In a perceptive article on ZDNet, writer and analyst Joe McKendrick makes the point that in the modern era, every company is a software and data company. He says:
“That widget manufacturer down the street likely has a data center and has developed and maintains a portfolio of its own custom software that keep evolving.”
His point is that even companies far removed from the space traditionally associated with information technology now rely heavily on data, analytics, and bespoke software to do their work. It's position I wholeheartedly agree with, but, of course, I'd take issue with the idea that every widget manufacturer has, or needs, their own data center.
Perhaps five years ago that was the case — smart manufacturers who were ahead of the curve where data is concerned may have invested in building their own infrastructure to handle data and analytics, but five years ago we lived in a very different infrastructure environment. The cloud was a young technology and it didn't have the trust of enterprise businesses — it had not yet proved either its usefulness or its security.
Since cloud technology matured, I'd argue that it is a mistake for "widget manufacturers" to invest heavily in in-house technology. Clearly, if your core business is manufacturing, design, product development, media, services, and almost anything else that's several degree removed from web services or technology, computing infrastructure management isn't your core competence.
Infrastructure management skills are at a huge premium. Infrastructure engineers are expensive and highly sought after. The cost to a traditional manufacturing company of hiring a team to put together an in-house data center that can compete with the cloud is prohibitive for all but the largest of companies. The time and expense of cultivating in-house infrastructure, network, and system administration expertise has never been higher.
But most companies have absolutely no need to cultivate that sort of talent pool. Infrastructure-as-a-service companies like Outscale have already done all the hard work of forging teams and infrastructure and made it available to other companies at a fraction of the cost of building in house.
Cloud providers benefit from economies of scale that other companies do not. They build data centers on a scale that other companies cannot. And because providing infrastructure is their business, they are incentivized to invest heavily in bringing on-board technical experts who are leaders in their field.
So, although I'd agree with the basic premise of McKendrick's article — every company is a data company — I think he misses the most important part of the transition to data-centric business: the cloud. The cloud has both driven the growth of data-focused business models and provided the infrastructure that businesses need to efficiently leverage data.