Honey I just blew up the ERP
Good thing about working from home is that I can watch webstreams from conferences - these days it's the fun they're having up in Davos.
Obviously much is about the crisis, what happened, how to fix it and how the heck did it happen. More and better regulation being one of the main themes of course, following the human inclination to grab more control when things goes... ehh.... out of control!But one thing is missing from the discussion - finding the root cause for it all. Is there a single thing that lies behind it all? If found all would be much easier, at least the issue of how to avoid such calamities again.
Is there a root cause? And if so, why is it not discussed?
So why is it not discussed? Because sometimes we take some things for granted, we cannot imagine the world without that something, we in fact forget or even cannot answer this simple question:
What assumptions are we making that we do not know we're making?And the root cause could be what? Allow me to posit that it is:
Double Entry Book Keeping.Not the accounting reports as such, I'm fine with those, it's the method of capture and representation of facts that I see as the culprit.
Sig has been poking away at the notion of accounting for several years now. He's right to do so. A system invented some 600+ years ago and which hasn't evolved since then can hardly be the basis upon which modern decision making is taken. Or is it one of those immutable 'truths' to which we cling in life?
As one who was trained as an accountant, I know how hard it is to give up things that seemed for so long to make perfect sense. But then time and again we see failures and question marks that have their root cause in something accounting related. Sig amplifies his argument with numerous examples, some of which I see others, I think are fraught with problem. His assertion for example that:
That is how a few well placed misplacements or closed eyes at Satyam could make them pretend to be much richer than they were.
...sounds a bit of a stretch when what we're really seeing is a web of deceit that encompasses many actors. Even so, the fundamental technical flaws that allowed Satyam to occur are there. For instance, the ability to allegedly inflate the number of employees, pay them and not have that register as an operational blip is a function of what you can do with book-keeping trcikery. It's called 'pass through' and is relatively easy to manipulate and hide from prying eyes when there are lax controls in place or where it is possible to fool people.
Sig's counter argument is to question the whole way 'we' represent transactions in systems:
Instead we need a direct and dynamic representation of the "transactions", something that is real time and where changes anywhere in the chain are precisely and immediately reflected everywhere independent on principles and rules.
In Sig's world, answers come by defining 'objects' according to their attributes and then ascribing relationships to them - example: widget (object) and transaction becomes a 'changed relationship.' It's a very simple but powerful notion and one that has huge ramifications.
For instance, yesterday Dick Hirsch was noodling on the intersection between social computing and business process. He starts by asserting:
Processes are usually static entities with a certain number of steps / tasks that may be spread over various roles / individuals. Of course, there might be various paths through this process but these paths are known. This static quality is necessary to assure that corporate "behavior" in the form of processes is repeatable and thus, measurable. Without this structure, "chaos would reign" and employees wouldn't know what to expect when confronted with a problem.
Dick is a business process expert at Siemens with whom I have worked. He is super bright, intensely analytical and able to parse complex problems in ways that most people will readily understand. He's also handy in the code writing stakes. Back to the plot.
Dick's view is fine except that's not how life tends to work in the real world. Or at least in Sig's. Process locks us down and prevents us from finding optimal or new ways of looking at the world. Only rarely do processes change for the better, usually on the grounds that inter-dependencies are too complex to risk changing 'the system.' Well yes and no. Dick attempts to come up with a solution that requires something of an intellectual leap of faith into the world of Web 2.0 style technologies:
...dynamic process definition would lead to process evolution in which more efficient processes might be more likely to survive. Such an evolution is only possible if a newly designed process could somehow be saved and performed again. Thus, there must also be an avenue to move these newly defined processes into the existing process landscape. If this is not done, the benefits of this new process might not be widespread. Therefore, these Web 2.0 tools need tight integration into existing process technology to support this transition.
This is the point where I disagree with Dick. There seems to be an implicit assumption that processes need to be in a continual state of refinement or lockdown when experience suggests that 80% of issues arise because of defects in process or unusual situations that cannot be addressed with pre-existing process definitions and tasks.
Capturing and acting upon the semantics of what defines those issues is where I believe Sig and Dick's thinking could usefully intersect. Sig prefers to talk about Barely Repeatable Processes (BRP) where the unusual requires attention but may be at the end of the long tail of process repeatability. Dick makes the intellectual leap to assert that social networks can be used as a pathway to solving those problems through crowdsourcing. I'm on the fence.
I've seen little evidence that crowdsourcing of that nature has any real value but then we are in the very early days of understanding how these networks operate internally to the organization. I've been in teams where crowdsourcing among a handful of the right people DOES work but I'm not convinced that is a proven model. It may be the start of some model that can be articulated and acted upon. It is one that will require considerable help from organizational social psychologists on a scale many will find intimidating if not downright intrusive. In the meantime, SAPper Steve Winkler also expresses disagreement with Dick:
...this kind of innovation is not for the faint of heart, because there's a decent chance you can get lost along the way. You can try something new and find out the short cut is actually a long cut.
aaah - the ol' status quo argument. Steve's point is well made in the context of organizations having spent millions on defining and refining business processes with the intention of driving efficiency. But in the 21st century, efficiency isn't enough. If it was then we would presumably see less layoffs? Effectiveness is what's called for and that almost certainly requires levels of collaboration and innovation that we've not seen in the past.
I'm sure Oliver Marks has plenty to say on this topic but I'll stop here and leave the ZDN community with these questions. What if Sig AND Dick are right? What if this means we can leave behind the highly constrained world of process built around rigid principles? Could this allow business to flourish in a world of constant innovation and if so, where is that best undertaken?