Galen Gruman
Executive Editor for Global Content

2008 InfoWorld CTO 25: Bill Chapman, Avnet

feature
Jun 2, 20082 mins

An acquisition almost goes bad -- and the rescue effort teaches a valuable lesson about SOA development

After electronics distributor Avnet bought Access Distribution, it became clear the expected “cost synergies” — redundancies that would create quick savings — weren’t there. The problem: Access’s custom pricing engine was enormously complex and couldn’t be easily ported over to Avnet’s infrastructure, leaving Access’s datacenter and applications in place. Yet Avnet’s management was counting on gaining those “synergies” to make the acquisition worthwhile.

[ Discover what insights you can take advantage of from the other 2008 InfoWorld CTO 25 winners. ]

Much of the challenge involved data architecture, figuring out how to handle the fundamentally different pricing models in the two companies — both of whose product catalogs easily surpassed 100,000 items.

Another significant challenge involved the people. “The biggest stress point is how the team views itself,” Chapman says, especially when that team now includes people from the acquired company whose processes — not just infrastructure — are largely being tossed out. “You can’t be hypercritical,” he says; instead, you should use the acquisition to identify “applications of services to a business case you hadn’t thought of.” (Under Chapman’s leadership, Avnet was an early pioneer of SOA, starting in 1999, so its own pricing engine’s development had been done as services.)

This open-minded, empathetic approach needed to apply not only to the acquired employees but also to services developed at Avnet in the past that were now being reworked to accommodate the new pricing approaches, Chapman notes. “It’s not that the person’s original service isn’t valid — it was done for a different purpose, and you’re now using it beyond its original scope, so of course you need to tune it.”

The pricing engine effort exposed a need for future services development to make an effort to anticipate future uses earlier, so at least not to inadvertently preclude such uses in the design or implementation, Chapman says. The good news is that the IT team quickly adopted this approach: “It’s almost a gaming process — a Sherlock Holmes detective process — and it’s become a catalyst for team-building.”