The Digital Tipping Point

McKinsey just published an interesting survey on the adoption of digital technology at organizations.  In a word, things are getting more digital all the time.  (To learn more about the survey go to www.mckinsey.com/insights/business_technology/  and search for The Digital Tipping Point.

As I skimmed the survey I kept thinking about the role of BPM in all this.  BPM isn't mentioned in the article, although they inquire about automation, innovation, and improving operations.

Consider a bit of history.  When computing began, software developers created very specific applications, integrated with their own databases.  Thus we had the Payroll Application, and the Customer Database, and, eventually, the spreadsheet program. By the 90s, it became obvious that all these stand alone applications were limiting what companies could do and a huge effort has been mounted at most organizations to integrate these various specific applications.  Much of that effort has focused on ripping out home-grown applications and substituting some variety of ERP.  In mosts cases the ERP programs aren't as useful as the programs they replaced, but they have the immediate advantage of sharing their data with other applications so the inconvenience is deemed worth it.

Today, with relational databases, the Internet and new, standard protocols in place, most companies expect to continue their march toward digitalization.  Unfortunately, as Michael Hammer proclaimed in the Nineties, if you pave cow paths, you end up with inefficient processes.   Most organizations have made some effort at process improvement.  Unfortunately, however, most organizations are still only at CMMI maturity level 2.  They have some major processes documented, but they don't have an integrated process architecture for the entire organization.  More important, they don't have integrated value chain processes with good measures and managers assigned to oversee the continuous improvement of each value chain.

Without a lot more work on processes, most companies that struggle to automate their current activities are going to find that they are building more silos and more barriers to efficiency.  A quick look at most of the BPMS applications that have been built in the pasts decade will confirm this.  With a few exceptions, most BPMS applications have been smallish applications that might as well have been done with a pre-2003 workflow tool, or a 4GL tool.  They automate a bit of a value chain, but not the whole thing.  They automate some specific activities, but don't provide hourly measurement information to managers designated to manage the value chain.  They are, in essence, just applications like the ones we might have developed in the Nineties, with some new bells and whistles.  And IT can monitor and update them online.

The ultimate task of rethinking how an organization works is not the job of IT but the job of the business executives running the organization.  If business executives try to delegate the analysis and improvement of their value chains to IT they will get paved cow paths, not because IT is unwilling or perverse, but simply because it isn't IT's job to re-imagine the way a business works.  IT automates what exists, not what might exist if it were reconceptualized.

Most businesses have yet to make the kind of process effort that is required.  Most business executives still don't really understand the process perspective and have yet to drive process ideas down into the core of their organizations.  As a result, digitalization is a nice idea, but it isn't going to provide the kind of improvement that executives expect when they delegate it to a IT team.

BPMS has the potential to revolutionize business.  The most exciting companies today are using variations on BPMS to develop more and more comprehensive and integrated understandings of their organizations and laying the groundwork for a time when an executive can look at a diagram online and see hourly measures of the flow of goods and services throughout the entire organization.  A few organizations will achieve this vision in the coming decade, and the rest will find themselves suddenly trying to catchup.

Those who still don't get it will imagine that they can hire IT groups to digitalize their organizations.  But they will be wrong then, as so many are now.  To digitalize an organization, properly, you first have to understand what you are trying to digitalize.  If you try automating an activity here and a subprocess there, you will simply use the latest technology to pave the current cow paths.  To digitalize an organization, properly, whether you do it now or in the future, first requires a clear understanding of the processes you are trying to maximize, and the bottlenecks you need to monitor to understand your flows.  At this point in time, digitalization  must support active management and control and continuous improvement.  Anything less isn't going to let your organization survive to the middle of this century.