Many organizations struggle to attack end to end processes from the top down during normal economic times, but today it is even tougher with the pressure to cut costs quickly. It makes justifying an attempt to do a large scope process almost impossible. There is also a danger of doing multiple BPM projects from the bottom up, in that, this promotes functional efficiency and excellence over organizational efficiency and excellence. Sometimes multiple bottom-up efforts clash, over the long haul, requiring potential rework when the important end to end process is attempted. This is known as sub-optimization. What is the answer?
Try the “T” modeling approach. This is where a basic thin model is done in an “end to end” fashion to set up an optimal bottom line impact identifying smaller scoped process. This means that you scope sub-processes in light of the overall process instead of bottom up process improvements without a proper process context. “T” modeling that requires the process design be done as a shallow cross functional exercise. This is then followed up by completing deeply focused process implementations that deliver maximum time boxed results without sub-optimizing the overall end to end process
There are some important methodology, tactics and technology implications to this approach. First and foremost, this puts premium on thin logical modeling that is quite often tied to architectural efforts. Some of today’s BPM vendors do not support this directly, but have partnerships with popular process modeling tools such as IDS Scheer, Casewise, IGrafx, Vision Wave and Nimbus so on and so forth, but others have capabilities like Lombardi, Metastorm, G360, and Tibco . There also a need to handle sub-processes that may have been implemented in numerous BPM execution engines, content engines, workflow tools, and activity monitoring capabilities. Typically the leading independent BPMS players support this kind of heterogeneity. I believe that the competitive atmosphere, world wide and the pressure to do more with less despite will require organizations to make moves before the standards are ready and fully supported.
5 responses so far ↓
1 Konstantin // Apr 7, 2009 at 4:08 am
Couple of questions:
in which way a classical BPMS of TIBCO (iProcess Suite) can support alignment of implemented processes with logical modelling and tie it to a much larger Business Architecture?
And how sub-processes implemented in say Pega can be handled (managed) by another BPMS like TIBCO?
Thanks for clarifications in advance.
/Konstantin
2 Jim Sinur // Apr 7, 2009 at 9:38 am
In order to have a deep and complete architecture linked to processes, you will likely link a EA (enterprise architecture) tool to the implemented process. There are a number of ways of doing this. See http://blogs.gartner.com/jim_sinur/2009/03/19/policing-model-synchronization-pros-and-cons/. If you have a light EA need, then you can point to the process instances in something as simple as word document. It depends on your desire for purity.
Sub processes implemented in foreign BPMSs will have to be wrapped and invoked. This is because these players are not interoperable through standards yet. There are very few vendors that stick to a pure standards based approach. It would be best to implement in one BPMS to support an end to end process, but not necessary. “T” modeling is a design method and it works independent of interoperability. Again, It depends on your desire for purity.
3 The Top Five BPM Starting Points Today // Apr 14, 2009 at 4:14 pm
[...] are my top five BPM starting points, but be careful not to sub-optimize now and pay later. See http://blogs.gartner.com/jim_sinur/2009/04/06/end-to-end-processes-that-fit-you-to-a-t/ How about [...]
4 What is the Greatest Hurdle Facing BPM? // Apr 29, 2009 at 6:19 pm
[...] If BPM sticks to small ideas and scopes to just save time and money while bettering the customer experience, it will be relegated to less than its potential going forward. Expanding BPMs influence to innovative end to end processes that are linked to important value chains will test what BPM really brings to the party. This means that BPM will have to play well in multiple contexts and be plugged into people and information. Please see http://blogs.gartner.com/jim_sinur/2009/04/06/end-to-end-processes-that-fit-you-to-a-t/ [...]
5 When is T&T Just Dynamite? // Jul 20, 2009 at 8:15 pm
[...] http://blogs.gartner.com/jim_sinur/2009/04/06/end-to-end-processes-that-fit-you-to-a-t/ [...]
Leave a Comment