For a good decade now, most IT organizations have been strongly focused on business process design. With businesses investing big bucks in ERP, supply chain management and customer relationship management systems, process design has been king, and in many cases it has met real business needs. But more and more, with those implementations mostly behind us, process design will take a back seat to other skills.

The best way to illustrate this is by example. A few years ago, a bank decided to get serious about managing its retail customer relationships more effectively. It installed a well-known CRM package that would give branch bankers all the customer data they would need for effective marketing; they would be able to offer the right products every time they interacted with customers.

But the expected growth did not occur. Calling on its best business thinkers and an able IT organization, the bank redesigned its processes. Everyone involved was certain that they had defined a branch process that was complete. It wasn't, and in fact, it made things worse. Customers began leaving in droves, often closing their accounts shortly after interacting with an account manager. The process design seemed to be having the opposite effect from what was intended. Where had it gone wrong?

To its credit, the bank reacted quickly, conducting a customer survey that started to reveal some answers. It discovered that customers were finding even the simplest of transactions time-consuming, aggravating, intrusive and overbearing. If they wanted to convert cash into certificates of deposit, the account manager, guided by the CRM, bombarded them with questions about every aspect of their finances. If they wanted to upgrade their credit card to one laden with extras, they were grilled as thoroughly as someone taking out a mortgage. Opening a second account ended up being a tedious rundown on all the financial products the bank offered.

Talking to account managers, it became clear that they shared customers' frustration and felt helpless to escape the rigid process design.

The bank once again turned to IT, but this time the focus would not be on the process itself, but rather on the imperative of fulfilling customer requests. IT undertook an architecture it called "the really good bank." It called for radically simplified proc­esses. The CRM system remained, but the rigid workflows were removed.

Parallel approaches were built in: If a customer asked about a CD, various options were calculated based on the computed "value" of the customer and presented immediately to the account manager. A purchase became a point-and-click transfer of funds, as fast as a withdrawal. All account types were bundled under one master account number, so new products were easy to add.

These few changes eliminated almost all customer aggravation. Signing up for a new product or changing the terms of an existing one became 30-to-60-second transactions that any teller could perform. The CRM system still presented the account managers with products to offer, based on what the bank "knew" about the customer, but it was less obtrusive. Account managers could comment quickly about some product that was available and even supply the customer with a flier while the transaction record was being printed. And if the client was interested, an equally fast process would get the customer signed up for the product then and there.

Process design works against this kind of revamp. IT needed to step back from technology and envision the future. The credibility this brings opens doors.

Bruce A. Stewart is CEO of Vancouver, British Columbia-based Accendor Research, an advisory services firm focused on management issues in the technology-enabled enterprise. He can be reached at bruce.stewart@accendor.com.