InformationWeek reports Red Hat CEO Jim Whitehurst’s condemnation of the software industry. Whitehurst catalogs a long list of industry ills including overpricing, a lack of productivity improvement in development organizations, and unimproved quantities of program bugs in commercial software.
The main problem, according to Whitehurst, is a commercial development model under which executives, programmers, and marketers get together in an effort to predict what their customers want-and then take five years to build it. As a result, “half of all IT projects fail,” said Whitehurst.
Scientific socialism implemented via five year plans didn’t work out for the Soviets, so it is no surprise that it doesn’t work for software either. The cure is to actually listen to what software customers need, then quickly deliver the results. (Sounds like HarrisData’s Collaborative Development and general approach to customers).
Repeating SAP mistakes? After avoiding one Soviet era pitfall (central planning), Whitehurst stumbles into another – Gigantism. Gigantism is the idea that one humongous manufacturing plant is better than many small or medium sized ones.
“All I need to worry about is whether our architecture participants include Google, Yahoo, and Amazon,” said Whitehurst, referring to some of Red Hat’s biggest customers. “Our solutions aren’t based on our ideas about what you want. Our customers are building them,” he said.
If the giants are happy, the masses will be too! This is the same problem that contributed to Leo Apotheker’s exit from SAP where focusing the on 10 largest customers led to mass dissatisfaction among the rest of SAP’s customers. How is replacing your own planning committee with a few giants going to produce software that your many large, medium, and small customers need? Why not listen to them?