Most executives that implement a PeopleSoft or SAP are surprised that productivity takes such a dive in the departments that these systems were supposed to automate. Departments that are dependent on the data see some productivity increase as information becomes more available, but many tasks that used to take a moment (or could, if you were pal-ly with the clerks) now takes a long time. The overall averages are lowered but the lowest times have increased dramatically. What’s going on?
One of the dirty secrets about process work is that almost no group follows the standard process. They’ll tell you that they do, but in the end they follow a whole slew of methods that they develop locally over time in response to issues. This tacit knowledge never gets fully captured by the requirements designers since they users themselves don’t explicitly know that they know it. (David Gilmour, in a recent HBR brief, reminds us that users also hoard knowledge because it makes them look good in front of the boss.) Thus, the new system lacks the full features of the previous one.
There are lots of terrible business processes out there. SAP and PeopleSoft implementations force companies to take a hard look at how things get done. Unfortunately, most of the time implementation modules are used to force a process on the company’s personnel. The personnel have to relearn all of the small tricks that they developed to get work done. It took years for the clerks to develop this institutional knowledge, albeit not explicit, and it will take years for them to develop new tricks.
Destroying local knowledge and process work-arounds is the leading cause of resistance to organizational change. The executives, who have never worked in a low-level job, force a change that reduces accumulated tacit knowledge to zero, meaning that the employees who were once so valuable are now on the same value level with new employees. Experience and its accompanying knowledge is thrown out the window, onto the dung heap below.
One interesting last point: Agrinomist and anarchist James Scott wrote Seeing like a State : How Certain Schemes to Improve the Human Condition Have Failed (Yale ISPS Series) which describes the process that states must use to gain control over things. He makes the point that standardizing is a natural form of Hyper Modernism. It also often really doesn’t work. It’s why so many Great Projects create so many problems. They don’t look at the situation on the ground.
It’s also why Positive Deviance methods can solve problems for almost nothing that stymy the best-funded agencies.
Unfortunately, things have to be rationalized before we can really turn to letting everyone be themselves. That seems counter-intuitive but it’s true. The industrialization process meant standardizing in a horrible way. People had to show up for work at a particular time. They had to work at the same time. They got paid the same pay for a days’ work or by piece. The reason is that it’s easier to keep track of things when you do that.
Decades later, we all are trained to use watches, show up on time, use a common language, adhere to wildly complex rules (think driving), in ways that you don’t get unless you try to put a new operation in a society that doesn’t have largescale efforts.
But today this isn’t necessary. We can and should enable local custom processes because it’s more efficient. We don’t have the same limitations that we did. We have massive computing power available to us. These aren’t individuals all doing their own thing.
Man, there’s a lot to talk about on this. Open Source Software, joint ventures, “a thousand points of light”….