Exceptions that prove the rule

by Andrew Hodgson

For business process automation projects in essential industries, exception handling can be a delicate balancing act.

On the one hand, no process, however cleverly designed, is completely bullet-proof. Exception management is necessary to ensure that unusual situations are handled effectively. If a system doesn’t raise exceptions when it should, problems will go unnoticed and the impact on customer service can be severe.

On the other hand, when a process has too many exception-points, it can generate exceptions at a rate and volume that simply overwhelm the support team. This not only creates backlogs that affect customer service; it also increases manual workload and hinders cost reduction.

It’s difficult to get the balance right, especially if exception management is overlooked at the process design stage. In the past, we’ve seen projects where a new system has gone live before it becomes clear that its approach to exception management is unworkable. This can lead to huge amounts of re-development, delays and dissatisfaction – if not outright project failure.

The best strategy, then, is to assess each possible exception-point on its own merits during the planning phase:

  • What value would this exception bring to the process? Does it help to improve customer service, or enable timely billing, or save costs – or does it just create more work?
  • Is this exception actually telling us something unique and useful, or just duplicating information that we already have?
  • Is this the best stage of the process to address this exception, or can it be caught and resolved earlier?
  • Can we reduce the incidence of this exception through automation? If not, what is the best way to handle the exception manually?

Performing this kind of analysis helps to prioritise the exceptions that are really necessary, and eliminate the ones that don’t add value. In some cases we’ve seen a list of proposed exception-points reduced by as much as 80 percent with this method – which can translate to similar savings on exception handling workload once the system goes live.

However, even the most careful exception design can’t completely guarantee success – especially in essential industries, where new regulations and new technologies mean that change is the only constant.

That’s why it’s important for exception handling to be as flexible as possible. Using a rule-based exception management system, where process owners can make changes to exception-handling logic without the need to recode underlying legacy systems, can provide a huge advantage.

With a rule-based solution, exceptions can be reviewed on a regular basis and tweaked to improve performance. If the business process itself needs to adapt to changes in the industry, the exceptions can easily be altered to match. And if certain areas are generating more exceptions than expected, it’s relatively simple to create more sophisticated rules that help to minimize the need for manual intervention.

When clients ask us about process automation, we advise them that all exceptions should be governed by rules. That’s our rule – and we make no exceptions.



Contact the author