BPM: Do we detect a process or invent it?

This was a question on ebizq.net at the start of the month, and when I first read it thought “I wonder how many people go for invent, or talk about abstraction, or the fact that workers are cogs in a process…” You can see the discussion here:

http://www.ebizq.net/blogs/ebizq_forum/2010/08/are-business-processes-detected-or-invented.php

 I have read through the discussion, and haven’t really wanted to make further comments because I felt that the discussion moves away from the question far too much. Typically us BPM thinkers like to “abstract” everything so we are not talking about a single process example, rather something that can be used across all processes within any business or market sector. However, though this is great, and I always go for abstraction and generic based thinking, it really doesn’t help a) answering straight forward questions like this one and b) making life easier for business decision makers in adopting BPM.

So what are the arguments for invention of processes and those for detection?

Invention

There are few times that as a business, you get to “invent” a brand new process, something that the business has never done before. The only time you may do this is if you are a new start up, or a company that has spread its wings into new markets, or a company that is re-structuring the way it operates. With this in mind, many process inventions aren’t actually new, rather they are tweaks of existing “known” processes. But let’s think “BPM” implementations and getting the process created. In this case I include designing processes as invention, processes that we are aware of and have had the luxury to address, refine and then define as a business process. These are invented / designed processes.

Typically a BA (Business Analyst) will have knowledge of a particular process. This is then analysed and then re-designed with maximum efficiency in mind. This can be communicated through a number of tools, visual aids and BPM type languages, but I won’t get bogged down with that. Rather, the process is now designed and ready to be implemented in our BPM platform.

For many, this is the best or only way to implement BPM solutions. So even though these processes are tweaks or conversion from existing processes they are “invented”.  There are many arguments that to design / invent a process, you must first detect it and have knowledge of it, I think this is simply an obvious statement. When talking BPM processing, Invention Vs Detection I take as how these processes came to light and how they end up in our BPM platform. As ever, let’s keep things simple.

Detection

It is common knowledge that organisations don’t know all the processes that they execute, it is also common knowledge that trying to design and implement all processes in a BPM platform will mean you only actually implement a small amount of processes carried out. So what does this mean in terms of invention? Well it means that invention alone will account for a fraction of the processes. So if you believe invention is the only way we can use BPM, then your organisation is missing out the benefits of BPM across any number of processes. Invention addresses known processes, we need something else to allow our BPM platform to detect and find the other processes out there.

Employees are paid good money to act as knowledge workers, they make key decisions and yet for many in the BPM world they are viewed as simple “cogs” in a process machine. If you take this view of people in the process, then you have to state that processes are invented rather than discovered. Since we know that designed and invented processes don’t make up all the processes out there, we know this thinking is simply wrong. This means that detection is key to successfully implementing a good BPM platform, and is key to making the most of your investment.

So how do we detect processes? There are many ways, some argue chasing a paper trail is good, which is true to an extent, however many processes may be in a person’s head and don’t involve paper. How do you detect these? Some talk about desktop activity, yet again this doesn’t capture the complete truth either, missing manual activities…So what is the answer.

I would argue that you need a BPM platform that is flexible / adaptive enough to allow your knowledge workers to update it. Even with designed processes, we see that elements have been missed or need tweaking. In such cases, knowledge workers with an adaptive platform, discover new processes / sub processes. They then update the process to make it “correct” (if they have permission and if indeed their updates are needed). The same applies to processes that are completely unknown. In these cases, knowledge workers detect unknown processes. These are then created within the BPM platform and made available for all, again if they are needed. Detected processes can then be reviewed and streamlined / re-invented because we now have knowledge of them.

The key to detection is choosing a BPM platform that is adaptive and trusting your workers knowledge.

Conclusion…

The question implies one is better than the other, or that in reality one thing happens without the other. The truth is, as an analyst or an organisation implementing a BPM platform, we do, and need both, “invention” and “detection”.

Invention is the starting point for implementing BPM. It is also the ultimate goal, to be able to design your processes to maximise efficiency, however in reality, this cannot be done for the majority of processes, simply because they aren’t known yet and/or the amount of time spent on analysis would be vast. So, in order for us to get as much knowledge about processes, and to therefore really maximise BPM in an organisation, we have to view detection as a massively important feature of the long term success of BPM.

This detection requirement is why I am a strong advocate of adaptive processes and adaptive platforms. It is also why I don’t like “lean” type solutions, or solutions that rely heavily on design based maps. These types of solutions typically presume that all processes are invented, and we know this simply isn’t the case…