Between Scylla and Charybdis; or, the Customization Dilemma

There they were, sailing along their merry way. Toward the horizon, a narrow strait approaches. As the boat gets closer, they notice a couple of strange characteristics; to one side a cliff and the other a whirlpool. Upon arrival, it becomes apparent that this is the cliff where the monster Scylla dwells. Looking to the other side, the monster Charybdis, spewing out huge amounts of water, causing deadly whirlpools. Each monster is close enough that to avoid one means meeting the other. Determined to get through, our intrepid hero Ulysses must make a decision.  The idiom “Between Scylla and Charybdis” comes from this story.  In more modern terms, we would translate this to “the lesser of two evils.”

PLM administrators, engineering managers, and IT teams are often give this same choice with equally deadly – well, unfortunate – outcomes. What is this dilemma? Customize the PLM system (beyond mere configuration) to match company policies and processes, or change the culture to bend to the limitations posed by “out of the box” configurations.

Companies will often say something to the effect of “We need the system to do X.” To which many vendors meekly reply “Well, it can’t exactly do X, but it’s close.” So what is a decisionmaker to do? Trust that their organization can adapt? Risking lost productivity and possibly mutiny? Or respond by asking “What will it take to get it to do X?” incurring the risk of additional cost and implementation time.

We can further elaborate on the risks of each.  When initially developing the customizations, there is the risk of what I call “vision mismatch.”  To the best ability, X is described with a full understanding of the bigger picture that is missed when the developer writes up the specification.  This leads to multiple revisions of the code and frustrations on both sides of the table.  Then, customizations have the longer-term risk of “locking” into a specific version.  While gaining the benefits of keeping your processes perfectly intact, the system is stuck in time unless the customizations are upgraded in parallel.  Some companies will avoid that by never upgrading…until their hardware, operating systems, or underlying software systems become unsupported and obsolete. Then the whole thing can come to a crashing halt.  Hope the backups work!

However, not customizing has its own risks. What if the new PLM system is replacing an older “homegrown” system that automated some processes that the new system cannot? (And a “homegrown” system comes with its own set of risks; original coder leaves the company, never commented code, no specifications, etc.)  For example, raising an issue automatically created an engineering change request while starting a CAPA process. The company has gained a manual process, thus exposing them to human error. Or, perhaps the company has policy that requires change orders go through a “four-eyes” approval process, to which the new system has no mechanism to support such a use case.

Customizing is akin to Charybdis, whom Ulysses avoided, deciding that it is better to knowingly lose a few crew members rather than risk losing the entire ship to the whirlpools. Not customizing  is more like Scylla, where there is lower risk, though a much higher probability to the point of almost certainty.

We’ve been through these straits and lived.  We’ve gone through with many companies, from large multi-nationals to the proverbial “ma and pa” shops.  Let us help you navigate the dangers with our PLM Analytics benchmark.