.  While other standards govern quality and process this guy decides he wants to create guns to such of a standard that all his guns look exactly the same.  Not only do the guns look the same but bits from Gun A can replace the same bits from Gun B.   

I know what you are thinking, sure this article is incredible well constructed and probably deserves a Pulitzer Prize but are we really talking about workflow?Well yes and no.  I believe that at its core workflow is really about the things that I have talked about.  Getting the most of the tools you have, providing standards for production, eliminating waste, continuous improvement and re-usability.  While workflow software might be a somewhat new thing, workflow itself has been around almost as long as we have.

I was looking at one of my favorite sites today, the workflow patterns site by Wil Van Der Aalst, and it occured to me that if you are writing in a procedural language such as C you are effectively doing the same thing as writing out control patterns.  So then it occured to me that there might be some constructs in languages such as C or VB that are not covered by the existing list of workflow patterns (as I understand them).

 Here are some VB,C language contructs that dont really have the same kind of thing in the workflow patterns.

1) Goto:

Yes an outdated language construct and with good reason.  The goto is used to exit out of a context and goto another one and it pretty much forgets where it came from.  With the workflow control patterns they are alot more controlled.  They don’t skip all over the place as is the case with goto.

2)  Kill Process / Exit Application

If I am in a proces within a process within a process there does not appear to be a way to exit the high level process without creating a whole bunch of control logic with the currently defined control flow patterns.

3)  Wait

I am not really sure if this qualifies as a control flow pattern but it kinda is.  When getting to this stage just don’t do anything for a period of time (irregardless of what else is going on in the system).

4)  Queue and Stack

You could argue that both of these structures are not control patterns but for the sake of arugement I am going to argue that they are.  Lets say we have 3 tasks running in parallel 1.1, 2.1, 3.1

after one point 1.1 comes 1.2 after

after one point 2.1 comes 2.2 after

 after one point 3.1 comes 3.2 after

here is the catch however for queue:  1.2 can only be complete if and only if 1.1 was complete before 2.1 or 3.1 or x.2 is completed i.e. First In First Out

here is another possible scenario 1.2 can only be completed if it has not been completed and it is x.1 has not been completed later than it or x.2 has been completed. i.e. Last In First Out.  The interleaved parallel routing construct doesn’t exactly cater for this behavior.  You may argue that it is more of a resourcing pattern but I would need to be convinced.   You could also argue that this is more of a multiple choice, which in many ways it is.  However the logic at the point of choice could become very difficult to express at the point of choosing which x.2 item is next.